Home

hp model 10 / 20 service manual

1. Fan Power supply pack units A P4 Pd S Pd IND GENE m We pao P 4 VA LO i Storage Rear view processor modules Figure 8 3 Model 10 Rackmount FRU Locations Model 10 Rackmount Chassis 8 5 punoouse qnouj Model 20 Rackmount Chassis Power supply units N Storage processor Disk drive slots modules 7 L 4 X N Pd Pa B A Fan 7 BBU pack i Front view Rear view Figure 8 4 Model 20 Rackmount FRU Locations 8 6 Model 20 Rackmount Chassis Front View Disk Array LEDs Several FRUs within the array contain LEDs Figure 8 5 shows front view LEDs Deskside Front View ll Jj I TEIL BOUTIQUE a _ Power LED green B Fault LED amber Rackmount Front View ower LED green Fault LED amber Disk Drive Slots Disk Module LEDs Ready LED green I Busy LED green Fault LED amber
2. Test ID Test Description DP Display Pass Count HE Help on Tests Commands HEX Help Extended 182378 ISA Bridge 182378 Tests DIR KBD87303 kbd87303 Keyboard Mouse Controller Tests DIR L2CACHE L2 Cache DIR LA Loop Always Mode LC Loop Continuous Mode LE Loop on Error Mode LF Line Feed Mode LN Loop Non Verbose Mode MASK Self Test Mask NCR1 NCR 1 53C8XX SCSI I O Processor Tests DIR NCR2 NCR 2 53C8XX SCSI I O Processor Tests DIR NCR3 NCR 3 53C8XX SCSI I O Processor Tests DIR NCR4 NCR 4 53C8XX SCSI I O Processor Tests DIR NCR5 NCR 5 53C8XX SCSI I O Processor Tests DIR NCR6 NCR 6 53C8XX SCSI I O Processor Tests DIR NV Non Verbose Mode PAR87303 Parallel Interface PC87303 Tests DIR PC16550 UART PC16550 Serial Input Output Tests DIR PCIBUS PCI PMC Generic QST Quick Self Test DIR RAM Random Access Memory Tests DIR RTC MK48TOx Timekeeping DIR SCC Serial Communication Controller Z85C230 Tests DIR SE Stop on Error Mode SRAM Static Random Access Memory Tests DIR ST Self Test DIR 8 14 Power Up Troubleshooting Table 8 4 Disk Array Power Up Tests for the SP620 SCSI SP only cont d Test ID Test Description TACH Tachyon Fibre Channel I O Processor Tests DIR VGA543X VGA Controller GD543X Tests DIR VME2 VME2Chip2 Tests DIR Z8536 Z8536 Counter Timer Input Output Tests DIR ZE Zero
3. Bracket Se retaining lt screws AC harness Figure 9 49 Removing the AC Harness Bracket Model 20 Deskside Component Removal 9 85 jueuie2e deH pue jeAoway Figure 9 50 Removing the AC Harness Plugs 9 86 Model 20 Deskside Component Removal On Off Switch Figure 9 51 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Remove the DC power distribution panel Figure 9 58 Note Record wire orientation before removing wires 3 Disconnect the four spade lug wires from the switch 4 Remove the on off switch from the distribution panel a Depress the retainer clips and push the switch out Note To install the on off switch press it into the distribution panel Use the flat blade of a screwdriver to depress the retainer clips Four spade lug nes Ss To wiring harness Retaining clips 2 on each side of switch Figure 9 51 On Off Switch and Power Plug Assembly Model 20 Deskside Component Removal 9 87 jueuie2e deH pue jeAoway Model 20 Rackmount Component Removal Remove the Array from the Rack Figure 9 52 CAUTION For these procedures the disk array must be powered off and unplugged Also it is recommended that ESD procedures be followed Use ESD kit P N 5182 4119 The disk array should be removed from the rack only
4. A5133A 17 8 GB Not supported 9 25 00 8 02 9 55 01 1 73 A5169A 4 2 GB Not supported 9 25 00 8 02 9 55 01 1 73 FAN x If a failure occurs in any fan a fault displays and all other fans speed up to compensate If FAN A or FAN B fails it must be replaced as soon as is possible Status codes are as follows OK Operating normally Failed Power Supply Unit VSC x The model 10 disk array uses one or two power supplies A and optionally B The model 20 disk array uses two or three power supplies A B and optionally C Also the disk array continues operating if one power supply fails but shuts down immediately if a second power supply fails If a power supply fails it must be replaced as soon as possible Status codes are as follows OK Operating normally Failed or not present If it is not apparent that a power supply has failed examine the fault LED on the VSC or look at the Unsolicited Event Log in Grid Manager BBU Each disk array supports a BBU The BBU is optional and is needed to enable cache Status codes are as follows OK Fully charged Presentation Utility Screen 6 39 Jobeue N pus NR Not ready charging After a power outage a BBU takes about 15 minutes to recharge From total depletion recharging takes about two hours The service fault LED is lit if the cache size is set larger than 0 and the BBU is not ready or not present Failed or not present ADP n BDP n UDP n
5. DISPMAP Displays a list of Logical Units LUNs found DOWNLOAD Downloads SP microcode and disk firmware to the specified device FRUINFO Displays information about a specific DISK Field Replaceable Unit FRU INQUIRY Performs a SCSI INQUIRY call and displays the information returned MODESENSE Performs a SCSI MODE SENSE call and displays the information returned UNBIND Performs a SCSI MODE SELECT call to unbind a LUN UNSOLLOG Displays the content of the Unsolicited Event Log for the selected SP UPDATE Automatically searches for and downloads SP microcode and disk firmware to the appropriate devices Other supported commands HELP Provides a detailed help on all the commands RESET Resets NIKEARRY and displays a list of LUNs UTILINFO Provides general information and hints about using NIKEARRY NIKEARRY 7 47 sonsouDeig pue seni ISOH Quick Start Instruction Executing NIKEARRY This example illustrates a typical use of NIKEARRY ODE gt NIKEARRY KKKKKKKKKKKKKKKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK kk kk Sk kk Sk KKK KKK KkKKKKK KkKKKKK KkKKKKK NIKEARRY KARKAR KERERE kkkkkxk Dee yer Copyright C 1994 thru 1998 by Hewlett Packard Company EM ASESOR All Rights Reserved WORKER ckckck ck KKK KKK Kekk This program may only be used by HP support personnel and KARNAN KERERE those customers with the appropriate Class license or TEREE E
6. Column heading Description Type Vendor LUN type or disk vendor as appropriate Product ID LUN or disk product identification as appropriate AprxCap LUN or disk approximate user capacity as appropriate where AprxCap Gbytes user_bytes 1 000 000 000 BSz LUN block size B R Build Rebuild percentage If information is not obtainable for example LUN not supported or disk faulted the unobtainable information is reported as a The disk section presents detailed information about the disk and vendor 7 76 Support Tools Manager STM SEAGATE SEAGATE SEAGATE IBM IBM IBM IBM IBM IBM IBM IBM IBM IBM IBM Product DGHSa8SY DGHSB83Y DGHSB83Y DGHSB83Y DG6HSBS3Y ST15158N ST15158N ST15158N DCHSB84F DCHSB84F OCHS SF OCHS SF OCHS SF OCHS SF OCHS SF OGHS18Y DGHS18Y DGHS18Y DGHS18Y amp Information Tool Log for C2300 Array on path 874 8 CLRRBS CLRRBS CLRRBS CLRRB3 CLRR83 CLRRB4 CLRRB4 CLRRBS CLRRBS CLRRBS CLAROS CLAROS CLAR18 CLRR18 CLRR18 CLRR18 Bab Bab Bab 8gb 2gb 8gb 2gb 2gb 2gb 8gb 8gb 8gb 8gb 8gb 8gb 8gb 8gb 8gb Serial Number 13008868 138855ED 13888585 138855E6 13888571 46235894 880435384 880436373 58873823 13143183 13131773 13867228 58187435 58127172 58141756 138842R5 13885ESE 138854F7 13885407 Figure 7 6 Information Log Screen Disks Section For more information
7. Ejectors Figure 9 9 Removing an SP Model 10 20 SP Removal Replacement 9 17 jueuie2e deH pue jeAoway Figure 9 10 Replacing an SP 9 18 Model 10 20 SP Removal Replacement Model 10 20 Power Supply Unit Removal Replacement Replace a power supply unit VSC by performing the following steps 1 Removing a power supply unit Figure 9 11 a Open the fan pack b Grasp the power supply unit by its handle c Hold down the latch d Pull the power supply unit out of the chassis 2 Installing a power supply unit Figure 9 11 a Locate the slot where the power supply unit is to be installed b Remove the power supply unit filler module from the slot c Slide the power supply unit into the slot until it engages the connector d Press firmly until the power supply unit seats fully into the chassis Model 10 20 Power Supply Unit Removal Replacement 9 19 jueuie2e deH pue jeAoway Figure 9 11 Removing or Installing a Power Supply Unit 9 20 Model 10 20 Power Supply Unit Removal Replacement Model 10 20 BBU Removal Replacement Install a replacement battery backup unit BBU by performing the following steps 1 Removing a BBU Figure 9 13 a b c d Open the fan pack model 20 only Grasp the BBU by inserting thumbs in holes and forefingers into latches Slide the latches inward and hold P
8. 0 cee ee eee ee eens 7 55 Disk FRU Map Model 10 Rackmount 0 0 cee ee eee eee ee 7 55 Information Tool Menu Options 0 0 eee eee eee eens 7 71 Information Log Screen Map Section 0 0 e eee eee eee eens 7 72 Information Log Screen Configuration Section 0 0 e eee eee 7 75 Information Log Screen Disks Section 0 0 eee ee eee ee eee 7 77 Activity Log Screens ioe senri erpii ag fX eee eke QE ee be 7 78 Firmware Update Tool Top Level Menu 0 0 eee eee eee eee eee 7 79 Current Revisions Display 0 0 eee eee een eect eee 7 81 xix sainbi4 7 10 7 11 7 12 7 13 7 14 7 15 7 16 7 17 7 18 7 19 7 20 7 21 7 22 7 23 7 24 7 25 7 26 7 27 7 28 7 29 7 30 7 31 7 32 7 33 7 34 7 35 7 36 7 37 7 38 XX Expert Tool Top Level Menu 0 cece eee eee he 7 83 Info Capacity Screen ccc cece ee ence hh hh 7 86 Info 2 Inquiry Screen 0 cee ene ehh hh 7 86 I O 2 Read Disks Screen 0 2 ccc ee e 7 87 I O Read Disks Select Blocks Screen 0 ee een ees 7 88 VO 2 Read Disks Output Screen 00 eee eee eee 7 88 VO Read LUNS Ser en doom Re REPE Rex Rid 7 89 I O gt Read LUNs Select Blocks Screen 0 cece eee eens 7 90 V O gt Read LUNs Output Screen 0 0 eee eee eens 7 90 Test Unit Ready Screen 4e A a eee ER ga a 7 91
9. SRevision 66 1 User profile for Grid Manager Emulate vt100 terminal TERM vt100 export TERM Configuring MUX Ports on HP 9000 Systems Set up the shell environment set u trap echo logout 0 Set up the shell variables EDITOR vi export EDITOR Set up the Grid Manager Mux Port TTYDEV ttylu pport Add lu and port to access first the SP export TTYDEV Ensure there is only one invocation of Grid Manager if ps ef grep v grep grep v getty grep TTYDEV then echo Grid Manager is currently in use Try again later exit else exec cu l S TTYDEV where l is the letter el not the number one fi Copy profile to the second login directory cd users gridmgrl Go to directory for second login cp users gridmgr0 profile profile Copy profile to second login directory vi profile Edit the TTYDEV ttylu pport line in the profile to set up the second login to access the second SP Connecting the Disk Array to a D Series CPU The D Series CPU has only two RS 232 ports which are reserved for the console and the UPS The EISA MUX cards listed in the next two sections should be used with this CPU The cu 1 command where I is the letter el not the number one uses a different alphanumeric character a instead of p for the port number that is dev ttylu aport where lu represents the system dependent tty number and port is the adapter port
10. Connector Plug JA retaining Screw Figure 9 58 Removing the DC Power Distribution Cable Model 20 Rackmount Component Removal 9 99 jueuie2e deH pue jeAoway Fan Cable Figure 9 59 Note The fan cable can be removed without removing the midplane Remove the two screws at the top of the midplane and tilt the midplane back to gain access to the fan cable connector When reinstalling the fan cable make sure there is enough slack in the cable so the fan cable folds properly when the fan pack is closed Unplug the fan cable from the fan pack and remove the fan pack Figure 9 6 Release the fan cable from the cable retainer Remove the four retaining screws and remove the distribution panel Remove the fan cable from the two internal cable retainers Disconnect the fan cable from the midplane eU oe ee LO p Remove the fan cable Midplane pee Midplane connector Figure 9 59 Removing the Fan Cable 9 100 Model 20 Rackmount Component Removal Address and Internal SCSI 2 Cables Figure 9 60 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Remove four Phillips head screws and lift the distribution panel away from the array chassis Note The three cables connected to the distribution panel are Two SCSI 2 cables One address cable 3 Disconnect the address cable from the address PCA Note When connecti
11. eet 140 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 l ennt 160 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 ll ett 180 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 enhn LV AAAA AA AAAA AAA AA Figure 7 42 I O Read LUNs Output Screen Support Tools Manager STM 7 119 sonsouBeiq pue saninn ISOH Tests Test Unit Ready hpterm Gf x Expert Tool Test Unit Ready for device C2300 Array on path 874 9 W salt rose hp com 15 43 214 34 Expert Tool Test Unit Ready for device C2300 Array on path 8 4 3 Array Date 04 29 98 SN 9621954 System Date 04 29 98 Array Time 11 11 48 FW 9 25 00 8 02 00 System Time 11 20 55 Array SP A 8 0 MN 7624 SP System Namet salt Note Lun Test Unit Ready commands will fail for Binding luns sense key 05 25 00 Hot Spare luns sense key 05 25 01 Inaccessible luns sense key 05 04 00 or I O interface status 1 gt Lun Status SenseKey Sense Data v location of Sense Key ASC and ASCO NII Figure 7 43 Test Unit Ready Screen Tests Verify Disks The Verify Disks and Verify LUNs functions are set up the same as the Read functions but the information is displayed in the Expert Tools main window Figure 7 34 1 Select Tests Verify Disks from the Expert Tools main menu Select the type of verification from the dialog box that appears Figure 7 44 2 Select the desi
12. pod Figure 8 5 Front LED Locations Front View Disk Array LEDs 8 7 punoouse qnouj Rear View Disk Array LEDs Figure 8 6 shows the rear view LEDs Te EJ Service LED ey c cd d Storage sel d Serv processors 5 ON Ready LED A i g reen ue gm ETE fm fv 1 EN Power iP ii units 1 i UU Ready Service LED Sen green red Latch E Replace LED amber Figure 8 6 Rear LED Locations 8 8 Rear View Disk Array LEDs Table 8 2 Model 10 and 20 Disk Array LEDs LED Front Panel POWER ON green ON The power supplies are operational and are providing 24 VDC to the midplane 24 VDC bus OFF The array is powered off or if the array has been running the disk array has lost the 24 VDC bus Front Panel The array has detected a fault in one Normal operation FAULT amber of the FRUs Disk module The disk module is operational and Drive is busy BUSY LED lit or there READY green ready to receive send data is a fault FAULT LED lit Disk Module Disk module is receiving or sending Drive is ready READY LED lit or BUSY green data or is hung there is a fault
13. 1 29 Mirrored Modules in a RAID 1 0 Group in a Model 20 Disk Array 1 30 Distribution of User and Parity Data in a RAID 5 Group 1 31 Example of a Hot Spare in a Model 10 Disk Array 0000000 1 34 Example of a Hot Spare in a Model 20 Disk Array 0 1 35 Disk Space Use in the Physical Disk Configurations 1 37 Configuration Management Worksheet for HP UX 9 04 1 43 Configuration Management Worksheet for HP UX 10 x and Later 1 44 Example Configuration with Two Adapters with Two Arrays HP UX 9 04 1 46 xvii sainbi4 1 24 2 2 2 3 2 4 2 5 2 6 2 7 2 8 2 9 2 10 2 11 2 12 2 13 3 1 3 2 3 3 4 2 4 3 4 5 4 6 5 1 5 2 5 3 5 4 6 1 xviii Example Configuration with Two Arrays and Two Adapters in Each of Two Hosts HP UX 10 x and Later 0 2 2 eee eee ee eee 1 46 Engaging the Brake Deskside Disk Array 00 0 0 e eee eee eee 2 8 Removing or Installing the Front Panel 00 0 0 eee ee eee ee eee 2 13 Mounting the Rails 0 0 cee Ih 2 14 Mounting the Model 10 0 0 ee ee ce eee ene eee 2 17 Mounting the Model20 1 2 0 och ae ea ee be eae a a 2 18 SCSI 2 Bus Connector Locations Model 10 0 cc ee eee eee 2 23 SCSI 2 Bus Connector Locations Model 20 0 cee cece eee 2 24 Single Host Configurations 00 0 c
14. 2 0 cece eee eee enn iea 7 78 Expert Tool mec ARR MR dae ee ee eee 7 82 Selected Expert Tool Screens 0 0 eee eee eee eee ences 7 85 mstm Interface sienpre E ede nd Gah tis eines Wee a MATURE 7 99 xi U9 U09 Information T00l 24 zi desea ee ba ee eee RN ed ele ed wd kee A 7 99 Firmware Update Tool 0 cee eee ene eee 7 107 Expert Tool m a eee ee eee eed eee 7 111 Selected Expert Tool Screens llle 7 114 SAM System Administration Management Tool 0000 7 127 Starting SAM acr ea a coded rdiei ue E UN E ei S VE 7 127 Making a Menu Selection 00 cee eee ccc eens 7 127 Binding a Logical Disk Unit Using SAM 00 0 0 eee eee 7 127 Unbinding a Logical Disk Unit Using SAM 0 0 cee eee 7 128 LVM Logical Volume Manager essere 7 129 Creating Physical Volumes 0 c eect ee eee eee eens 7 129 Creating Needed Directories for Volume Groups 7 130 Creating Volume Groups 2 0 0 cece cece eee een nee 7 130 Creating Logical Volumes 0 0 cece eee eee e 7 131 Using Logical Volumes vk d REA dae ee Lae ees 7 131 EMS Hardware Event Monitoring 0 0 cee eee eee 7 132 Setting Up the Hardware Event Monitor 0 e cece eens 7 132 8 Troubleshooting About Field Replaceable Units FRUs 0 0 2 cece eee eee eee 8 2 Model
15. Note 9 Gbyte and 18 Gbyte disk modules are not supported on MPE ix 9 Gbyte LP and 18 Gbyte disk modules are not supported on the model 10 disk array RAID levels 5 1 and 1 0 data redundancy are supported on HP UX RAID level 3 is not supported on HP UX RAID level 5 is supported only as a data disk on MPE iX RAID level 1 is supported as a boot disk on MPE iX RAID levels 3 0 1 0 and independent mode are not supported on MPE iX RAID levels 5 1 0 1 0 and independent mode are supported on models A3231A A8232A A3388A A3389A A3539A and A3540A RAID levels 5 1 0 and 1 are supported on the models A3549A and A3550A RAID level 0 and independent mode are not supported on the models A3549A and A3550A Dual storage control processors SPs When two SPs are installed one SP can be hot swapped that is replaced with array power on Redundant power supply units VSCs and fan packs that can be hot swapped Global hot spare configurability Reference R 3 o2ue49JoH e Two internal synchronous single ended SE SCSI 2 buses in the model 10 e Five internal synchronous single ended SE SCSI 2 buses in the model 20 The disk array can continue to operate if one of the internal SCSI 2 buses fails e Auto trespass HP UX 10 x and later R 4 Reference Models Options and Add On Products Models Available e HP A3549A Model 20 High Availability SCSI Disk Array deskside e HP A3550A Model 20 High Avai
16. Example NIKEARRY gt UNBIND 1 Would You Like To Continue y n y Enter RESET to remap the system 7 62 NIKEARRY UNSOLLOG The UNSOLLOG command displays the Unsolicited Event Log of a given array to the console UNSOLLOG takes a LUN index as a parameter If no parameter is provided the value of the TESTUNIT environment variable is used as the LUN index number If TESTUNIT is not set the user is prompted for a LUN index Example NIKEARRY UNSOLLOG Enter test array cr 0 5 0 PLEASE WAIT WHILE THE UNSOLICITED SENSE DATA IS READ LOG LUN FRU CRU EVENT MESSAGE EXTENDED STATUS 2056 02 04 EO 0x906 Unit Shutdown 0x0000000000000000 2055 03 DO 0x906 Unit Shutdown 0x0000000000000000 2054 00 02 CO 0x906 Unit Shutdown 0x0000000000000000 2053 00 01 BO 0x906 Unit Shutdown 0x0000000000000000 2052 00 00 AO 0x906 Unit Shutdown 0x0000000000000000 2051 02 04 EO 0x602 Drive Enabled 0x0000000000000000 2050 03 D0 0x602 Drive Enabled 0x0000000000000000 2049 00 02 CO 0x602 Drive Enabled 0x0000000000000000 2048 00 01 BO 0x602 Drive Enabled 0x0000000000000000 2047 00 00 A0 0x602 Drive Enabled 0x0000000000000000 2046 02 04 EO 0x906 Unit Shutdown 0x0000000000000000 2045 03 DO 0x906 Unit Shutdown 0x0000000000000000 2044 02 04 EO 0x602 Drive Enabled 0x0000000000000000 2043 03 DO 0x602 Drive Enabled 0x0000000000000000 2042 02 04 E0 0x906 Unit Shutdown 0x0000000000000000 2041 03 DO 0x906 Unit Shutdown 0x0
17. Figure 9 48 Removing the Address and SCSI 2 Cable 9 83 Model 20 Deskside Component Removal Internal AC Harness Figure 9 49 and Figure 9 50 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Remove the outer covers Figure 9 40 3 Remove the following CRUs a Fan pack Figure 9 6 b Power supplies Figure 9 11 c SPs Figure 9 9 4 Remove the inner metal cover 11 screws Figure 9 41 5 Remove the distribution panel 6 Remove the internal AC harness bracket Remove four retaining screws and lift the internal AC bracket away from the chassis Note Record wire orientation before removing wires 7 Remove the green wire from the grounding lug behind the on off switch 8 Disconnect the blue and brown wires from the on off switch 9 Pull the cable out 10 Remove the three AC plugs from the internal AC harness bracket Note The plugs must be oriented as shown in Figure 9 50 when they are installed in the bracket If they are not oriented correctly the power supply units cannot be plugged in When reinserting the cable tape the two spade lug wires and the green neutral wire together This makes it easier to thread the cable 9 84 Model 20 Deskside Component Removal CAUTION Make sure not to pinch the SCSI 2 cables when installing the distribution panel
18. If the logs are turned off failing CRU data is not reported to the root user or the e mail distribution list Since most systems use MUX connections for array s RS 232 ports if an unplanned system outage occurs this feature cannot be disabled without doing a lot of non standard work Hard Media Error versus Hot Spare When a 0x920 Hard Media Error occurs the disk array faults the disk module that encountered the error If the disk module was configured in RAID 1 RAID 1 0 or RAID 5 and a hot spare is not configured in the disk array a rebuild does not begin until the faulted disk module is replaced If a global hot spare was configured a rebuild starts immediately on the hot spare Unsolicited Event Log 8 67 punoouse qnouj During the rebuild the SP microcode misinterprets the state of the disk module that faulted The microcode believes the disk module has just been inserted and tries to spin it up If the disk module does spin up it does not interrupt the rebuild If the 0x920 error is encountered again the disk to disk copy aborts and the disk module is faulted The disk module is held in the fault state by the SP microcode until it is replaced If the disk to disk copy completes it could mean the 0x920 error was recovered Even so the disk module should be replaced Another 0x920 error could occur on the disk module which would start another rebuild to the hot spare Bound but Unassigned Problems Problems will occur
19. Menu item Option Property Description File Select File filename Specifies the microcode or firmware file to be used Save As filename Saves a file using a user specified name Print N A Prints a file Exit N A Exits the Firmware Update Tool List Current N A Lists SP microcode and disk module firmware Revisions revisions currently in use Figure 7 33 File Selection N A Reports the file previously specified using the Select File option on the File menu Instructions N A Lists basic instructions for Firmware Update Tool use 7 108 Support Tools Manager STM Table 7 18 Firmware Update Tool Menu Items and Descriptions cont d Menu item Option Property Update Controller Download Firmware update PROM and restart controllers Description Downloads microcode to the SCSI disk array s database drives burns PROM code contained in the download file into the SP s PROMs and restarts the SP using the new microcode Download and restart controllers Downloads microcode to the SCSI disk array s database drives restarts the SP using the new microcode Download only Downloads microcode to the SCSI disk array s database drives but does not take effect until the SPs are restarted with a Restart Controllers only command or by power cycling the SCSI disk array Listing Microcode and Firmware Revisions Restart Restarts the SPs using microcode residing on the
20. Misc Status No detected fault in this subsystem Disable logging of cache statistic Initialize and enable the system cache SP Memory Size 2MB 0x2MB Usable Cache Size OMB 0x0MB Cache Page Size 2KB 0x2KB Cache State Cache Disabled Low Watermark 80 High Watermark 96 SP A owned Cache Pages Q 0x0 SP B owned Cache Pages QO 0x0 Unassigned Cache Pages 0 0x0 Current Date amp Time Thursday 5 12 1994 2 57 24 Optimized RAID 3 Memory O 0x0 PROM Revision 0 7 9 1 0 System Serial Number LUN Tables LUN State Bound but not assigned LUN Status OK LUN State Bound but not assigned LUN Status OK RESET The RESET command does a soft restart of NIKEARRY It resets all its variables and LUN lists to their original values rescans the buses for devices and displays them as with the DISPMAP command NIKEARRY 7 61 sonsouBeiq pue seninf ISOH UNBIND The UNBIND command allows the user to attempt to reconfigure the system by dissolving a LUN and removing it from the logical system Any FRUs which were bound to the specified LUN become unbound and inaccessible UNBIND takes a LUN index as a parameter If no parameter is provided the value of the TESTUNIT environment variable is used as the LUN index If TESTUNIT is not set the user is prompted for a LUN index CAUTION Unbinding a physical unit destroys all data on that physical unit LUN Unbind only when absolutely necessary
21. Multi bus Two hosts Two SCSI 2 adapters one per host Two SCSI 2 buses one per host Two SPs Disk Array Configurations Resembles two basic configurations side by side and provides modest high availability with the external SCSI 2 bus not shared between two hosts With RAID levels other than 0 or an individual disk applications can continue to run after a disk module failure HP UX 9 04 There is no automatic transfer of LUN ownership or path for this configuration type HP UX 10 x If one SP fails the System Administrator can access the LUNs owned by the failed SP by accessing the remaining operational SP with the other host If a host adapter fails the System Administrator can access the LUNs in the array with the other host adapter The System Administrator must change the path and the SP ownership of the LUNs for the failed SP or failed host adapter by using LVM and Grid Manager respectively 4 3 009 seues 0006 dH uoneunByuo25 Table 4 1 Array Configurations cont d Multi initiator Two hosts Two SCSI 2 adapters one per host One SCSI 2 bus Adds backup host to two SP configuration With RAID of any level other than 0 applications can continue after the failure of any disk module If one host one SCSI 2 adapter or one SP fails the other host can take over the failed host s disks if the array and hosts have been appropriately configured with LVM If both hosts intend to two per host
22. 0x3C Description Recovered error remap failed An attempt to do a remap after a Recovered error failed for some reason other than one of the following another media error a timeout cannot remap no LBA provided or bad defect list All of these errors are itemized separately Ox3D Media error cannot remap The device reported a Media error 03 xx but it did not report the bad LBA remap was not enabled Ox3E Media error remap failed An attempt to do a remap after a Media error failed for some reason other than one of the following another media error a timeout cannot remap no LBA provided or bad defect list All of these errors are itemized separately Ox3F Media error bad defect list Device returned a Media Error check condition which specifically indicates that the device s defect list is no good 03 19 or 03 1c A remap is not possible for these errors Ox7E Extended media error A remap of a bad block failed Usually seen with event codes 0x820 or 0x920 0x80 CRU Powered Down During a sniff the device was found to be powered down 0x81 CRU Powered Up Dead While powering up a CRU died 0x82 CRU Defined Powered Down Switched from defined state to powered down 0x83 CRU Defined Dead Drive found dead while in defined state 0x84 CRU Too Small Drive is too small to fit in the assigned LUN 0x85 CRU Unsupported Drive is not supported and has been shutdown 0x
23. 1 Make sure that the legs are securely fastened to the disk array 2 Make sure that the legs are fully extended Engage the Brake Engage the brake on the front bottom of the model 10 disk array to prevent it from rolling Figure 2 1 Be careful that the disk array does not tip over The model 20 has two brakes on its front bottom Engage both brakes on the model 20 Unpack the Disk Array 2 7 uonejeisuj Figure 2 1 Engaging the Brake Deskside Disk Array Check the Display Orientation Switch If installing a model 10 deskside disk array set the display orientation switch on the back of the disk array to the Tower position The switch is the topmost switch in the set of 10 SCSI ID switches located on the back of the disk array Refer to Figure 2 10 for the location of these switches The topmost switch should be moved to the right for deskside installation Unpack a Rackmount Disk Array 1 Move the disk array on its shipping pallet to the location where it is to be installed 2 Cut the plastic bands around the shipping carton and pallet Open the shipping carton by cutting the tape over the seams 2 8 Unpack the Disk Array WARNING The shipping carton is held onto the pallet with strapping bands Wear protective glasses while cutting these bands They are under tension When cut they spring back and can cause serious eye injury 3 Remove the cardboard box on the top of the disk array The b
24. Silece desired senian m Head ALL diske ALL blocks nn listed bn mer aulurtscm E55 C Mead SELECTED dika ALL blocks Blezka Blk 1a Blocks 2 4 Blocks Blecka Blocks Blocks Blecksl Blocks Bloka Blk Blezkri Blocks Bleka m ae aill F r dL EI upra xai LT Sap will Sipe uL El Viri tak EI br 2m3507 i 253037 Pa 7a 17253797 FREESAT 34878818 bs a be he be ba be Ten ymu arent tr ead the daka ng eelecteds opecifiied Lm Figure 7 13 I O Read Disks Screen Support Tools Manager STM 7 87 sonsoubeiq pue seni ISOH Poe Met AD on Array Serial 15 7003 p55 Fire Beck to reed ftis 17283507 m 1 Leet Block en reed MIS 1728235907 152 1 Mocka por mad cmd i 170 at 1 Macks b display D In Nese Last block bo read miari hn 1m iras black on reel b li raas hp com 15 43 27 H Esp ri Ten Head Dishe fee deside C98 Arrey on prih Be A Array Daves BE S598 SH 35 riy amp B5B farum Dara Boe Pa oe Array Tims L7 40 00 FH 3 53 81 1 73 00 Tejmrme Tirs 13 12 J7 rrai 5P A iram FEID jira M ialt D 5K BB data block JOIES 1 23 5 F B3 IBI Z3 45 BET BS BLZM5SETBS BLZ3455785 Done eno anon none ood ann ann te eap Fe BOB pO Bee Beda aos papi jiji Bes Bue Jiji agi Bone GUB UB BOBO BUD OE B ania i GUI ORIG DODO SUDO DOD anon Dono A anon Dons sog anon amn mga gig hlB
25. 2 Average Requests in Queue 0 Number of Remapped Sectors 0 Number of Blocks Read 120C6 Number of Blocks Written A3B6 Sum of Blocks Seeked 0 Upper 4 Bytes Hit CR to continue or E to Exit HARRAY Section 17 External Exerciser 7 29 sonsoubeiq pue seninf ISOH 7 30 FRU SLOT NUMBER AO cont Sum of Blocks Seeked 24D4A Lower 4 Bytes Sum of Queue Length on Arrival 58D Number of Arrivals with NonZero Queues 58C Number of 100 ms Ticks with this FRU idle Number of 100 ms Ticks with this FRU busy FRU Vendor Identification SEAGATE FRU Product Identification ST12550N FRU Product Revision Level HP03 FRU SLOT NUMBER A1 LUN REPORT UTILITY COMPLETED 4 09 PM HARRAY Section 17 External Exerciser MECH Non exclusive non destructive Returns information for disk mechanisms of the LUN under test Examples Several examples are shown Each one is for a different MECH option starting with option 0 and ending with option 5 Example 1 MECH command with option 0 HARRAY gt MECH The following commands are available at the mech level 0 Test Unit Ready TUR 1 Inquiry 2 Read Capacity 3 Read 4 Read Media Test RO MT 5 Verify via scsi commands 99 Exit Enter number 0 5 or 99 to Cancel 4 gt 0 The following mechs FRU s are available for test FRU NUMBER 0 FRU Slot Number A0 FRU NUMBER 6 FRU Slot Number A1 FRU NUMBER 12 FRU Slot Number A2 FRU N
26. 2 Install the MUX card in any appropriate slot on the Series 800 system 3 Connect the modem cable from the MUX card to the MDP panel 4 Power up and reboot the Series 800 system 2 36 Connect Asynchronous Console Cables 5 Verify that the MUX card is properly configured Configuration should have occurred automatically when the system was rebooted Check the following a The power LED on the MDP panel illuminates b The configuration is correct To verify the configuration issue an ioscan f command and find the lu number of the MUX card Make a note of this number Look for the special device files created for this MUX card by running the l dev ttylu command where is the MUX card s lu number The appropriate driver for the MUX card for HP UX 9 04 is mux2 major number 58 Issue a Isdev command to find the new major number for the mux2 driver Make sure this value is the same as the value that appeared in the major number field returned by the ll dev ttylu command Set Up the Disk Array on a Series 700 MUX Port To set up the disk array 1 Power down the Series 700 system 2 Connect the 24542H cable to one of the RS 232 ports on the Series 700 Core I O board The serial ports on the Series 700 are identified differently on each model 1 and 2 on the 725 100 and RS 232 A and RS 232 B on the 735 and 755 The ports on the host systems are 9 pin serial ports The ports on the disk array are 25 pin serial ports HP 245
27. 20 Firmware Pass Number 0 Number of Hard Errors 0 Maximum Outstanding Requests 1 Average Outstanding Requests 0 Busy Ticks F Idle Ticks 26FA83 Number of Reads 0 Number of Writes 0 CONTROLLER REPORT UTILITY COMPLETED 4 06 PM HARRAY Section 17 External Exerciser 7 17 sonsouDeig pue seni ISOH DESCRIBE Non exclusive non destructive The DESCRIBE command is used to view the vendor id product id and revision number of each of the disks configured in the LUN under test Example HARRAY gt DESCRIBE PDEV 24 14 0 is an C1300WDR5 DESCRIBE UTILITY 4 06 PM This LUN is bound as a Raid 5 Disk Array FRU SLOT Number AO FRU Vendor Id SEAGATE FRU Product Id ST12550N FRU Revision Number HP03 FRU SLOT Number Al FRU Vendor Id SEAGATE FRU Product Id ST32550N FRU Revision Number HP08 FRU SLOT Number A2 FRU Vendor Id SEAGATE FRU Product Id ST12550N FRU Revision Number HP03 FRU SLOT Number A3 FRU Vendor Id SEAGATE FRU Product Id ST12550N FRU Revision Number HP03 DESCRIBE UTILITY COMPLETED 4 06 PM 7 18 HARRAY Section 17 External Exerciser DIAG Exclusive non destructive The DIAG command is used to restart the controller firmware Example HARRAY gt DIAG PDEV 24 14 0 is an C1300WDR5 INITIATE DIAGNOSTIC UTILITY 5 50 PM WARNING This command will restart controller firmware Continue Y N N y Diagnostic Cmd successfu
28. An individual unit like a RAID 0 group provides no high reliability features As a result all its disk space is available for user data 1 36 Planning Disk Unit Configurations O user data L parity data 80 user data 20 parity data O Or Ls RAID 5 group 50 user data 50 redundant data Disk Mirror RAID 1 mirrored pair Individual disk unit No user data E 100 user data Hot Spare A redundant user data 100 user data O O s RAID O group non redundant array 50 user data O 3 50 redundant data Aa As Pog RAID 1 0 group Figure 1 20 Disk Space Use in the Physical Disk Configurations Planning Disk Unit Configurations 1 37 uonduioseg jonpold Guidelines for Logical Disk Unit Configurations After the array configuration is determined refer to chapters 3 4 or 5 the logical disk units must be configured Disk unit configuration must be based on the site s system environment and needs This section explains disk unit configuration choices and tradeoffs There are six disk unit configurations from which to choose e RAID 5 group individual access array e RAID 1 mirrored pair e RAID 0 group non redundant individual access array e RAID 1 0 group mirrored RAID 0 group e ndividual disk unit e Hot spare Note RAID 0 and individual disk units are not supported on models A3549 or A3550 because they do not provide data redundancy For
29. Packard maintains an active program of auditing production products to ensure these specifications remain true when products are again tested under the same conditions The limits of these specifications do not represent the optimum for long trouble free operation and are specifically not recommended for maximum customer satisfaction The recommended conditions are stated separately where appropriate Recommended operating range 20 C to 25 5 C 68 F to 78 F Operating range 10 C to 40 C 50 F to 104 F Non operating range 40 C to 70 C 40 F to 158 F Gradient maximum 20 C 36 F per hour Relative humidity 2096 to 8096 non condensing maximum wetbulb at 26 C Recommended humidity range 3096 to 5096 non condensing Altitude 3000 m 10000 ft Heat Output BTUs hr 1611 for model 10 2614 for model 20 Non Operating Environment Shipping and Storage Temperature range 40 C to 70 C 40 F to 158 F Gradient maximum 10 C hr 36 F hr Relative humidity 1596 to 9096 non condensing Altitude 4600 m 15000 ft Battery Disposal The BBU for both disk array models contains batteries that may be harmful to the environment if not disposed of properly Please follow the local state or country regulations regarding the proper disposal of these batteries Reference R 11 o2ue49JoH Acoustics Acoustic Levels Sound Pressure Model Sound Power Operator s Position Model 10 59 dBA 46 dBA
30. Remove the Array from the Rack Figure 9 29 CAUTION For these procedures the disk array must be powered off and unplugged Also it is recommended that ESD procedures be followed Use ESD kit P N 5182 4119 The chassis should be removed from the rack only when the internal parts of the array must be accessed 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 All of the CRUs must be removed because the disk array is too heavy to lift out of the rack if they are not removed Remove the following a b Fan pack Figure 9 6 IPCB Figure 9 21 SPs or fillers Figure 9 9 Power supplies or fillers Figure 9 11 Front panel Figure 9 18 BBU located in the lower front by pressing in on the two tabs at the sides of the unit Figure 9 13 All disk modules or fillers Figure 9 3 Be sure to label the disk modules as they are removed from the disk array 4 Remove the four retaining screws that hold the chassis to the rack Figure 9 29 5 Remove the empty chassis from the front of the rack Figure 9 29 9 50 Model 10 Rackmount Component Removal WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array Fan pack oooboooooo 000000000 ol o ol o o ol o o E e o o o o o o o p 9090
31. There is a serial port for each SP on the back of the disk array These are labeled CONSOLE SP A and CONSOLE SP B Use these serial links to configure the disk array in the following ways e Direct connect hard terminal configuration 2 33 Connect Asynchronous Console Cables uonejjeisuj e Terminal emulation through a MUX using Grid Manager discussed in chapter 6 Set up and physical attachment steps vary depending on how the disk array is configured discussed in Planning Disk Unit Configurations chapter 1 Direct Connect Terminal Configuration The simplest method of using the disk array s serial ports is to connect a terminal directly to each port For this configuration the following hardware is required e HP or EM100 capable Terminal Model 700 96 product number C1064W option ABA e HP RS 232 Cable part number 5061 5003 Two terminals and two RS 232 cables are required for a two controller disk array For a single controller disk array one terminal and one cable is required Cables are shipped with the disk array or with an SP upgrade To set up the hard terminal configuration 1 Using the RS 232 cable connect CONSOLE SP A on the back of the disk array to the serial port on the back of the HP terminal 2 34 Connect Asynchronous Console Cables 2 Power up the HP terminal Set the HP terminal to VT100 Emulation Mode or EM100 mode a Press the MENU key then the System key until the function
32. Two SCSI 2 buses Two SPs Two SPs access a particular disk simultaneously in this configuration each host should access the shared disk using the same primary controller to avoid continuous automatic controller ownership transfer which is not supported as a dynamic feature Multi bus Provides the highest availability and best array performance for dual multi initiator host configurations With RAID of any level other than 0 applications Two hosts can continue after the failure of any disk module If one host one Four SCSI 2 SCSI 2 adapter or one SP fails the other host can take over the adapters failed host s disks automatically when primary and alternate paths have been configured using LVM CAUTION The disk array is not supported as a boot device in either multi initiator or shared bus configurations on HP UX 10 x or later 4 4 Disk Array Configurations Basic Configuration for Model 10 and Model 20 Disk Arrays The basic configuration has one host with one SCSI 2 adapter connected by an external Fast Wide differential SCSI 2 bus to the SP in the array illustrated in Figure 4 1 The system can survive the failure of a disk module within a redundant RAID group or mirror but it cannot continue after failure of the SCSI 2 adapter or SCSI 2 cable Table 4 2 has more information Host computer Disk array SCSI 2 SCSI 2 bus SPA adapter RS 232 System console To MUX port or Lm
33. e The green POWER LED on the front of the disk array illuminates e The green READY LED on the disk drive modules illuminates e The fan pack starts operating The green READY LED on the power supply unit s illuminates e The green READY LED on the SP module s illuminates If any of these visual indicators are absent the amber FAULT LED on the front panel will illuminate One of the power up tests will display a message on the disk array console that a fault was detected The SERVICE or FAULT LED on the faulted CRU will be illuminated Power Up Sequence for the SP620 SCSI SP When the disk array is powered up each SP620 SCSI SP cycles through a series of self tests which test internal disk array components and functions As each test is performed one line of information about it including test ID description and status displays on the screen While a test is running its status is Running Its final status displays at the far right of the line either Passed if the test was successful or Failed if it was unsuccessful 8 12 Power Up Troubleshooting The information about each successive test overlays the preceding tests output line so that only the information about the current test is visible If a test fails its information remains displayed No additional tests are run and the disk array power up sequence halts The power up sequence also loads the SP microcode from the database disk modules onto the SPs The databas
34. lssf dev ttyl asio0 card instance 1 hardwired at address 2 0 4 dev ttylpO Type more etc inittab to identify the baud rate Example of tty information in the inittab file ttypl 2 0ff etc getty h ttyOpl 9600 ttyp2 2 0ff etc getty h tty0p2 9600 ttyp3 2 off etc getty h tty0p3 9600 Ports 1 through 4 are set at 9600 baud Set Up the Disk Array on an HP 3000 Series 900 DTC After connecting the disk array s SCSI 2 cables on HP 3000 Series 900 systems connect the disk array s RS 232 ports Plan this installation by considering the following 2 38 Connect Asynchronous Console Cables e Determine the total number of RS 232 I O ports and cables needed based on the number of disk arrays to be installed on the system One RS 232 port is required for each SP in each disk array e Ensure that enough 25 pin RS 232 ports are available on DTCs on the HP 3000 Series 900 system Add more DTCs and 25 pin RS 232 distribution panels J2085 option 101 or J2085 option 102 if needed e On the basis of the overall HP 3000 Series 900 system configuration choose a set of LDEV numbers to be assigned to the disk array s RS 232 SP connections Connecting RS 232 ports requires the following hardware e One or more DTCs DTC16iX J2062A DTC16MX J2063A DTC72MX J2070A with sufficient 25 pin RS 232 interface connectors supplied by J2085 option 101 8 port MDP and J2085 option 102 8 port DDP distribution panels so that each RS 232 port on eac
35. v more wv wv more wv ek Figure 7 40 I O Read LUNs Screen 7 118 Support Tools Manager STM amp hpterm P x Expert Tool Read Luns for device C2300 Array on path 874 9 For Lun on Array Serial 9621954 First block to read lt 0 689080313 J Last block to read 0 68908031 68908031 Blocks per read cmd 1 127 127 Blocks to display lt 0 200 10 uuu Note Last block to read must be gt first block to read Figure 7 41 I O Read LUNs Select Blocks Screen s hpterm P x Expert Tool Read Luns for device C2300 Array on path 8 4 9 IW salt rose hp com 15 43 214 34 Expert Tool Read Luns for device C2300 Array on path 874 9 Array Date 04 29 38 SN 9621954 System Date 04 29 98 Array Time 11 11 11 FH 9 25 00 8 02 00 Sustem Time 11 20 18 Array SP A 3 0 MN 7624 SP System Name salt LUN 0 data block 0 0123 45 67 89101 23 45 67 89 0123456789 0123456789 0 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 sesesese eet 20 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 le ett 40 0000 0000 0000 0000 0000 0000 0000 OOOO 0000 0000 lett 50 0000 0000 0000 0000 0000 0000 0000 OQQA 0000 0000 lett 80 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 leen 100 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 vent 120 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
36. 200 Verifying addresses 100 thru 1000 901 blk s 200 per Mech VERIFY complete 0 error s encountered HARRAY Section 17 External Exerciser 7 35 sonsoubeiq pue seninf ISOH READ Non exclusive non destructive Reads a block of data from the LUN Example HARRAY gt READ PDEV 24 14 0 is an C1300WDR5 READ UTILITY 4 45 PM Enter block address Block address Reading Block 31415 The Data Follows 0 1 2 3 O 00 00 00 00 10 00 00 00 00 20 00 00 00 00 510 00 00 End of data from block READ UTILITY COMPLETED lt cr gt to keep 0 12219263 current 0 31415 hex 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 31415 4 45 PM READ command with a block number as an option HARRAY gt READ 51413 PDEV 24 14 0 is an C1300WDR5 READ UTILITY 4 46 PM Reading Block 51413 The Data Follows 0 1 2 3 0 00 00 00 00 10 00 00 00 00 20 00 00 00 00 510 00 00 End of data from block READ UTILITY COMPLETED 7 36 hex 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 51413 4 46 PM HARRAY Section 17 External Exerciser REQUEST SENSE Non exclusive non destructive Returns sense information Example HARRAY gt REQUEST SENSE PDEV 24 14 0 is an C1300WDR5 REQUEST SENSE UTILITY 4 10 PM Sense bytes hex 00 70 00 00 00 00 00 00 OA 08 00 00 00 00 00 00 00 00 Me 00 00 00 00 0G 00 080 00 18 00 00 00 00 00 00 00
37. A flaw on the SEL line of the host SCSI bus would cause the NCR 720 SCSI controller on the SP to invalidate its onboard SSID register During the validation the microcode examined the register found a bit reset and panicked Fixed in microcode revision 8 20 0x00A0002F Bad information passed Occurred if an SP had just cleared its queue and was responding to a transfer request These codes are part of the SCSI bus synchronous and wide transfer negotiation handshake Fixed in microcode revision 8 20 Unsolicited Event Log 8 83 punoouse qnouj Table 8 17 Microcode Panic Codes Not a Complete List cont d Value 0x00A00043 Description l O SCSI command Occurred when executing an I O SCSI command from the host during a SCSI bus reset condition Fixed in microcode revisions 8 29 8 61 and 9 xx 0x00A00047 SCSI errors SCSI errors on the host SCSI bus might cause the microcode to panic with a 0x00A00016 or 0x00A00047 code Fixed in microcode revision 8 20 0x00A00081 Bad hardware Resets the backend SCSI bus indicating a possible hardware problem Check the SP SCSI I O adapter and all cabling connections 0x00A000B0 SCSI initiator problems Initiator was not responding Fixed in microcode revision 8 20 Panics 0x00A000BO0 0x00240010 and 0x00220050 occurred in SCSI environments that had severe SCSI protocol problems due to malfunctioning initiators or noisy conditions Fixed in microcode
38. B1 Free B3 Free B4 Free OK OK Help Figure 8 28 Bind a LUN Screen Troubleshooting with SAM 8 97 Bunoouse qnoJj Disk Devices Hardware Volume Total Path Use Group Mbytes Description 52 4 0 Unused 2048 Seagate ST12400 SCSI Disk Drive 52 5 0 LVM vg00 1292 HP C24 4 SCSI Disk Drive 52 6 0 Unused 645 HP C24 35 SCSI Disk Drive 8 5 a m ees HP C1300 Disk Array Controller 8 6 a HP C1300 Disk Array Controller 8 6 0 Unused aie 955 HP C1300 LUN RAID5 5 disks 8 6 1 Binding HP C1300 LUN RAID1 0 4 disks Figure 8 29 Binding Display Screen The upper left hand corner of the screen in Figure 8 28 shows the currently selected LUN number and the maximum rebuild time for that LUN Bind Type Accesses a pop up menu where the bind type is selected Sectors per Stripe Element Specifies how much disk space is consumed by each stripe element in striped RAID types If the bind type does not support disk striping this box is absent Enable Read Cache Enable or disable the read cache for the LUN Enable Write Cache Enable or disable the write cache for the LUN The remainder of the screen portrays the array and the available devices in the array Text at the top indicates how many disks can or must be selected depending on the bind type Only disks listed as Free can be bound as a LUN 8 98 Troubleshooting with SAM When satisfied with the configuration settings
39. Blocks per read cmd 1 125 125 Blocks to display lt 0 200 10 wou Note Last block to read must be gt first block to read Figure 7 38 I O Read Disks Select Blocks Screen s hpterm ioj x Expert Tool Read Disks for device C2300 Array on path 8 4 9 WW salt rose hp com 15 42 214 24 Expert Tool Read Disks for device C2300 Array on path 8 4 9 Array Date 04 29 38 SN 9621954 System Date 04 29 98 Array Timez 11 08 56 FH 9 25 00 8 02 00 System Time 11 18 03 Array SP A 3 0 MN 7624 SP System Name salt DISK AG data block 0 0123 45 67 89101 23 45 67 89 0123456789 0123456789 0 0010 1400 5765 6420 4EBF 7620 2035 2031 313A 3439 Wed No v 5 11 49 20 3A33 3920 3133 3937 OABD 6164 6520 6279 2073 6C65 739 1997 m ade by sle 40 7665 726F BEB5 0000 0009 1900 0002 0001 0004 0303 verone ssesereses 60 0000 0000 466C 6172 6520 S36F 6674 7761 7265 2053 Flare Software S 80 7973 7465 BDZE 0000 0000 0000 0000 0000 5361 756E ystem Saun 100 6120 636F BE74 726F BCBC 6572 2E20 2020 0000 0000 a controll er 120 0000 0000 0000 0000 0000 S86C2 0000 0000 0000 0000 lent 140 0000 0000 0029 4054 0082 0000 0003 179C 0000 0000 8T sssssesese 160 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 ertt 180 0000 0000 0000 0000 0000 0000 2240 2823 2920 4858 BOD Hi VM A AAAA AAAA AAAA AAAA AAAA AAA AAAA AAAA AAAA Figu
40. Change Unit Rebuild Time Displays dialogue for changing rebuild time 3 Change Unit Auto Assign Displays dialogue for changing auto assign 4 Change Unit Caching Parameters Displays dialogue for changing unit caching 2 Change Storage System Parameters 1 Change Storage System Caching Parameters Change caching parameters N Y Cache State Enable Disable Disable Cache Page Size in KB 2 4 8 16 64 Cache Size in MB 64 BBU Testing Enable Disable Enable BBU Test Day of the Week 1 7 1 BBU Test Hour 0 23 0 BBU Test Minutes 0 59 0 6 20 Grid Manager Menus FE FE 2 Change Main Menu Password Displays dialogue for changing password 3 Change System Date and Time Displays dialogue for changing date and time 4 Select Storage System Options Package Displays dialogue for changing system type 5 Change Storage System Fast RAID 3 Parameters Displays dialogue for changing FAST RAID 3 cache vault area 3 Change SP Parameters 1 Enable Disable Statistics Logging Displays dialogue for controlling statistics logging 2 Enable Disable Periodic Error Reporting Displays dialogue for controlling periodic error reporting 3 Change Console Parameters Displays several modifiable console parameters default is 9600 baud 8 bit no parity 4 Update SP PROMs and Reboot SP Displays dialogue for updating SP PROMs 5 Change Host Interface Options Addressing model LUN TARGET LUN Target init
41. Download and restart controllers Downloads microcode to the SCSI disk array s database drives and restarts the SP using the new microcode Download only Downloads microcode to the SCSI disk array s database drives but does not take effect until the SPs are restarted with a Restart Controllers only command or by power cycling the SCSI disk array Restart controllers only Restarts the SPs using microcode residing on the database drives Disk Firmware N A Supplies firmware to all compatible disk modules within the SCSI disk array that is those with the same product ID This applies even to disk modules owned by another SP The firmware then resides permanently within those disk modules until another disk module firmware update is performed Access the Current Revisions screen Figure 7 9 by selecting List Current Revisions from the Firmware Update Tool menu 7 80 Support Tools Manager STM e Firmware Update Tool for device type C2300 Array on path 874 8 D x File List Update Help CURRENT REVISIONS Mcode Ry Prom Rev This SP Peer SP Array Dt Array Tm Serial 3 55 81 1 73 08 A 9 82 25 38 17 19 39 95 7535 656 The array does NOT support commands to disks participating in active Hot Swap operations Such disks are excluded from the following list Product I DG6HSB3Y DG6HSB83Y DGHS83Y CLRR83 DGHS83Y CLRR83 DGHS83Y CLRR83 SEAGATE ST1515 N SEAGATE ST1515 N
42. Illustrated Parts Breakdown for the Model 10 Model 10 deskside front bezel Illustrated Parts Breakdown for the Model 20 Figure 9 2 Parts Breakdown for the Model 20 Array Illustrated Parts Breakdown for the Model 20 9 5 jueuie2e deH pue jeAoway Table 9 2 Model 20 Disk Array CE Replaceable Parts see Figure 9 2 Figure New HP Exchange index part number part number Description 1 005 42853 A3232 69107 Model 20 midplane 2 005 039649 N A LED cable 3 N A N A Chassis 4 005 043512 N A Internal synchronous SE SCSI 2 cables 5 N A N A Internal AC harness bracket 6 113 000634 N A On off switch 7 005 039646 N A Power supply AC cable 8 005 039648 N A Fan cable 9 005 039645 N A SP address cable 10 005 041325 N A RIP PCA N A 5063 5396 A3388 69021 4 Gbyte disk module N A 5063 5398 A3232 69121 2 Gbyte disk module N A 5063 5393 A3666 69021 4 Gbyte LP disk module N A 5063 5394 A3541 69021 9 Gbyte HH disk module N A 5063 5333 A5132 69021 9 Gbyte LP disk module N A 5063 5334 A5133 69021 18 Gbyte LP disk module N A A3232 60005 A3232 69005 Battery backup unit BBU N A A3232 60003 A3232 69103 Storage processor SP N A A3549 69102 Storage processor SP620 SCSI SP PCA N A A3232 60004 A3232 69104 Power supply unit VSC N A A3232 60006 A3232 69006 Fan pack 9 6 Illustrated Parts Breakdown for the Model 20 Tab
43. RUL Allows the user to view Unsolicited Sense Logs SHOW MECHS Displays configuration information for each possible FRU in the system SUSPEND Suspends HARRAY and returns control to the DUI AUTO TRESPASS Allows the user to enable disable AutoTresPass WTR MT Performs a write then read media test on the disk The user is given the option of testing the complete data path or disk media only COMMAND SYNTAX HELP CMD ALL The following screen shows help with a command as an option HARRAY help ro Do you want a description or just syntax D S D d COMMAND DESCRIPTION NON EXCLUSIVE NON DESTRUCTIVE COMMAND The RO MT command enables the user to initiate a Random Read Only Media Test on the LUN COMMAND SYNTAX RO MT HARRAY Section 17 External Exerciser 7 27 sonsoubeiq pue seninf ISOH INQUIRY Non exclusive non destructive The INQUIRY command returns information about the disk array Example HARRAY gt INQUIRY PDEV 24 14 0 is an C1300WDR5 INQUIRY UTILITY 4 08 PM Peripheral Device Type 0 Direct Access Device Device Type Modifier 101 Raid 5 ANSI approved version 2 AENC 0 7 28 Controller but not as TrmIO 0 Controller RelAd 0 Controller WB32 0 Controller WB16 1 Controller Sync 1 Controller Linked 0 Controller CmdQu 1 Controller SftRe 0 Controller Reset Controller Vendor Identification Controller Product Identification Control
44. Remove a SIMM Figure 9 16 4 a Pull out on the clips b Lift the SIMM out of the socket Install a SIMM Figure 9 17 a Locate the socket where the SIMM is to be installed b If only two SIMMs are being installed they must be installed in the following sockets SP model SIMM sockets As shown in SP SIMM 1 and SIMM 3 Figure 9 14 SP620 SCSISP SIMM 1 and SIMM 2 Figure 9 15 This information is also silkscreened on the SP board Place the SIMM into the socket at an angle as shown in Figure 9 17 Ifinstalling the SIMM on an SP controller press the SIMM into place upright until it locks into the retaining clips as shown in Figure 9 16 The SIMMs on the SP controller are inserted at about a 60 degree angle and then straightened to an upright position to lock them into place If installing the SIMM on an SP620 SCSI controller press DOWN on the SIMM in the opposite direction of the arrow in Figure 9 16 until it locks into the retaining clips The SIMMs on the SP620 SCSI controller are inserted at about a 75 degree angle and then pushed down to about a 30 degree angle to lock them into position Also the key notches are on opposite sides on the two controllers so the SIMMs have to be rotated when transferring them from one type of controller to the other SP to SP620 SCSI or vice versa 5 Re install the SP Figure 9 10 6 Enable disk cache To have caching an array must have two SPs The cache on each S
45. SEAGATE ST1515 N IBM DCHS 4F CLARG4 IBM OCHS 4F CLRR84 IBM DCHS SF CLAR S IBM DCHSB8S3F CLAR S IBM DCHSB83F CLRR83 CO CO CO amp A J A CD CO CO Co CO CD CO CO DJ n P9 CO n CO CO CO CO CO Interrupt Abort Figure 7 9 Current Revisions Display Updating Microcode and Firmware To use the Firmware Update Tool the microcode or firmware file to be downloaded to the disk array must be available on the host system The procedure to update the microcode or firmware with the Firmware Update Tool starts with the following two steps 1 Select the microcode or firmware file to be used for the update with the Select File option on the File menu 2 Identify the type of file selected either SP microcode or disk module firmware with the Update menu Support Tools Manager STM 7 81 sonsouBeiq pue seninf ISOH SP Microcode Updates If the SCSI disk array uses two SPs both should be enabled before firmware is updated so that they are updated simultaneously The download update PROM and restart controllers property from the Controller Firmware option on the Update menu is recommended for general use The other options provide a greater degree of control needed only in special circumstances Disk Module Firmware Updates The peer SP must be disabled when downloading disk module firmware in a dual SP SCSI disk array configuration If a peer SP is enabled the Firmware Update Tool requests that it be disabled an
46. and loop ID of peer SP Serial Serial number of the SCSI disk array FAULT DETECTED Displayed only if the SP reports a SCSI disk array fault the Information Tool completion status will be set to Information Warning Array Date SCSI disk array s date in mm dd yy format Array Time SCSI disk array s time in hh mm ss 24 hour format Cabnt Type Cabinet type for example deskside or rackmount Fan A B Fan status for example OK or NR not ready VSC A B C Power supply units status 7 102 Support Tools Manager STM Table 7 14 xstm Status Fields for the Information Tool Log cont d Status field values Description BBU Battery Backup Unit status Cache Amount of physical memory on selected SP Systm Physical memory reserved for SP s use Read Physical memory reserved for reads and status Write Physical memory reserved for writes and status Firmware Rev SP firmware revision Microcode Rev SP microcode revision Prom Rev Levl PROM revision level Auto Trespass Auto trespass setting Stat Logging Statistics logging setting The physical layout contains information about the status of the disks and LUNs Table 7 15 Disk State for the Map in the Information Tool Log Status field values Description blank removed Off off Pow powering up Unb unbound Rdy ready Reb rebuilding Ena enabled Bin binding Fmt formatting Eq
47. contained To provide this modularity some information is repeated in each section Table 4 1 Array Configurations Configuration type Basic One host One SCSI 2 adapter One SCSI 2 bus One SP Features With RAID of any level other than 0 or an individual disk applications can continue to run after the failure of a single disk module within a physical disk unit LUN Applications cannot continue to run if an SP SCSI 2 adapter or external SCSI 2 bus fails Two SP One host One SCSI 2 adapter One SCSI 2 bus Two SPs Provides the same features as the basic configuration but with better performance and higher availability With RAID of any level other than 0 or an individual disk applications continue to run after the failure of a single disk module HP UX 9 04 If one SP fails the System Administrator must reassign ownership of the failed SP s LUNs to the operational SP by using Grid Manager Also the path to the operational SP must be used and can be changed by way of LVM Once these changes have been made access to all LUNs on the array is possible If the SCSI 2 adapter or external SCSI 2 bus fails no I O operations are possible between the host and the array HP UX 10 x If one SP fails ownership of the failed SP s LUNs can transfer automatically to the remaining operational SP If a primary and an alternate path have been configured in LVM LVM can switch automatically to the alternate path These op
48. e The information stored can be easily restored from backup Individual disk units are flexible The performance of an individual disk unit compares with a standard SCSI disk not in a disk array The slight performance degradation results from SP overhead Note RAID 0 and individual disk units are not supported on models A3549 or A3550 because they do not provide data redundancy For microcode revision 8 56 and later these RAID levels can only be configured in FE mode with the Grid Manager utility See chapter 6 Grid Manager for more information about FE mode 1 40 Planning Disk Unit Configurations Hot Spare A hot spare is recommended if in RAID 5 RAID 1 0 or RAID 1 groups high availability is so important that data redundancy needs to be regained as soon as possible if a disk module fails A hot spare provides no data storage but enhances the availability of each RAID 5 RAID 1 and RAID 1 0 group in a disk array Performance The RAID configurations differ in their performance data availability and disk space features Table 1 1 compares the read and write performance of array configurations and individual disk units with the performance of a single spindle disk Read and write performance given is the theoretical maximum performance relative to the performance of a single disk Table 1 1 Relative Performance Compared to a Single Disk Physical disk Tolerance for single configuration Read performance
49. right parenthesis following the menu number indicates a menu one level down from the Main Menu a hyphen indicates a menu two levels down from the Main Menu a period indicates a menu three levels down from the Main Menu Grid Manager Menus for Microcode Revision 7 23 Menu number from the Grid Manager Main Menu 1 Presentation Utility Displays the Presentation screen 2 Bind Physical Units 1 RAID 5 Group Individual Access Array Displays dialogue for making a RAID 5 LUN 2 RAID 3 Group Parallel Access Array Displays dialogue for making a RAID 3 LUN 3 RAID 1 Mirrored Pair Displays dialogue for making a RAID 1 LUN 4 RAID 0 Group Non redundant Individual Access Array Displays dialogue for making a RAID 0 LUN 5 RAID 1 0 Group Mirrored RAID 0 Group Displays dialogue for making a RAID 1 0 LUN 6 Individual Disk Unit Displays dialogue for making an independent mode disk Grid Manager Menus 6 19 Jobeue N pus 7 Hot Spare Displays dialogue for making a hot spare 3 Unbind Physical Units Displays dialogue for unbinding a LUN 4 View Unsolicited Event Log Displays the Unsolicited Event Logs starting at entry number 1 Must use FE mode to see all entries USER mode only shows 0x9xx and OxAxx errors Allows selection of a filter for either error code or FRU 5 Change Parameters 1 Change Physical Unit Parameters 1 Change Default SP Ownership of Unit Displays dialogue for changing SP ownership 2
50. 0 0 4 1 0 0 4 2 0 and 0 4 3 0 rather than the expected 4 13 0 etc In addition to the host system seeing incorrect SCSI addresses the Grid Manager screen shows no SCSI address associated with the SP selected via Grid Manager When in this situation enter FE mode by typing CONTROL P followed by in upper case FE and follow the menu path 5 Change Parameters gt 2 Change Storage System Parameters 5 Select Storage System Option Package Select either A or 2 whichever is being used After the correct option has been selected the array reboots automatically after the correct option has been entered the array reboots Cache is enabled automatically and FE mode is turned off after the reboot completes FE mode is turned off after the reboot completes The storage system parameters can be modified only by an HP service representative and should be set correctly at the time the disk array is installed Changing the Storage Systems Option Package Two option configurations are available on the disk array Option A is used for HP UX 9 04 HP UX 9 05 9 07 and MPEAX 5 5 Option 2 is used for HP UX 10 x Changing Storage System Parameters 6 65 Jobeue N pus Note The file system on the disk array must be unmounted before a new options package is selected The array reboots automatically when the new options package is selected The HP UX operating system also must be rebooted after the new options package has been selected on the
51. 0 100 Sniff Period To set the sniff period from the Grid Manager main menu follow the menu path 5 Change Parameters 1 Change Logical Unit Parameters 6 Change Unit Verify Configuration Passes Indicates the total number of verify operations that have been performed Applies only to the first and fourth page reports Historical Totals Correctable and Uncorrectable Totals Lists the number of errors which were detected during verification A correctable error indicates that the SP was able to fix the detected problem while an uncorrectable error indicates it was unable to do so Errors are normally only correctable on redundant unit types Contact your HP service representative for further information about the following error classifications Checksum Write Stamp Time Stamp Shed Stamp Coherency LUN Configuration Summary To view information about each LUN owned by the active SP from the Grid Manager Main menu follow the menu path 4 View Maintenance Information 4 Display LUN Configuration Summary This screen indicates the current state of each LUN the type of LUN RAID level the Read Write cache status for each LUN and the number of disk modules bound into the LUN and the disk array cache states Viewing Maintenance Information 8 49 punoouse qnouj m xterm Miei X SP A SP SCSI ID 15 LUN Configuration Summary S N 9621954 Unit Cache LUN State Tupe Read Write Disks 0 ENA RAID
52. 00 20 00 00 00 00 Sense Key Code 0 00 No Sense Key Information No additional sense information REQUEST SENSE UTILITY COMPLETED 4 10 PM HARRAY Section 17 External Exerciser 7 37 sonsoubeiq pue seninf ISOH RESET Exclusive non destructive Issues a bus reset to the device Example HARRAY gt reset PDEV 24 14 0 is an C1300WDR5 RESET UTILITY 4 46 PM Issuing Bus Reset to device RESET UTILITY COMPLETED 4 46 PM 7 38 HARRAY Section 17 External Exerciser REVISION Non exclusive non destructive Returns microcode revision number Example HARRAY gt REVISION PDEV 24 14 0 is an C1300WDR5 REVISION UTILITY 4 10 PM Firmware Major Number 8 Firmware Minor Number 20 REVISION UTILITY COMPLETED 4 10 PM HARRAY Section 17 External Exerciser 7 39 sonsoubeiq pue seninf ISOH RO MT RANDOM MEDIA TEST Non exclusive non destructive Example HARRAY gt RO MT RANDOM MEDIA TEST UTILITY Enter t 4 11 PM test is performed by executing a sequence of reads or writes length of each transaction can be User specified length Randomly selected lengths he transaction length type Do you want to continue on errors Enter t he loop count 1 2295 E1 R U R r Y N N y TE RANDOM MEDIA TEST beginning MON APR 29 1996 4 12 PM RANDOM Test Pass Count 1 4 12 PM The Random Media Test is 25 complete 4
53. 1 From the Grid Manager Main menu follow the menu path 5 Change Parameters 2 Change Storage System Parameters 1 Change Storage System Cache Parameters Change Storage System Cache Parameters At any prompt enter to GO BACK TO PREVIOUS MENU or for help Current storage system cache parameters Cache State DISABLED Cache Page Size 2 KB SP Memory 16 MB Cache size 16 MB Change caching parameters N Figure 6 14 Change Storage System Cache Parameters Screen Disable 2 To change the cache page size or cache size values type Y for yes to the Change caching parameters prompt The following prompt will appear Cache State Enable Disable Disable 3 Press RETURN to accept the default of Disable Ifthe default is not Disable go to step 3 in the previous section Disabling and Enabling Disk Array Caching The following prompt will appear Cache page size in KB 2 4 8 16 2 4 Enteranew cache page size or accept the default size Setting up Disk Array Caching for Microcode Revision 7 23 6 57 Jobeue N pus Generally the cache page size should be 2 or 4 Kbytes for database applications and 8 Kbytes for general HP UX file server applications The following prompt will appear Cache size in Mbyte 8 5 Enter the size of the cache desired or accept the default size then enable cache Unless a special performance environment is being set up a cache size equal to the SP memory should be
54. 104 Model 20 Rackmount Component Removal Figure 9 62 Removing the AC Harness Plugs Model 20 Rackmount Component Removal 9 105 jueuie2e deH pue jeAowaYy On Off Switch Figure 9 63 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds Remove the DC power distribution panel Figure 9 58 Disconnect the four spade lug wires from the switch Remove the on off switch from the distribution panel De a GP Depress the retainer clips and push the switch out Note To install the on off switch press it into the distribution panel Use the flat blade of a screwdriver to depress the retainer clips Four spade lug wires Nx To wiring harness T Retaining clips 2 on each side of switch Figure 9 63 On Off Switch and Power Plug Assembly 9 106 Model 20 Rackmount Component Removal GLOSSARY battery backup unit BBU A unit that provides limited backup power in case of a power outage A BBU is required to enable disk array cache If power fails the BBU allows the SP to write the cache image to the cache vault area on disk A BBU can be replaced under power without interrupting applications bind In the context of disk arrays the procedure by which one or more disk modules are formatted into one physical disk unit LUN Disk modules can be bound as one of the following types of physical disk units RAID 5 group independent access array RAID 1 mi
55. 21 7 1 7 2 7 4 7 5 7 6 7 7 7 8 7 9 NMMGR 3000 Terminal Profile for Disk Array Grid Manager RS 232 Ports 6 8 Running GridMgFAX i iud lll ex RE Eee ERUCLE RIDERE ELE 6 12 Notification of Insufficient User Capabilities 0 0000 6 13 Grid Manager Main Menu sslseeeeeeeeeeee ee 6 14 Access Denied Ldev Already in Use by Another Session 6 15 Grid Manager Main Menu 0 cece eee eee 6 17 Typical Presentation Utility Screen 0 0 cee ee eee 6 33 Soft Disable Enable SP sss 6 42 Bind Logical Units Menu Microcode revision 9 Xx lssesees 6 46 Grid for Binding Modules for the Model 20 0 cece eee eee eee 6 47 Unbinding a Logical Disk Unit 0 0 00 0 6 52 Change Storage System Cache Parameters Screen Enable 6 55 Change Storage System Cache Parameters Screen Disable 6 57 Storage System Memory Parameters 0 0 0 cece eee eee eee eee 6 61 Change Storage System Cache Parameters Screen Disable 6 63 Host Interface Options 0 s cece e 6 69 Main Menu Password Utility 0 0 eee ee eee eee eee 6 71 Change Main Menu Password Utility llle 6 72 System Inventory Information 0 0 cece cece eee eee ee ene 6 73 Physical Disk Inventory Information 00 0 ce eee eee eee eens 6 74 Disk FRU Map Model 20 Deskside
56. 6 39 weight R 9 power troubleshooting 8 41 power up the disk array 2 43 power up 8 15 failure 8 15 sequence for the SP 8 10 sequence for the SP620 SCSI SP 8 12 tests for SP 8 10 tests for SP620 SCSI SP 8 13 power up troubleshooting 8 10 Presentation Utility checking status 8 86 screen 6 33 PROM code R 2 6 34 6 37 7 21 8 19 8 27 PROM revision on Presentation screen 6 37 R rackmount disk array Index 1 5 xepu unpacking 2 8 RAID concepts 1 23 levels supported R 3 1 18 performance 1 41 technology 1 18 1 23 1 38 RAID 0 1 25 1 36 1 38 1 40 RAID 1 1 18 1 26 1 36 1 38 1 39 hardware mirroring 1 26 RAID 1 0 1 18 1 27 1 36 1 38 1 39 for model 10 1 29 for model 20 1 30 RAID 5 1 18 1 30 1 35 1 38 1 39 6 44 read cache 8 95 READ HARRAY command 7 36 reading disks modules 7 84 7 87 7 118 7 116 LUNs 7 84 7 88 7 89 7 90 7 113 7 117 7 119 rebuild 8 24 time in STM 7 85 7 114 rebuild abort events and messages 8 65 recovered error log reporting 8 66 redundant power supply units R 3 removal AC line filter model 10 deskside 9 48 AC line filter model 10 rackmount 9 68 address and internal SCSI cables model 20 rackmount 9 101 address and internal SCSI cables modle 20 deskside 9 82 BBU 9 21 components model 10 deskside 9 29 components model 10 rackmount 9 50 components model 20 deskside 9 70 components model 20 rackmount 9 88 DC power distribution cable model 20 deskside 9 78 DC power dist
57. 7 131 UTILINFO Nikearry command 7 67 V verify disks 7 84 7 92 1 113 7 120 LUNs 7 84 7 94 7 95 7 113 7 121 7 123 verify results display of 8 47 VSC 1 14 W weight R 8 BBU R 9 disk module R 9 fan pack R 9 power supply unit R 9 SP R 9 WRT MT HARRAY command 7 44 X xstm 7 68 xstm interface of STM 7 69 7 98 Index 1 9 xepu 1 10 Index Reader Comment Sheet Hewlett Packard High Availability SCSI Disk Array Service Manual We welcome your evaluation of this manual Your comments and suggestions will help us improve our publications Remove this page and mail or fax it to 916 785 2299 Agree Disagree N A The manual is well organized O O The information is technically accurate Information is easy to find Step by step procedures are easy to perform There are enough examples and pictures OO 0 0 O0000 0 QOO 0 9000 O0000 0 O O0 0 0 O00 OQ OQ The examples and pictures are useful Comments Name Phone Title fax Company E mail Address ZIP City amp State Country Check here if you would like a reply Hewlett Packard has the right to use submitted suggestions without obligation with all such ideas becoming the property of Hewlett Packard ESBU Roseville Information Engineering A3232 90010 E1299 BUSINESS REPLY MAIL FIRST CLASS MAIL PERMIT NO 256 ROSEVILLE CA POSTAGE WILL BE PAID BY ADDRESSEE Attention I
58. 7 23 8 18 disk module about 1 9 capacity 7 84 7 86 7 113 7 115 firmware updates 7 82 7 111 inquiry 7 84 7 86 7 113 7 115 reading 7 84 7 87 7 113 7 116 removal replacement 9 8 status codes 6 36 status on Presentation screen 6 36 verify 7 84 7 92 7 113 7 120 weight R 9 disk striping 1 24 display orientation jumper 9 73 9 92 switch 2 8 2 16 DISPMAP Nikearry command 7 52 DOWNLOAD HARRAY command 7 20 downloading disk module firmware 7 82 7 111 SP microcode 7 82 7 110 dual SP 1 20 E EISA 64 port MUX card 6 6 8 port MUX card 6 5 EMC compliance R 13 enable disable switch 6 42 9 15 enabling disabling caching 6 62 environmental specifications R 10 event notification 7 132 EXIT HARRAY command 7 25 Expert Tool of mstm 7 111 7 126 Expert Tool of xstm 7 82 7 98 extended status decoding 8 70 F fan cable removal 9 39 9 59 9 80 9 100 fan pack aboutthe 1 13 removal replacement 9 11 weight R 9 fan status on Presentation screen 6 39 fault field on Presentation Screen 6 34 features R 2 file system create on a Series 700 8 101 Firmware Update Tool of mstm 7 107 7 111 Firmware Update Tool of xstm 7 78 7 82 firmware disk 7 23 8 18 FRUINFO Nikearry command 7 56 FRUs 8 2 G grid cursor moving on Presentation screen 6 41 Grid Manager about 6 2 FE mode 6 17 initiating on HP UX 6 10 initiating on MPE iX 6 11 inventory screens 6 73 main menu 6 17 menus for microcode rev 7 23 6 19 menus for mi
59. 9 30 7 Remove the inner chassis Figure 9 31 9 66 Model 10 Rackmount Component Removal Note The on off switch is attached to the inner chassis 8 Remove the four spade lug wires from the back of the switch 9 Use a flat blade screwdriver to press down on the retaining clips 10 Work the switch back and forth to loosen Retaining clips 2 on top 2 on bottom 4 4 Four P spade lug lt 77 To wiring Figure 9 38 Removing the On Off Switch Model 10 Rackmount Component Removal 9 67 jueuie2e deH pue jeAoway AC Line Filter Figure 9 39 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 Remove the following a b C g Fan pack Figure 9 6 SPs or fillers Figure 9 9 Power supplies or fillers Figure 9 11 Front panel Figure 9 18 BBU located in the lower front by pressing in on the two tabs at the sides of the unit Figure 9 13 All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array IPCB Figure 9 32 4 Remove the four retaining screws Figure 9 29 WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array 5 Remove the empty chassis from the rack Figure 9 29 6 Remove 11 screws and remove the t
60. 9 30 7 Remove the inner chassis Figure 9 31 Model 10 Rackmount Component Removal 9 63 jueuie2e daH pue jeAoway 8 Cut the cable tie securing the AC cable to the chassis 9 Disconnect the associated spade lug wires one green neutral wire and remove the AC cable Bracket retaining screw One at each end AC harness Figure 9 36 Removing the AC Harness Bracket 9 64 Model 10 Rackmount Component Removal Figure 9 37 Removing the AC Harness Plugs Model 10 Rackmount Component Removal 9 65 jueuie2e deH pue jeAoway On Off Switch Figure 9 38 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 Remove the following a b C g Fan pack Figure 9 6 SPs or fillers Figure 9 9 Power supplies or fillers Figure 9 11 Front panel Figure 9 18 BBU located in the lower front by pressing in on the two tabs at the sides of the unit Figure 9 13 All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array IPCB Figure 9 32 4 Remove the four retaining screws Figure 9 29 WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array 5 Remove the empty chassis from the rack Figure 9 29 6 Remove 11 screws and remove the top panel of the chassis Figure
61. A only Federal Communications Commission Radio Frequency Interference Statement Warning This equipment generates uses and can radiate radio frequency energy If it is not installed and used in accordance with the instruction manual it may cause interference to radio communications It has been tested and found to comply with the limits for a Class A computing device pursuant to part 15 of FCC Rules which are designed to provide reasonable protection against interference when operated in a commercial environment Operation of this equipment in a residential area is likely to cause interference in which case the user at his own expense will be required to take whatever measures may be required to correct the interference If this equipment causes interference to radio reception which can be determined by turning the equipment off and on try these measures Re orient the receiving antenna Relocate the computer with respect to the receiver Move the computer equipment away from the receiver Plug the computer and the receiver into different branch circuits Consult your dealer or an experienced technician for additional suggestions VCCI Class 1 For Japan Only p enh CORE PRUBRESERESATMAMBS VCCI OBE LES S BARAMRE CT COREE CERT 6c Sk Eee CFCEMHVET COMBIMICHAAM MIDGET O KIBRENSCEMBVYET R 24 Reference Declaration of Conformity according to ISO IEC Guide 22 and EN 45014 Manufacturer Hewlett Packard Company Hewlett
62. A Binds selected disk modules into a LUN with a specified RAID level Unbind LUNs N A Unbinds a LUN LUN Configuration Auto assign Enables or disables the auto assign state of a LUN Cache Parms Enables or disables write and read caching for a LUN Default SP Sets the default SP for a specified LUN Rebuild Time Sets the maximum amount of time the SCSI disk array should take to rebuild a failed disk module on a LUN System Configuration Auto Trespass Sets the auto trespass state on or off BBU Test Time Specifies the day and time the BBU self test is to run Date Time Sets SCSI disk array date and time to match the host system Host I F Options Sets or changes the Host interface settings for the selected SCSI disk array See Figure 7 26 Serial Number Sets or changes the disk array s serial number System Cache Specifies system cache states See Figure 7 24 System Memory Specifies system memory configuration values See Figure 7 25 System Type Selected Expert Tool Screens Controls certain operating characteristics of the array HP supported types are 2 and A The following screen captures Figure 7 11 through Figure 7 26 are examples of selected Expert Tool screens Support Tools Manager STM 7 85 sonsouDeig pue seni ISOH Info Capacity lt ranm hp om i Ih 433 7158 38 Fuperr Teed Dapecing fer deire
63. A0 BO CO DO and E0 are a RAID 5 group Those in slots Al and B1 are a RAID 1 mirrored pair and the module in A2 is a hot spare see Figure 1 19 Eventually module D0 fails Immediately the SP begins rebuilding the RAID 5 group using the hot spare When it finishes the RAID 5 group consists of disk modules AO BO CO A2 and EO No hot spare is now available When the failed module in DO is replaced the SP starts copying the structure on A2 to DO When it finishes the RAID 5 group once again consists of modules A0 BO C0 DO and EO and the hot spare in A2 becomes available for 1 34 Planning Disk Unit Configurations use if any other module fails A similar sequence would occur if module A1 in the mirrored pair failed In summary e The RAID 5 group is AO BO CO DO and E0 the RAID 1 mirrored pair is Al and B1 the hot spare is A2 e Disk module D0 fails e The RAID 5 group is AO BO C0 A2 and E0 no hot spare is available e Failing disk module DO is replaced with working module e The RAID 5 group is AO BO CO DO and EO the hot spare is A2 Hot Spare mLOM A0 m Bt AO BO CO DO EOJA2 A1 B1 o o o o o j Deskside Model I Rack
64. Check SP connector for gouging Pin on midplane connector or host adapter may be damaged or bent Also check external SCSI 2 cable for damage SP READY LED on with disk array failure Power up tests fail at letter W Check that database disks are installed in the proper slots No SP microcode being loaded from disk Trying to unbind disks and get dual board unbind error Trying to unbind disks that this SP does not own The other SP owns the disks Access the other SP to unbind the disks ISL gt IOMAP SYSMAP gt IOMAP doesn t recognize the disk array or second SP 8 30 Check term power light on host adapter Check SCSI address conflict on bus Check SCSI 2 cable length Check SCSI piggyback SP to SP cable Check terminator or termination on bus Troubleshooting the Disk Array SIMM Troubleshooting for the SP Only This information does not apply to the SP620 SCSI SP but to the older model SP The SP has four SIMM single in line memory module sockets that can be configured with 2 a no cache PCA 8 16 32 or 64 Mbytes of memory If during normal operation a parity error in the SIMM is detected by the disk array which results in a panic the panic code information can be used to determine which SIMM failed The format of a panic code is error code address for example 0x00000012 0x05123454 In this example the error code 0x012 indicates that a RAM parity error was detected at addres
65. Display Verify Results The first of four reports is displayed Viewing Maintenance Information 8 47 Bunoouse qnoJj m xterm Miei X SP SCSI ID 15 S N 9621954 Verify Results for Unit 0 State Sniff Running Percentage Complete 8 Sniffing Enabled Sniff Period x100 ms 30 Corrected Uncorrectable i if FRI I I Time I I Enter Choice I _ Enter N P for HELP to GO TO THE PREVIOUS HENU i Figure 8 16 First Page of a Unit Verify Results Report The four reports available on this menu are accessed by typing N or n for next or P or p for previous The four reports are Historical Total of all Powerfail Recovery Verifies Lists the historical total of all errors detected during recovery of a powerfail situation Currently Running Full Unit Verify Lists the errors detected during a verify for a selected unit Most Recently Completed Full Unit Verify Lists errors detected during the most recently completed verify Historical Total of all Full Unit Verifies Lists historical totals of all errors detected during full verifies of the selected unit non powerfail situations 8 48 Viewing Maintenance Information Each report screen can contain the following information Field Description State Indicates that a Verify or Sniff is currently running Sniffing Indicates that this feature is Enabled or Disabled Complete Indicates the percentage complete for the operation
66. E1 E2 and E3 The box for each compartment containing a disk module shows a three letter status code Table 6 1 If a disk module is not bound the displayed code is UNB with two asterisks The box for a bound module shows the code and logical unit number of the RAID group individual unit or hot spare to which the module is bound A box without text indicates an empty compartment that is one not containing a disk module or one containing a non operational disk module Table 6 1 Disk Module Status Codes Code Disk module status BIN n Binding being bound into logical disk unit LUN No other configuration functions can be performed while disks are being bound 6 36 Presentation Utility Screen Table 6 1 Disk Module Status Codes cont d Code Disk module status ENA n Enabled bound with its primary route via this SP If disk array has another SP this module s box is marked RDY on that SP s display For a hot spare the code is ENA H it means that the hot spare has been integrated into RAID group whose logical disk unit number is shown EQZ n Hot spare only Data is being copied from the hot spare onto a replacement disk module in the logical disk unit whose number is shown FMT n Being hardware formatted Generally modules do not need hardware formatting OFF Powered off by the SP which can happen if the wrong type of module is inserted a message also appears in the Unsoli
67. Errors ZP Zero Pass Count All power up tests and the power up sequence should complete within three minutes Visual indications of power up are e The green READY LED on the front of the disk array illuminates e The fan pack starts operating e The green READY LED on the disk drive modules illuminates e The green READY LED on the power supply units illuminates e The green READY LED on the SPs illuminates e A blinking asterisk appears in the Heartbeat field of the Presentation Utility screen see chapter 6 Grid Manager for details e No FAULT REPLACE or SERVICE LEDs are illuminated If the disk array fails to power up suspect parts are the SP SIMMs or database disk modules If the power up tests fail all CRUs in the disk array must be examined for illuminated FAULT REPLACE or SERVICE LEDs If the power up sequence fails the amber FAULT LED on the disk array s front panel illuminates and the failing component s SERVICE REPLACE or FAULT LED also illuminates In addition to the LEDs illuminating the halted power up test displays a Failed status on the console Table 8 5 identifies causes and solutions to the most common power up failures The Disk Array Installation Troubleshooting Checklist located in chapter 2 contains additional items to check Power Up Troubleshooting 8 15 punoouse qnouj Troubleshooting the Power Up Sequence CAUTION The following troubleshooting steps are intended t
68. Figure 9 2 9 6 xxvii seiqer xxviii Features Models Options and Add On Products Specifications Replaceable Parts Model 10 Replaceable Parts Model 20 Hewlett Packard Software License Terms Regulatory Statements REFERENCE R 1 o2uoJ9JoH Features Compatibility Computer systems HP 9000 Series 700 HP 9000 Series 800 HP 3000 Series 900 Operating systems HP UX 9 04 Version A B9 04 3B or later HP UX 9 05 Patches required HP UX 9 07 Patches required HP UX 10 x HP UX 11 x MPE iX Release 5 5 patches required Microcode Revision 7 16 Revision 7 23 Revision 8 20 Revision 8 27 Revision 8 51 Reference PROM 7 977 PROM 7 977 PROM 7 99 PROM 8 02 PROM 1 49 4 Gbyte HH half height disk support for the SP on model 10 and model 20 only 2 Gbyte disk support for the SP only for the SP only 4 Gbyte LP low profile and 9 Gbyte HH disk support for the SP on model 10 and model 20 only 4 Gbyte HH disk support for the SP620 SCSI SP only Revision 8 56 PROM 1 56 for the SP620 SCSI SP only Revision 8 58 PROM 1 57 4 Gbyte LP and 9 Gbyte HH disk support for the SP620 SCSI SP only Revision 9 26 07 PROM 8 02 4 Gbyte LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP on model 20 only Revision 9 56 07 PROM 1 73 4 Gbyte LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP620 SCSI SP only
69. GO BACK TO PREVIOUS HENU Set password Y N N 1 Enable password mode Y N N Enter new password CR ME for none Verify new password Figure 6 18 Main Menu Password Utility Grid Manager Passwords 6 71 JoDeue N pud Note If the disk array already has a password a prompt will ask that it be entered on the Presentation screen If the password is not known or has been forgotten type clariion_all Now the existing password can be removed or a new password can be typed as explained in the next section Changing a Password To change or remove a password start at the Grid Manager Main menu and follow the menu path 5 Change Parameters 2 Change Storage System Parameters 3 Change Main Menu Password The Password menu displays Answer the screen prompts e xterm Password utility Enter for HELP to GO BACK TO PREVIOUS HENU Enter old password password li Change password Y N N Enable password mode Y N N Enter new password CR ME for none Verify new password Figure 6 19 Change Main Menu Password Utility 6 72 Grid Manager Passwords Enter old password password Change password Y N N Y Enable password mode Y N N Enter new password CR NL for none Verify new password If a password is assigned but password mode is not enabled all of the Grid Manager functions can be performed but a prompt will request the password every time th
70. Garbage on terminal display Check for the following terminal configuration Datacomm config Baud rate 9600 Parity Databits None 8 Terminal config Termmode EM100 Array powering up but no display or garbage on terminal 8 16 Power Up Troubleshooting Check that the correct RS 232 cable part number 5061 5003 is being used Also check the internal address cable for proper orientation to connector Table 8 5 Power up Troubleshooting cont d Condition Symptom Subsystem fault during power up sequence Solution A failed FRU should not stop the initial power up tests Power up should complete and the faulted FRU should be indicated by an FRU FAULT LED Change the faulted FRU after power up tests complete Subsystem fault during power up sequence BBU may have discharged during power down Subsystem fault will be indicated until BBU is fully recharged BBU recharge message can be found in Unsolicited Event Log in Grid Manager FE mode is required to view the log Power Up Troubleshooting 8 17 Buljooysajqnoly Troubleshooting the Disk Array This section describes how to troubleshoot power in the system and the disk array and operate the RIP PCB The problems the array detects with FRUs are indicated by host error log inputs messages under system status messages to the Unsolicited Event Log and visible indications of the failed FRU illuminated fault LEDs Problems with the FRUs
71. Grid Manager Power Up the Disk Array 2 43 uonejjeisuj Disk Array Installation Troubleshooting Checklist Revision 1 3 Check for physical damage g g Disk array packaging and disk array physically undamaged 1 1 1 6 m cabinet packaging and cabinet physically undamaged Check site power OOadUddaud U J Proper use of ground and neutral with correct gauge wire Neutral to ground voltage less than 3 volts Ground connected to main service panel with correct gauge wire Raised floor properly grounded No single wire daisy chaining of ground between power receptacles Power source used exclusively by computer room No power surges or sags Proper voltage on all phases C Other RFI EMI Check site environment g g g g 2 44 Ambient temperature within specifications 20 C to 25 5 C 68 F to 78 F Perforated tiles properly placed to enhance proper airflow No temperature hot spots within computer room Humidity within specifications 40 to 60 non condensing Proper clearance between surrounding products Disk Array Installation Troubleshooting Checklist Revision 1 3 Check SCSI and RS 232 cables and SCSI terminators J No bent or damaged pins on SCSI and RS 232 cables No bent or damaged pins on SCSI terminators No damaged or loose screws on SCSI and RS 232 connectors SCSI and RS 232 cables and terminators tightly secured to disk array bulkhead SCSI cable tightly secured to host a
72. HB56D436BR 7A MH4M36AXJ 7 MHAMS36BXJ 7 HB56D436B 7RB SCM421186SG70 SM5364000G 7DG KMM5364100AG 7 SM536044002P3G7 KMM5364103AKG 7 16 Mbyte SIMM N A 005 040535 N A DC power distribution cable N A 005 040685 N A Address RS 232 PCA N A 111 001497 N A Display orientation jumper N A 123 002825 N A Model 10 20 caster N A 5063 5397 N A Model 10 20 disk slot filler panel N A 005 041695 N A Model 20 rackmount front bezel N A 005 041696 N A Model 20 deskside front bezel Reference R 19 o2ue49JoH Hewlett Packard Software License Terms The disk array hardware described herein use Licensed Internal Code LIC The LIC including any updates or replacements and any LIC utility software are collectively referred to as Software CLARiiON Storage Systems Inc owns the copyrights in the Software and all copies of the Software HP distributes the Software under license The following Terms and Conditions shall govern the Software and Customer s use of the Software DEFINITIONS 1 R 20 a b Software is one or more programs capable of operating on a computer processor or controller which is either listed separately as a Software Product on the Price List included with another Product on the Price List or fixed in hardware and not removable in normal operation Use means storing loading instal
73. In preparation for its weekly BBU test the disk array disabled caching This message is followed by a 0x663 message 0x665 Configured for Single SP The disk array detected change to a single SP configuration Unsolicited Event Log 8 59 punoouse qnouj Table 8 10 Informational Codes contd Event code Description Ox666 Configured for Dual SP The disk array detected change to a dual SP configuration 0x667 Cache Recovering The SP is recovering the contents of the Write Cache following a reboot 0x668 Cache Recovered The SP has successfully completed recovering the contents of the Write Cache following a reboot 0x66A Soft Vault Load Failure The SP could not load the saved contents of the Write Cache from disk However there were no units with saved data so Write cache does not need to be loaded This situation can occur after SP memory is reconfigured 0x66B Single Bit Error Detected The SP has detected a recoverable single bit error 0x66C Fan Fault Detected The SP has detected a single fan fault 0x66D Peer SP timed out The host SP has timed out waiting for the peer SP to reply to a request 0x680 Invalid Data Sector Read The hardware found a bad checksum on a data sector 0x681 Invalid Parity Sector Read The hardware found a bad checksum on a parity sector 0x682 Invalid Sector Read The hardware found a bad checksum on a sector There is not enough inf
74. KEK KKK KKK KKK KKK KK KKK KKK KKK KK KKKKK KKK KKK KKK KK ko B When you are done and want to quit using Grid Manager x type either a Control Y or a Control K at any of the hs B Grid Manager Enter Choice prompts for input m ck ck ck ck ck ck ck kk ck ck ck ck ck ck KKK KKK KKK ck ck ck KKK KKK ck ck ck ck ck ck ck cock ck ck ck ck ok KKK KKK KKK KK KK Type a Carriage Return when you are ready to proceed Figure 6 3 Running GridMgr iX GridMgr iX enforces access limitation based on the logged on user s capabilities which provides some degree of security against unauthorized access to the disk array device configurations System Manager System Operator or Diagnostician Capability is required to run GridMgr iX If GridMgr iX is started without the required capability access is denied 6 12 Initiating Grid Manager on MPEAX as shown in Figure 6 4 and GridMgr iX halts operation and returns to the session s MPE iX CI Command Interpreter run gridmgr pub sys GridMgr iX A 01 01 C COPYRIGHT Hewlett Packard Co 1996 BAT APR 27 1996 6 239 PM ERROR Insufficient user capabilities You must have SM System manager OP System Operator or DI Diagnostician capability in order to run the GridMgr program END OF PROGRAM Figure 6 4 Notification of Insufficient User Capabilities When GridMgr X starts up it normally indicates how many GridMgr Disk Array Ldevs have been configured on the system and list
75. KKK KKK KKK KKK KKK KKK KKK KKK KK KKK KKK KKK b Firmware downloaded SUCCESSFULLY ck ck ck ck ck kc kk kc ck ck ck ck ck KKK KK KKK KKK ck ck ck KKK KKK ck ck ck KKK ck ck ck ck ok cock KKK KKK ko kc ko ko ko ko Please wait for NIKE to reboot ku anaes LU seconds 933339 seconds dimann osos oO BECORBOS crescesse ea 40 seconds chrine ss 90 Seconds ss eseseees 2200 Minute NIKEARRY UTILINFO UTILINFO includes a quick description of the utility how to use it and available commands Example DFDUTIL gt utilinfo UTILINFO for NIKEARRY CLARiiON Disk Array Utility PURPOSE This UTILITY gives HP field engineers low level control of the disk array Among the supported functions are BIND UNBIND and access to several informational mode pages Access to these pages allows the user to do such things as reset a disk array to it s default configuration TO RECONFIGURE a DISK ARRAY 1 Use DISPMAP to display the list of available LUNs Use TESTUNIT to select a LUN array 2 Use FRUINFO to determine which Logical Units LUNs Field Replaceable Units FRUs 0 through 4 are bound to 3 If they are all bound to LUN 0 with bind type 1 RAID 5 the default configuration is set Otherwise use UNBIND to unbind LUNs containing FRU 0 through 4 4 Use BIND to bind disk FRU 0 4 to LUN 0 with a bind type of 1 RAID 5 5 Use FRUINFO MODESENSE Page 2a and RESET to verify new configuration NIKEARRY 7 67 sonsoubeiq p
76. LUN 0 as Ldev 30 for example io ad 30 id hpc2300wd path 10 16 4 1 0 io ld 30 DEV 30 DEVNAME OUTDEV 0 MODE ID HPC2300WD RSIZE 128 DEVTYPE DISC PATH 10 16 4 1 0 MPETYPE 4 MPESUBTYPE 2 CLASS DISC SPOOL io gt lp 10 16 4 1 0 PATH 10 16 4 1 0 LDEV 30 ID HPC2300WD TYPE DISC PMGR SCSI DISK AND ARRAY DM PMGRPRI 8 MGR LOGICAL DEVICE MANAGER MAXIOS 0 5 14 Configuring Disk Array SCSI LUNs for MPEAX Configure disk array LUNs 1 and 2 as Ldevs 31 and 32 for example io ad 31 id hpc2300wd path 10 16 4 1 1 io ad 32 id hpc2300wd path 10 16 4 1 2 For more information refer to the MPE AX System Startup Configuration and Shutdown Reference Manual part number 32650 90042 Configuring Disk Array SCSI LUNs for MPEAX 5 15 006 seues 000 dH uoneiun6yuo 5 16 Configuring Disk Array SCSI LUNs for MPEAX GRID MANAGER About Grid Manager Configuring MUX Ports on HP 9000 Systems Configuring RS 232 Ports on HP 3000 Series 900 Systems Initiating Grid Manager on HP UX Initiating Grid Manager on MPE ix Displaying the Grid Manager Main Menu Grid Manager Menus Presentation Utility Screen Enabling Disabling the SP620 SCSI Storage Processor Creating Logical Disk Units LUNs Unbinding a Logical Disk Unit LUN Setting up Disk Array Caching for Microcode Revision 7 23 Setting u
77. Maintenance Change Caching Parameters FETTE GGG GG GGG GG BG GG GG GG G GG Gg GG G Gg G Gg Gl 9 1 g 4 n ng g gn ng ngn g Change Caching Parameters T FKF TrrgrktgrPb g YFTJEKETFHFErFIAFHrEFUTITIEFI TFT Se A LUN Number 0 X Enable Write Cache j z X Enable Read Cache PEE ES EE LEP EAC on gg LET EEL ELE KLE ET LETT PEEK ng OK J Cancel Help T FETTE GG GGG GG G B GG GG GG GG G G GG G Gg Gg Gg gg gGz gg gn lg gn L nrnngnngng ng Figure 8 27 Cache Parameters Screen Enabling Write Cache Enabling the write cache speeds up write operations to the LUN because the array can store the information in memory without having to wait until it is through servicing other requests before it can commit the data to disk Managing writes to a LUN with cache disabled is normally reserved for applications where it is critical that the data be committed to disk before returning from a write0 system call in case of a power failure or other system problem that might result in lost data In other applications where performance improvement outweighs the importance of absolute data integrity in the case of power loss or other failure run the LUN with write cache enabled Enabling Read Cache Enabling the read cache speeds up read operations from the LUN Each time the LUN is referenced the array can read more information from the disk than was requested After delivering the requested data to the system the extra d
78. Mbyte 10 Mbyte 64 Mbyte 2 Mbyte 41 Mbyte 21 Mbyte Specifying Memory Configuration for Microcode Revision 8 xx 9 xx Table 6 10 Change Storage System Memory Configuration on an SP620 SCSI SP 8 Mbyte Minimum Total cache Buffer Write cache Read cache 8 Mbyte 4 Mbyte 4 Mbyte 1 Mbyte 16 Mbyte 4 Mbyte 8 Mbyte 4 Mbyte 32 Mbyte 4 Mbyte 18 Mbyte 10 Mbyte 64 Mbyte 4 Mbyte 40 Mbyte 20 Mbyte Unless a special performance environment is being set up a cache size equal to the SP memory should be set up Grid Manager does not allow more memory than either SP has to be specified If less than this amount is specified the remaining SP memory is unused Valid cache sizes are 0 4 8 and 16 Mbytes in a model 10 0 4 8 16 32 and 64 Mbytes ina model 20 and 8 16 32 and 64 Mbytes in a model 20 with an SP620 SCSI SP If caching has never been enabled the cache size is 0 A cache size of 0 prevents the disk array from enabling caching on power up Disabling and Enabling Disk Array Caching Because of cache mirroring the maximum cache size is 16 Mbytes for the model 10 disk array and 64 Mbytes for the model 20 On power up the disk array enables caching if the prerequisite hardware is working and the cache size parameter is a valid non zero value Determine whether disk array caching is disabled or enabled by looking at the Cache State status on the Presentation Utility screen Disk array caching can be disabled or enable
79. Packard GmbH Hewlett Packard Japan LTD 8000 Foothills Boulevard Hewlett Packard Str 1 5 5 22 Shinmachi Hino shi Roseville CA 95747 D 71083 Herrenberg Giltstein Tokyo 191 0002 USA Germany Japan Declares that the product Product Name Disk Array Product Numbers A3231A A3232A A3388A A3389A A3539A A3540A A3540AZ A3549A A3550A A3550AZ Product Options All conforms to the following Product Specifications Safety IEC 950 1991 A1 A2 EN60950 1995 A1 A2 A3 EMC CISPR 22 1993 EN 55022 1994 Class A EN 50082 1 1992 Generic Immunity including IEC 801 2 1984 prEN 55024 2 1992 3kV CD 8kV AD IEC 801 3 1992 prEN 50082 2 1992 3 v m IEC 801 4 1988 prEN 55024 4 1992 0 5kV Signal Lines 1 0kV Power Lines supplementary information The product herewith complies with the requirements of the Low Voltage Directive 73 23 EEC and the EMC Directive 89 336 EEC and carries the CE mark accordingly This product was tested in a typical configuration with HP 9000 Series 800 computer system Roseville California May 20 1998 Frank L Sindelar ESSD R Quality Manager European Contact Your local Hewlett Packard Sales and Service Office or Hewlett Packard GmbH Department ZQ Standards Europe Hewlett Packard Str 1 D 71083 Herrenberg Gultstein Reference R 25 o2ue49JoH R 26 Reference PRODUCT DESCRIPTION About the Disk Array About Disk Array Components Disk Arra
80. Recovery eeess 4 10 Multi Bus Configuration Error Recovery 0 0s eee cece eee eens 4 12 Multi Initiator Error Recovery 0 e eee eee ene eens 4 14 Multi Bus Initiator Error Recovery 0 00 e eee ee eee cee eee eee 4 16 Array Configurations 0 0 cece cece eee ene ene me 5 3 Basic Configuration Error Recovery 0 2 eee e eee eee eee eee 5 6 Two SP Configuration Error Recovery eee ee eee eee eee eee 5 8 Two Adapter Two SP Configuration Error Recovery esess 5 10 Multi Bus Configuration Error Recovery 0 0s cece eee eens 5 12 Disk Module Status Codes 0 0 cece eee eee ete eee n eens 6 36 XXV seiqer 6 2 6 3 6 5 6 6 6 7 6 8 6 9 6 10 7 1 7 2 7 3 7 4 7 5 7 6 7 7 7 8 7 10 7 11 7 12 7 13 7 14 7 15 7 16 7 17 7 18 7 19 xxvi Minimum Microcode and PROM Revisions for Supported Disk Modules 6 38 Cursor Movement Keys ave WR pA DRAIN PSU DAR ES 6 41 Restrictions and Recommendations for Binding Disks into LUNs 6 44 Bind Options for Different Groups 0 e eee eee ee eee eee 6 50 SIMM Sizes for Model 10 and Model 20 0 0 00 cee eee eee ee 6 53 SIMM Sizes for Model 10 and Model 20 sees 6 59 Storage System Memory Configuration Values 2 2 eee eee 6 60 Change Storage System Memory Configuration 0 00 e e
81. Replacement 0 002 eee eee eee eens 9 14 Model 10 20 Power Supply Unit Removal Replacement sess 9 19 Model 10 20 BBU Removal Replacement eeeeeeeeeeeeeee 9 21 Model 10 20 SIMM Removal Replacement 0 2 sees eee eee eee 9 24 Model 10 Deskside Component Removal esee 9 29 Outer Covers Figure 9 18 2 0 0 0 occ ec cee een eee ences 9 29 Midplane Figure 9 19 and Figure 9 20 0 0 eee eee eee eee 9 31 Interconnect Printed Circuit Board IPCB Figure 9 21 9 35 LEDs LED Cable Figure 9 22 00 cee cece een eh 9 37 Fan Cable Figure 9 23 2 0 ccc cece eh 9 39 SCSI 2 Cables Figure 9 24 0 cee ec eeeee eens 9 41 Internal AC Harness Figure 9 25 and Figure 9 26 4 9 43 On Off Switch Figure 9 27 sss 9 46 AC Line Filter Figure 9 28 0 0 0 eee cece eee een eens 9 48 Model 10 Rackmount Component Removal 00 0 0 e eee eee 9 50 Remove the Array from the Rack Figure 9 29 000 9 50 Midplane Figure 9 30 and Figure 9 31 0 0 0 eee eee ee eee ee 9 52 Interconnect Printed Circuit Board IPCB Figure 9 32 9 55 LEDs LED Cable Figure 9 33 0 0 cece eee cee eee ene 9 57 Fan Cable Figure 9 34 2 0 0 2 0 eee eee nee eee 9 59 SCSI 2 Cables Figure 9 35 2 0 cee eee eee eee ences 9 62 Internal AC Harness Figure 9 3
82. Table 2 5 SCSI ID Switch Settings cont d 6 ON OFF OFF ON 7 OFF OFF OFF ON 8 ON ON ON OFF 9 OFF ON ON OFF 10 ON OFF ON OFF 11 OFF OFF ON OFF 12 ON ON OFF OFF 13 OFF ON OFF OFF 14 ON OFF OFF OFF 15 OFF OFF OFF OFF Note The SCSI 2 adapter in the host computer system typically is set to SCSI ID 7 Make sure the SP is set to an ID number that is not used by any other device on the same SCSI 2 bus 2 32 Set SCSI IDs for SPs in the Disk Array s Connect Asynchronous Console Cables To bind disk modules into physical disk units LUNs set up disk array caching and monitor disk array operation each SP must be connected to a terminal MUX or DTC port If terminals are being used one terminal is required for each SP If a system MUX port is being used a MUX port is required for each SP If connecting to a DTC on an HP 3000 Series 900 system see Set Up the Disk Array on an HP 3000 Series 900 DTC later in this chapter oooco f g Model 10 SP console ports Model 20 ES SE i O aD co uc c 0c SP console ports Figure 2 12 SP Console Port Locations The configuration and management of each disk array SP are handled through a serial link
83. The fans continue to run for a few seconds 2 Access the rear of the disk array 3 Remove the following a Fan pack Figure 9 6 b SPs Figure 9 9 c Power supplies Figure 9 11 d BBU located under the power supply units Figure 9 13 e All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array f Front panel Figure 9 40 Remove the four retaining screws that hold the chassis to the rack Remove the disk array chassis from the rack Remove the 11 screws from the top metal cover Figure 9 53 SS To Remove the midplane a Remove the two screws that secure the midplane to the chassis One screw at each end of the top of the midplane b Disconnect the LED cable from the midplane c Remove the RIP PCA retaining screw Figure 9 54 d Lift the midplane up approximately 1 2 inch e Tilt the midplane approximately 2 inches and remove the RIP PCA f Disconnect all cables from the midplane g Be careful not to damage the midplane connectors Model 20 Rackmount Component Removal 9 91 jueuie2e deH pue jeAoway Note The presence or absence of a jumper on plug P1 changes the orientation of the Grid Manager Presentation Utility screen If a jumper is on P1 Grid Manager displays a deskside system If there is no jumper Grid Manager displays a rackmount system When replacing the midplane take the jumper off the old midplane and place it in the same orientatio
84. This field is at the bottom left of the screen For disk arrays with caching enabled the number n in ADP n and BDP n show the percentage of dirty pages modified but yet unwritten in the memory of SP A and SP B respectively CAUTION Never disable both SPs at the same time because data can be lost The cache status lines ADP SP A s dirty pages and BDP SP B s dirty pages give the percentages of SP A s and SP B s cache pages that were modified but not yet written to disk The UDP unassigned dirty pages line gives the percentage of cache pages belonging to a LUN not enabled for either SP For example the dirty pages belonging to a failed SP become unassigned When disk array caching is enabled and an SP accesses a LUN with caching enabled either the ADP or BDP percentage should change depending on whether the LUN being accessed is owned by SP A or SP B The number n in UDP n indicates the percentage of unassigned dirty pages in the memories of all SPs Unassigned dirty pages are dirty pages belonging to a LUN not enabled for either SP that is not accessible from either SP Unassigned pages would result for example if an SP fails and its cache contains dirty pages In such a case the dirty pages belonging to the failed SP become unassigned pages If the LUNs owned by the failed SP are transferred to the working SP any unassigned pages for those LUNs transfer automatically to the working SP The working SP writes these unassigned p
85. VT100 emulation capability GridMgr iX allows Grid Manager on the disk array to be accessed from a remote system site using dial up modem vt3k or DSLINE network access It also allows Grid Manager to be accessed from HP workstations using the X windows facility GridMgr iX uses the NUMGR 3000 program to configure the Ldevs on the HP 3000 Series 900 system and the RS 232 I O connections to the disk array A named device class is used to designate device ports as Grid Manager devices When GridMgr X starts up it automatically locates all the Grid Manager devices configured on the system and displays them in a menu The user can select any SP on any disk array attached to any system that can be monitored from the RS 232 connection Initiating Grid Manager on MPEAX 6 11 Jobeue N pus GridMgr iX is run from the program file gridmgr pub sys Figure 6 3 shows the GridMgr iX main screen run gridmgr pub sys GridMgr iX A 01 01 C COPYRIGHT Hewlett Packard Co 1996 FRI APR 26 1996 98 30 AM There are 29 Grid Manager Disk Array Ldevs configured on this system Their Ldev numbers are 102 103 104 106 107 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 Which of these Disk Array Ldevs do you want to use You can select Ldev number zero to mean none of them Ldev number 103 MANAGER PUB SYS on Ldev 5 GridMgr Nike Ldev 103 User Ldev 5 KKK KKK KKK KKK KKK
86. Write Cache Size in MB 0 SP A Read Cache Size in MB 0 SP B Read Cache Size in MB 0 Optimized RAID 3 Memory Size in MB 0 3 Change Main Menu Password Displays dialogue for changing password 4 Change System Date and Time Displays dialogue for changing date and time 6 30 Grid Manager Menus FE 5 Select Storage System Options Package Displays dialogue for changing system type FE 6 Change Storage System Serial Number Displays dialogue for changing the stored serial number 3 Change SP Parameters 1 Enable Disable Statistics Logging Displays dialogue for controlling statistics logging 2 Enable Disable Periodic Error Reporting Displays dialogue for controlling periodic error reporting 3 Change Console Parameters Displays several modifiable console parameters default is 9600 baud 8 bit no parity 4 Update SP PROMs and Reboot SP Displays dialogue for updating SP PROMs system cache must first be disabled 5 Soft Disable Enable SP Displays dialogue for enabling and disabling the SP FE 6 Change Host Interface Options Addressing model Target initiated negotiation Substitute busy for qfull DISABLED ENABLED Mode page 8 DISABLED ENABLED ENABLED Recovered error log reporting DISABLED ENABLED ENABLED LUN TARGET MIXED Allow non mirrored Cache DISABLED ENABLED DISABLED DISABLED ENABLED LUN ENABLED DISABLED Auto trespass DISA
87. Write performance disk module failure Individual disk unit 1 0 1 0 No RAID 0 group with 3 Up to 3 0 with Up to 3 0 with No modules 3 module group 3 module group RAID 5 group with 5 Up to 5 0 with Up to 1 25 with Yes disk modules 5 module group 5 module group RAID 1 mirrored pair 1 0 1 0 Yes RAID 1 0 group with 6 Up to 3 0 with Up to 3 0 with Yes modules 6 module group 6 module group RAID 1 0 group with 10 Up to 5 0 with Up to 5 0 with Yes modules 10 module group 10 module group A RAID 5 group individual access array provides simultaneous I O operations to more than one disk module at a time On write operations these configurations may perform as many as four I O operations e Read user data e Read parity information Planning Disk Unit Configurations 1 41 uonduioseg jonpold e Write user data e Calculate and write parity information Because some I O operations occur simultaneously performance depends on the number of disk modules in the group The I O performance of RAID 5 benefits from write caching A RAID 1 mirrored pair requires one I O operation for a read and two I O operations for a write one to each disk module in the pair A RAID 0 group non redundant individual access array or RAID 1 0 group mirrored RAID 0 group can have as many I O operations occurring simultaneously as there are disk modules in the group The performance of a RAID 1 0 group is slower than that of a RAID 0 group because a RAID
88. able to help determine an Planning Disk Unit Configurations 1 47 uonduioseg jonpold optimum value for your environment Otherwise the default value of 128 sectors is recommended If SP cache has been purchased determine how it is to be configured The SP caching option requires redundant SPs because the cache is mirrored plus a fully charged BBU a For the model 10 disks must be housed in slots A0 A1 A2 A3 and A4 For the model 20 disks must be housed in slots AO BO C0 DO and EO These disk modules are used as cache vault drives b In general the cache size must be matched to the amount of SP memory in each controller 8 16 32 or 64 Mbytes c The cache page size may be fine tuned to meet the needs of your computing environment Legal values are 2 4 8 or 16 Kbytes For a general purpose HP UX system 8 Kbytes is recommended If primarily running database applications 2 or 4 Kbytes are generally optimal but it is best to consult the database application documentation Note Setting system cache to both read and write cache is recommended for RAID 5 T configurations because this will provide best overall performance Fill in the Hardware Path and Device File columns HP UX 9 04 considering host system slots the SP target addresses and the desired LUN numbers Likewise fill in the Primary Path Alternate Path and Device File columns HP UX 10 x and later Note Ifthe array is to be configure
89. an unrecoverable media error on the LUN in question 0x04 Hardware Error 04 00 00 The command could not be completed by the SP because of an unrecoverable error This can be the result of a single disk failure in a non redundant LUN RAID O or individual disk or multiple disk failures in a redundant LUN RAID 1 RAID 3 RAID 5 or RAID 1 0 04 05 00 Same as 04 00 00 This status is used in place of 04 00 00 for some nonstandard system types 04 3A 00 An SP microcode update was attempted but there were fewer than two microcode storing disks in the cabinet at the time Two disks are required so that the new code does not override all copies of the current version of the code in case the new code has a problem If the new code does not load correctly the SP will reload the old code 04 44 00 A drive firmware update failed on one of the internal disk drives 0x05 Illegal Request 05 04 00 The LUN specified in the Identify message accompanying the command descriptor block CDB is not assigned to this SP The initiator can assume that the Auto Assign attempt failed 05 1A 00 The length of a parameter on a SCSI Mode Select for example is incorrect 05 20 00 The opcode field in the CDB is invalid 05 21 00 The Logical Block Address specified in the CDB is out of range 05 24 00 A field in the CDB is invalid 05 25 00 The LUN specified in the Identify message accompanying the CDB does
90. and the High Availability Storage System Monitor dm ses enclosure for complete event notification To install and configure the required monitors refer to the EMS Hardware Monitors User s Guide 7 132 EMS Hardware Event Monitoring TROUBLESHOOTING About Field Replaceable Units FRUs Model 10 Deskside Cabinet Model 20 Deskside Cabinet Model 10 Rackmount Chassis Model 20 Rackmount Chassis Front View Disk Array LEDs Rear View Disk Array LEDs Power Up Troubleshooting Troubleshooting the Disk Array Upgrading from an SP to a SP620 SCSI SP Cache Configuration Instructions Power Block Diagram Viewing Maintenance Information Unsolicited Event Log Troubleshooting with Grid Manager Troubleshooting with SAM Creating a File System on a Series 700 System 8 1 Buljooysajqnoly About Field Replaceable Units FRUs The disk array comes in two configurations e The deskside cabinet e The rackmount chassis The disk array contains seven major FRUs Table 8 1 explains who should replace a given component Table 8 1 Table of FRUs Customer Engineer Customer Component replaceable replaceable Disk module Yes Yes Power supply Yes Yes SP Yes Yes Fan pack Yes Yes BBU Yes Yes Interconnect PCB model 10 Yes No RIP PCA model 20 Yes No Midplane Yes No Note Other components are replaceable only by trained Customer E
91. cabinet documentation for information Install a Rackmount Disk Array in a Cabinet 2 13 uonejjeisuj Attach the mounting rails to the cabinet as shown in Figure 2 3 a Determine the location in the cabinet where the disk array is to be mounted Note that the model 10 is seven EIA units the model 20 is nine EIA units b Install four clip nuts shown as A in Figure 2 3 over the two holes in the front and the two holes in the back of the cabinet at the height where the left rail is to be mounted c Mount the left rail by inserting the tabs at the front and rear of the cabinet B in Figure 2 3 d Install the four screws to anchor the rail in the cabinet C in Figure 2 3 e Install the right rail in the same manner Front of cabinet Ooocooooooocooo000000 Rear of cabinet A Install clip nuts 9eeseesessess9e9 Insert rail tabs in slots C Install screws in rails O00 OOOOOOOOOOCO Figure 2 3 Mounting the Rails 2 14 Install a Rackmount Disk Array in a Cabinet 6 Mount the disk array chassis in the cabinet Figure 2 4 and Figure 2 5 a Count up seven holes from the rail and slide a clip nut over this hole For the model 10 count up eight more holes shown as A in Figure 2 4 for the model 20 count up 13 more holes A in Figure 2 5 and slide a second clip nut over this hole Perform this step for both sides of the cabinet installing a total of
92. cable Close the fan pack Slide the lock unlock lever to the locked position Model 10 20 Fan Pack Removal Replacement 9 11 jueuie2e deH pue jeAoway Locking lever Replace light amber Model 10 fan module Model 20 fan module Figure 9 5 Model 10 20 Fan Packs Disconnect fan Figure 9 6 Removing a Fan Pack 9 12 Model 10 20 Fan Pack Removal Replacement Connect fan cable Figure 9 7 Installing a Fan Pack Model 10 20 Fan Pack Removal Replacement 9 13 jueuie2e deH pue jeAoway Model 10 20 SP Removal Replacement Install a replacement or add an additional SP by performing the following steps 1 Observe antistatic precautions CAUTION It is recommended that the ESD wrist strap be grounded to the disk array chassis or suitable ground and strapped to your wrist when removing or handling the SP and that the SP be placed on the ESD conductive sheet when it is removed from its carton Use ESD kit P N 5182 4119 2 Removing an SP Figure 9 9 a 9 14 Disable disk cache to force the unposted data to be written to the disk modules If this is not done the unposted data will be lost See Disabling and Enabling Disk Array Caching in chapter 6 for information Open the fan pack Set the SP power switch to disable The SP620 SCSI SP does not have an enable disable switch It is disabled using Grid Manage
93. caching caching is available on a per LUN basis Three options are available for LUN caching e Read and write cache can be enabled e Read or write cache can be enabled e Read and write cache can be disabled It is recommended that both read and write LUN caching be enabled LUN caching cannot be operational unless storage system cache also is enabled Enabled LUN caches randomly share the storage system cache pool No specific cache addresses are assigned on a LUN by LUN basis The storage system cache state can be observed on the Grid Manager Presentation screen See chapter 6 Grid Manager for more information Use the Grid Manager utility or the SAM utility described in chapter 7 to enable disk array caching specify basic cache parameters and enable or disable read and or write LUN caching Configurations with Two SPs or Two Hosts or Both A number of configurations with two SPs two hosts or two hosts and two SPs provide higher availability and better array performance than the basic configuration in which one host with one SCSI 2 adapter is connected to a single SP over one external Fast Wide differential SCSI 2 bus A two SP configuration in which one host with one SCSI 2 adapter is connected by one external SCSI 2 bus to two SPs provides higher availability than the basic configuration 1 20 Disk Array High Availability Features because a second redundant SP dual SP is available if an SP fails Any configuration
94. continue through the remaining healthy path once auto trespass has occurred however the failed hardware must be replaced as soon as possible LVM has an automatic switchback feature that moves I O traffic back to the primary path once it is operational Auto trespass is supported on HP UX 10 x and later 1 22 Disk Array High Availability Features Planning Disk Unit Configurations RAID Concepts RAID technology defines some new terms and uses existing terms in new ways This section defines the most important RAID concepts RAID technology groups individual disk modules into one logical disk unit also called RAID group LUN or Logical Unit Number and can appear as physical disk unit on some Grid Manager screens to improve reliability and or performance RAID technology supports high availability either by making two copies of all user data written to disk or by using parity information to recreate the user data if a disk module fails in the RAID group The disk modules can be combined into the following storage units e RAID 0 e 6RAID 1 e 6RAID 1 0 e RAID 5 e ndividual disk unit e Hotspare Note RAID 0 and individual disk units are not supported on models A3549 or A3550 because they do not provide data redundancy For microcode revision 8 56 and later these RAID levels can only be configured in FE mode with the Grid Manager utility See chapter 6 Grid Manager for more information about FE mode Individu
95. controllers only database drives Disk Firmware N A Supplies firmware to all compatible disk modules within the SCSI disk array that is those with the same product ID This applies even to disk modules owned by another SP The firmware then resides permanently within those disk modules until another disk module firmware update is performed Access the Current Revisions screen Figure 7 33 by selecting List Current Revisions from the Firmware Update Tool menu Support Tools Manager STM 7 109 sonsoubeig pue seni ISOH e hpterm CI xj Firmware Update Tool for device type C2300 Array on path 874 9 File List Update Help INSTRUCTIONS 1 Use File commands to select firmware file to be used 2 Use List commands to display current information 3 Use Update commands to start update controller disk a Controller updates should have peer SP if any enabled b Disk updates must have peer SP if any disabled CURRENT REVISIONS Rev Mcode Rv Prom Rev This SP Peer SP Array Dt Array Tm Serial HP 2 9 25 00 8 02 00 A 9 B lt 10 04 28 88 11 25 15 9621954 Note The array does NOT support commands to disks participating in active Hot Swap operations Such disks are excluded from the following list Rev Fru Vendor Product ID ApprxCap HP1A RO IBM DGHSOSY CLRROS 8 8 gb Figure 7 33 Current Revisions Display Updating Microcode and Firmware To use the Firmware Update Tool the m
96. cstm Command line for use in scripts usr sbin cstm The following sections describe the xstm and mstm interfaces for the System and Tools menus The same functionality is available through commands in the cstm interface 7 68 Support Tools Manager STM xstm Interface Each SP in the SCSI disk array is shown on the system map drawn by STM as an icon labeled C2300 Array A dual SP array is represented by two icons Figure 7 3 on p 71 of this chapter Click on either icon to access the diagnostic tools described below Access these tools through the Tools pull down menu on the system map screen These tools along with an indication of whether a license is required to run each are listed in Table 7 6 To set the license select System License Install HP Only from the xstm menu bar Enter the license password in the dialog box and click OK Table 7 6 Available Support Tools Support tool license required Tool type Description No Yes Information Provides quick access to the most useful information X Firmware Update Updates SP microcode and disk module firmware X Expert Provides low level access to the SCSI disk array for X configuration and troubleshooting Information Tool The Information Tool displays SCSI disk array configuration and status information including Auto trespass SCSI disk array and LUN e Caching parameters SP and LUN e Capacities LUN and disk module e Compo
97. default for each selection is the current setting fiddressing model LUN TARGET HIXED LUN Target initiated negotiation lt DISABLED ENABLED Substitute busy for qfull Hode page 8 ENABLED Allow non mirrored cache fiuto trespass Allow PQ 3 No spaces in PID Spindle Synch Trespass Logging Figure 6 17 Host Interface Options Allow non mirrored cache is a new parameter which is not supported It must always be disabled If the disk array contains only one SP this parameter allows cache to be enabled However caching cannot be mirrored because a single SP is present If the SP fails the data in cache may not be posted onto the disk module which may result in data integrity problems Changing Storage System Parameters 6 69 Jobeue N pus To Check the Host Interface Options Settings 6 70 Enter FE mode by typing CONnTROL P followed by in upper case FE Go to the Grid Manager Main menu Follow the menu path 5 Change Parameters 3 Change SP Parameters 5 Change Host Interface Options Exit FE mode by typing CONTROL P followed by in upper case USER Changing Storage System Parameters Grid Manager Passwords Adding a Password To add a password start at the Grid Manager Main menu and follow the menu path 5 Change Parameters gt 2 Change Storage System Parameters 3 Change Main Menu Password The Password menu displays Answer the screen prompts e xterm Password utility Enter for HELP to
98. disk array Before updating the options package cache must be disabled When cache is disabled its contents are dumped to the cache drives When the option package update has been completed a microcode panic may be reported by the peer controller This panic is caused by the disk array trying to dump cache at the start of the reboot Even though the panic may be reported the cache data was saved prior to the reboot and the options package was installed successfully This is an example of the Unsolicited Event Log with the panic code Event Data CRU Event Message 1 12 22 94 13 45 14 SP B 0x907 Microcode Panic 0x00800077 Extended Status 0x008db414 To Set the Array to the Desired Option 1 Enter FE mode by typing ConTROL P followed by in upper case FE 2 Go to the Grid Manager Main menu 3 Follow the menu path 5 Change Parameters gt 2 Change Storage System Parameters 1 Change Storage System Cache Parameters 4 Disable cache if it is enabled 5 Onthe same screen select option 4 Change Storage System Option Package and press RETURN On an HP UX 9 04 HP UX 9 05 9 07 or MPE iX 5 5 system type A for the system type On an HP UX 10 x system type 2 as the system type 6 66 Changing Storage System Parameters After the correct option has been entered the array reboots Cache is enabled automatically and FE mode is turned off after the reboot completes 6 Reboot the HP UX operating system so t
99. eee eee ee een eee eee 7 8 Fan Failure Message sees 7 8 Controller Failure Message 2 0 cece eee ee ene eens 7 8 Power Supply Failure Message 0 cece eee cece eee eee 7 9 BBU Failure Message 0 cece cee ne 7 9 HARAYMON Error Reporting 20 e eects 7 10 HARAYMON on HP UX 9 0x versus HP UX 10 x 2 0 0 eee ee eee 7 10 HARRAY Section 10 Diagnostic Trouble Tree 20005 7 12 HARRAY Section 17 External Exerciser 2 0 0 e ee ee eee ee eee 1 13 AUTO TRESPASS HP UX 10 x and later 00 0 0 ee eee ee 7 14 CAPACITY wre usa hig utils telat wate tee ee eet eR 7 15 CLEAR LOGS wicks e4h4 0 ian Sere ola wie bred whee EH pees dee 7 16 CONTROLLER REP Us Bete eee ER NER de eese 7 17 DESCRIBE 5 Cos gt cee edd Syd Eee ses sh 1 18 DIAG m RPelzkmbImeuriwhetkgeNe hae ae dem bets Im 7 19 DOWNLOAD 252532 Dr pelea eed be rtr tee en drei UE Sa e eda 1 20 Updating Storage Processor Microcode for Non Root Disk Arrays 7 20 Updating Disk Firmware 0 0 cece eee e 7 23 EXIT uuu Wiikey soi eedem CE EE EIER ime ee pete 7 25 HELP vpusti yh seed eke 2 aba Se Gal Mees ba aes 7 26 INQUIRY oct MAS ED e URTREUNSRERWGUGUR T REREETE PURIS 1 28 LCUN REPORT ele tees aeter pate denn 1 29 MECH x uua Ede BIER ku MERE Us eee LU as 7 31 READ netus qr E EE emper dus 1 36 REQUEST SENSE ek ene Goede edie ad x Ra dada 7 37 REVISION gt ciccaie
100. failure of a single disk module SP No 9 04 Yes 10 x HP UX 9 04 If one SP fails the System Administrator must reassign ownership of the failed SP s LUNs to the operational SP by using Grid Manager Also the path to the operational SP must be used and can be changed by using LVM Once these changes have been made access to all LUNs on the array is possible HP UX 10 x If one SP fails ownership of the failed SP s LUNs can transfer automatically to the remaining operational SP If a primary and an alternate path have been configured in LVM LVM switches automatically to the alternate path These operations are transparent to the user Fan pack Yes 9 04 Applications continue running System Administrator or service provider replaces module Power supply Yes 9 04 9 04 Yes 10 x 9 04 Yes 10 x If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply The System Administrator restarts the operating system if necessary SCSI 2 adapter cable No 9 04 No 10 x If the SCSI 2 adapter fails no I O operations are possible between the host and the array Authorized service provider replaces adapter or cable and System Administrator restarts operating system and applications If the SCSI 2 adapter fails no I O operations are possible between the host and the array Can regain access t
101. four clip nuts b Slide the chassis onto rails and into cabinet B in Figure 2 4 and Figure 2 5 WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to install it into a cabinet CAUTION Do not lift the disk array chassis by the fan pack handle at the back of the disk array because doing so can damage the fan pack c Fasten front of chassis into cabinet with four screws C in Figure 2 4 and Figure 2 5 d Insert the keeper tab in the slot at the rear of each rail and fasten the rear of the disk array chassis to this tab D in Figure 2 4 and Figure 2 5 Note The disk array must have an SP in slot SP A If the disk array has a single SP in which case the disk array can be used only in a basic configuration make sure the SP is installed in slot SP A If an SP must be installed in slot SP A see Model 10 20 SP Removal Replacement in chapter 9 7 Reinstall any disk modules that were removed refer to Model 10 20 Disk Module Removal Replacement in chapter 9 for instructions being sure to put each module back in its original slot Install a Rackmount Disk Array in a Cabinet 2 15 uonejeisuj CAUTION Handle disk modules carefully as they can be damaged easily 8 Reinstall the disk array s front panel Figure 2 2 9 Reinstall any trim removed from the cabinet and close any cabinet doors 10 If installing a rackmount model 10 disk array set
102. hot spare the original failed disk module can be replaced The rebuild on the hot spare continues to completion Once the rebuild has completed the disk to disk copy begins immediately Table 8 7 Sense Key Additional Sense Code Additional Sense Code Qualifier Values Not a Complete List Sense key Description 0x00 No Sense 00 00 00 There is no sense information to return at this time for this LUN 0x01 Recovered Error 01 5D 00 The SP is returning an Unsolicited Sense Block to the host See Recovered Error Log Reporting on p 66 of this chapter This sense key is used only when Recovered Error Reporting is enabled 0x02 Not Ready 02 04 00 Logical unit not ready Cause is not reportable 02 04 02 Logical unit not ready Initializing command is required 8 24 Troubleshooting the Disk Array Table 8 7 Sense Key Additional Sense Code Additional Sense Code Qualifier Values Not a Complete List cont d Sense key Description 02 04 01 The LUN is reporting that it is becoming ready performing an Assign The initiator should retry after a short time interval such as 500 ms 02 04 03 The LUN reporting this sense data is indicating it is inoperable and manual intervention may be necessary 0x03 Media Error 03 11 04 Data cannot be returned for a read operation because of an unrecoverable media error on the LUN in question 03 11 0B A Verify operation failed because of
103. in the array are not as easily detected e Problems with the midplane may indicate that other non faulted FRUs in the array have failed e Afaulty RIP PCB can be determined faulty when a redundant SP in the array is found to be at fault and nothing reported to the host that the SP had an error during normal operation of the array General Troubleshooting Information The following is a list of the current media types and manufacturing part numbers for HP UX 9 04 updates These media cannot be ordered from SMO Media Type Mfg Part Number QIC Tap B3108 13855 DDS Tape B3108 13472 CD ROM B3108 14000 SAM does not require RS 232 port connections when running on HSC bus platforms running HP UX 10 10 or greater RS 232 port connections are required for NIO systems running HP UX 10 10 Neither HSC nor NIO platforms require RS 232 connections for SAM when running HP UX 10 20 Following are the disk firmware levels at the time of publication Refer to Service Notes for updates 8 18 Troubleshooting the Disk Array Product Number ST12550N ST32550N ST15150N DCHS04F DGHS04Y DCHSO9F DGHS09Y DNESO9Y DGHS18Y DNES18Y Description Firmware Level 1 6 2 Gbyte Seagate HP03 1 2 Gbyte Seagate HP08 1 6 4 Gbyte Seagate HP03 1 4 Gbyte IBM 1 4 Gbyte LP IBM HP53 HP3C 1 6 9 Gbyte HH IBM HP53 1 9 Gbyte LP IBM 1 9 Gbyte LP IBM HP1A HPJO 1 6 18 Gbyte HH IBM HP1A 1 6 18 Gbyte LP IBM HPJO Following are th
104. in which they were bound Disk arrays are shipped from the factory with the disk modules bound into RAID 5 LUNs RAID technology is explained later in this chapter Figure 1 6 Disk Module Storage Control Processor The storage control processor also called a storage processor or storage controller and generically referred to as an SP is a printed circuit board that resides in the disk array cabinet Two SP models are available for the model 20 disk array One which is referred to as an SP620 SCSI SP where it is necessary to distinguish between the two models provides slightly better performance under certain applications SP and SP620 SCSI SP boards are not interchangeable and cannot be mixed in the same disk array If an SP fails it 1 10 About Disk Array Components can be removed and replaced with array power on but it must be replaced with an SP of the same type For high availability the disk array can support two SPs The second SP provides a second path to the disk array so both SPs can connect to the same host or two different hosts With two SPs the disk array can support disk array caching in which each SP temporarily stores modified data in its memory and writes the data to disk at the most expedient time Four sockets located on the SP printed circuit board are available for adding additional single in line memory modules SIMMs which are used for caching SIMMs for the SP are interchangeable
105. individual disk go to step 9 6 Usethe arrow keys or the space bar on the keyboard to select the next disk to bind and press RETURN If configuring a mirrored pair skip to step 9 7 Highlight and select the appropriate number of disks to be bound in the RAID 0 RAID 1 0 or RAID 5 group 8 Type b at the Enter Choice prompt and press RETURN 9 This produces the Bind Options menu Refer to Table 6 5 to review specific bind options for the logical disk unit type to be configured Press RETURN in response to the questions on the screen If selecting default values proceed to the note in Enable Caches LUN Cache on p 48 of this chapter 6 46 Creating Logical Disk Units LUNs Bind RAID 5 Group Independent Access Array Enter for HELP to GO TO BIND HENU Enter Choice i Figure 6 11 Grid for Binding Modules for the Model 20 Logical Unit Number LUN The default logical unit number also called Logical Unit Number LUN is the next hexadecimal number available starting with 0 The first LUN bound regardless of the number of SPs or hosts attached to this disk array has the default LUN number 0 the second 1 the third 2 and so on It is possible to specify a non default number hexadecimal For the next LUN bound the default number will be the lowest number skipped When binding a hot spare the LUN number assigned is used only while the hot spare is on standby When the hot spare is integrated into a disk grou
106. is eight e The LUNs can be numbered from 0 through 7 e Most S700 systems using the disk arrays will be unable to support them remotely due to the lack of available ports for connecting to the disk array s serial port The disk array s serial port must remain connected to an EM100 capable terminal in order to access Grid Manager Any disk array failure probably will require an on site visit from an HP service representative e Multi bus connections and multi initiator connections are not supported on HP UX 9 05 and 9 07 software platforms e VMis not available on HP UX 9 05 and 9 07 e SAM support is not available on HP UX 9 05 and 9 07 for the disk array File systems must be built on the array using HP UX commands Auto trespass is not supported on HP UX 9 05 and 9 07 and should be disabled when the disk array is installed by setting the Storage Systems Option Package to option A by using Grid Manager see Changing Storage Systems Option Package on the Disk Array in chapter 6 e Supported host adapters are Fast Wide Differential Core I O on Series 735 755 Fast Wide Differential Core I O on J Series Add on Fast Wide Differential Adapter A4107A 3 4 Installing and Reconfiguring on S700 Systems Disk Array Configurations Three hardware configurations are available on the HP9000 Series 700 Table 3 1 lists the hardware components that make up each configuration and summarizes the features of each A section for eac
107. key menu appears across the bottom of the screen b Press F8 config keys c Press F5 terminal config d Press the arrow keys to position the cursor at the TermMode entry e Press F2 next choice repeatedly until EM100 or VT100 appears in the TermMode field then press F1 save config f Check for an indication at the bottom of the screen indicating EM100 mode or VT100 g Access the datacomm configuration screen and set the baud rate to 9600 and the parity databits to NONE 8 h Press F1 save config to save the configuration 3 Power up the disk array The boot screen should display Changes can be made to the factory default configuration by connecting the disk array to a terminal without connecting the disk array to the host computer system with SCSI 2 cables A terminal can be directly connected to a Series 700 system by using an HP Core I O connector to RS 232 array cable 24542H one per SP to attach to an RS 232 port on the core I O card Terminal Emulation through a MUX on HP UX For users who want the greatest configuration capabilities this option provides a way to access Grid Manager using either the system console or a remote terminal without having a dedicated hard terminal for configuration purposes To use terminal emulation to access the Grid Manager configuration tool on Series 700 and 800 systems it is necessary to set up a serial link to each disk array controller console port via a serial multiple
108. kk kk Sk kk Sk kk kk Sk kk ck Sk ck ck ck ck Sk kA kA A amp Indx Path Product ID Also Known As AKA 0 10 4 8 15 0 HPC2300WD disk array Ctrl SP620 Controller PHOENIX 1 10 4 8 14 0 HPC2300WD disk array Ctrl SP620 Controller PHOENIX 2 10 4 8 15 0 HPC2300WD disk array Ctrl SP620 Controller PHOENIX Note Indx Index number used for referencing the device Enter the NIKE Index will display indices Default will quit 0 Kk kk kk eee ek ke kk ck ccc ccc ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck sk sk sk sk sk ko kx kx About to download firmware SP 955 01 onto 10 4 8 15 0 kkkxkxkxkxkxkxkxkxkxkxkxkkxkxkxkxkkxkxkkkxkkkxkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkxk Do you wish to continue with the download y n q y KKKKKKKKKKKKKKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK sk sk kx kx kx 5 Firmware downloaded SUCCESSFULLY ck ck ck ck ck kc kk kc ck ck ck KKK KKK KKK KKK KKK KKK KKK KKK KKK ck ck ck KKK KKK KKK ko ko ko ko ko ko ko Please wait for NIKE to reboot igs wee aa ee 10 seconds PE uu s duce 20 seconds s hs s e IS e US 30 seconds eU 40 seconds Sivas corta te e 50 seconds Pow vg um 1 00 minute 7 54 NIKEARRY FRU Mapping Figure 7 1 and Figure 7 2 illustrate how disk array slot IDs map to FRU IDs FRU FRU FRU FRU FRU FRU PRU FRU PRU FRU 1 z 3 4 12 13 14 15 16 AD BO CD D ED A2 B2 C2 D2 E2 FRU FRU FRU FRU FRU FRU FRU PRU
109. male cable e Option 804 10 0 m 68 pin HD male to 68 pin HD male cable e Option 840 2 0m V cable 68 pin HD male Cable options for HP 9000 V Class Enterprise Servers e Option 851 Part A4801 63002 10 0 m cable with in line terminator e Option 871 Part A4801 63010 2 5 m V cable with in line terminator e Option 873 Part A4801 63012 2 3 m V cable with in line terminator e Option 875 Part A4801 63004 5 0 m cable with in line terminator Upgrade and Add On Products e A3234A 16 Mbyte write cache module must be ordered in sets of four e A3235A 4 Mbyte write cache module must be ordered in sets of four e A3237A Fast Wide SCSI 2 cache ready SP for model 20 R 6 Reference e A3238A DC power supply for model 20 e A32D4A BBU for model 20 e A3396A 4 Gbyte HH disk module e A3666A 4 Gbyte LP disk module e A5169A 4 Gbyte LP disk module for model 20 only Note 9 Gbyte and 18 Gbyte disk modules are not supported on MPE iX or the model 10 disk array e A35D4A1A 9 Gbyte HH disk module e A5132A 9 Gbyte LP disk module e Ab5133A 18 Gbyte HH disk module e A3551A Add on Fast Wide SCSI 2 SP620 SCSI SP PCA Supported on model 20 only Cache option is mandatory 006 64 Mbyte cache 4 x 16 Mbyte SIMM cards Reference R 7 o2ue49JoH Specifications Dimensions Model Height Width Depth Rackmount 26 7 cm 10 5 inches 48 3 cm 19 0 inches 76 2 cm 30 0 inches model 10 Deskside 62 9 cm 24 8 inch
110. modules model 10 and model 20 two power supplies required both operate but only one is needed If one fails the second N 1 redundant can continue normal operation of the disk array without disruption Two power supplies however are required to power on the model 20 disk array regardless of how many disk modules are installed e More than 10 disk modules model 20 only three power supplies required all operate but only two are needed If one fails two of the three power supplies can continue normal operation of the disk array without disruption 1 14 About Disk Array Components The power supplies take the AC input voltage and convert it to 18 and 24 VDC on the model 10 and to 18 24 and 48 VDC on the model 20 The DC voltages provided by the power supplies are used by the fan pack SPs and disk modules Power supplies are not interchangeable between the model 10 and model 20 disk arrays The power supplies are customer replaceable and hot replaceable A failed power supply can be replaced without disrupting an operating disk array provided the array contains redundant power supplies RN 7 259250 aca CE p EE E KRI B XR RE E SEEN p amm REED OOOOOO00 O2 OO OQOQQOOOO 252929 OOOOOOOCO QOOOOOO00 OOOOOOOO OOOOCOOCOO OOCOOOCOOO OOOOOOO OOOOOOOOOOOOOOOOOOOO OOCOOOOOOOOOOOCOCOOOOCOCOCO OOOOOOOOOOOOOOCOOOOOQ OOOOOOOOOOOOOOOCOOOOCO Om OOCOOOOOOOOOOOOO 9 Oo OOOOOOOOOOOOOOCO
111. n n 2 1 RDY RAID n n 2 2 RIY Spare 1 3 BIN RAIDS Off Off 3 System Cache States Write ENABLED Read ENABLED Enter Choice Jj Enter N P R for HELP to GO TO THE PREVIOUS HENU Figure 8 17 LUN Configuration Summary Screen To navigate the LUN Configuration Summary screen type the following single letter commands Key Action Norn Moves to the next screen There are more screens if there are more LUNs than will fit onto one screen Porp Moves to the previous screen Rorr Refreshes the screen contents to reflect the current state of the disk array Disk Failure Information To display information on the most recent disk failures enter FE mode Then from the Grid Manager Main menu follow the menu path 4 View Maintenance Information 5 Display Disk Failure Information This submenu displays four reports for a maximum of 16 failures 8 50 Viewing Maintenance Information The reports are Device Information Block Active Device Request Block Device Queue Element and Dispatcher Information e xterm P x SP A FH SP SCSI ID 15 Display Disk Failure Information S N 9621954 Obituary 4 of 5 Bevice Information Block Date 04 23 1998 Thursday at 14 17 46 00 Bus Number 0 Fru Number 0 Target ID 0 Slots Available 1 Type 0x41 Hax Active 2 Capacity 4110034 fictive Count 1 Queueing Enable No Queue Hold No Request Count 1270 Spindle Synch Allowed No Fat R
112. number EISA 8 Port MUX Card J2482A The accessory kit for this card is part number J2482 60002 The kit contains a fanout cable with eight RJ45 connectors to provide connection to the disk array with the standard 17355D cable An adapter cable part number J2489A for a package of 16 cables must be used The software needed to use the MUX card is the EISA MUX driver part number J2467A Configuring MUX Ports on HP 9000 Systems 6 5 Jobeue N pus EISA 64 Port MUX Card J2483A The accessory kit for this card is part number J2483 60002 The kit contains the power module adapter which is needed to connect to the disk array with the standard 17355D cable Use the J2489A adapter cable and software as listed in the above paragraph for the 8 port MUX card The RJ45 port module part number J2485A and the port module power supply part number J2486A may also be necessary for this setup Figure 6 1 shows a pinout diagram for the J2489A cable Signal Pin Pin Signal 10 20 9 22 8 4 Signal ground 7 7 Signal ground Transmit data 6 3 Transmit data NC 5 Receive data 4 2 Receive data 3 6 2 8 1 9 10 1 13 1 eeeeeeceeeeo RJ 45 Cable 25 14 End view DB 25 Female end Figure 6 1 Cable J2489A Pinouts 6 6 Configuring MUX Ports on HP 9000 Systems Configuring RS 232 Ports on HP 3000 Series 9
113. operational SP by using Grid Manager Also the path to the operational SP must be used and can be changed by using SYSGEN or ioconfig Once these changes have been made access to all LUNs on the array within the SCSI LUN limits is possible Fan pack Yes Applications continue running System Administrator or service provider replaces module Power supply Yes If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply The System Administrator restarts the operating system if necessary SCSI 2 adapter cable No If the adapter cable fails I O operations to array disk units fail Authorized service provider replaces adapter or cable and System Administrator restarts operating system and applications If the SCSI 2 adapter fails no I O operations are possible between the host and the array Can regain access to LUNs associated with failed SP by making changes manually in Grid Manager and SYSGEN ioconfig 5 8 Two SP Configuration Two Adapter Two SP Configuration The two adapter two SP configuration shown in Figure 5 3 has one host with two SCSI 2 adapters each connected by one external Fast Wide differential SCSI 2 bus to a different SP in the array For better performance with this configuration some logical disk units LUNs can be bound on one SP the rest can be bound on the other SP The SP that binds a
114. or A2 because p1 and s1 are in slots AO and DO p2 and s2 are in slots BO and E0 and p3 and s3 are in slots CO and A2 When binding modules into a RAID 1 0 group it is necessary to select them in the following order pl s1 p2 s2 p3 s3 and so on Rackmount model Deskside model Primary mirror image p1 p2 p3 Secondary mirror image s1 s2 s3 Mirrored pairs p1 and s1 p2 and s2 and p3 and s3 Figure 1 16 Mirrored Modules in a RAID 1 0 Group in a Model 20 Disk Array RAID 5 Group Individual Access Array RAID 5 is similar to RAID 0 except that it adds a data protection feature This configuration the default configuration for the disk arrays usually consists of five modules The model 10 disk array can have three to 10 modules bound as a RAID 5 group The model 20 can have three to 16 modules This type of RAID group uses disk striping With RAID 5 technology the hardware reads and writes parity information to each module in the disk array If a module fails the SP can reconstruct all user data from the user data and parity information on the other disk modules When a failed disk module is replaced 1 30 Planning Disk Unit Configurations the SP automatically rebuilds the disk array using the information stored on the remaining modules The rebuilt disk array contains a replica of the information it would have con
115. pallet with strapping bands Wear protective glasses while cutting them The bands are under tension when cut they spring back and can cause serious eye injury 3 Remove the cardboard box on the top of the system The box contains cables manuals and patch media Remove the four foam corners from the shipping carton Lift the shipping carton off the disk array Cut the plastic tape securing the plywood ramp to the disk array Lower the ramp then lift off the plastic protective covering 9 mU goo Remove the two wing nuts from the foam plywood support near the ramp Push the protruding bolts down through the plywood 9 Grasp the foam plywood support near the head of the arrow 10 Push against the top of the disk array Pull the foam plywood support from beneath the disk array and lower the disk array to the shipping pallet 2 6 Unpack the Disk Array WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array 11 Go to the opposite side of the disk array 12 Gently push the disk array off the foam plywood support on that side of the pallet 13 Carefully roll the disk array off the pallet and down the ramp CAUTION When unpacking a deskside disk array make sure it does not tip over when it rolls down the ramp Check the Anti Tip Legs The model 10 deskside disk array is shipped with anti tip legs attached to each side of the chassis
116. password On the next line type SYSDIAG At the DUI prompt type HARRAY PDEV x x x section 17 where x x x is the PDEV number If HARRAY installed successfully the PDEV is removed from system use and is identified for example PDEV 8 0 0 is a C1300WDR5 HARAYMON Messages HARAYMON messages contain five types of information The date and time of the failure This could be the time the failure was detected by the system rather than the time the component failed As a result the timestamp on the entry in the Grid Manager s Unsolicited Event Log may not match the timestamp on the HARAYMON message Text explaining the type of FRU failure for example Drive Failure or Fan Unit Failure The product id of the disk array reporting the failure The physical device path of the LUN reporting the failure This may not be the LUN that failed The failed unit which is the component that needs to be replaced The failed unit may not be in the path of the LUN reporting the failure but most likely is part of another path High Availability Disk Array Monitor HARAYMON 7 7 sonsouDeig pue seni ISOH CAUTION Always check for failure LEDs before removing a module Disk Module Failure Message If HARAYMON detects a disk module failure it displays a message on the console and e mails a copy to the specified user Following is an example message Mon Jul 11 10 23 17 MST 1994 High Availability Array Monitor Drive Failur
117. reads or writes The length of each transaction can be R Randomly selected lengths U User specified length Enter the transaction length type R U R u Enter transfer length in sectors 1 127 2 100 The sources of the data pattern to be used are U User specified pattern R Random pattern Which pattern source would you like U R R r Generating random pattern Do you want to continue on errors Y N N y Enter the loop count 1 255 1 2 RANDOM MEDIA TEST beginning MON APR 29 1996 4 49 PM RANDOM Test Pass Count 2 4 49 PM The Random Media Test is 25 complete 4 49 PM The Random Media Test is 50 complete 4 49 PM The Random Media Test is 75 complete 4 49 PM The Random Media Test is 100 complete Media Test pass number 1 of 2 complete 4 49 PM The Random Media Test is 25 complete 4 49 PM The Random Media Test is 50 complete 4 50 PM The Random Media Test is 75 complete 4 50 PM The Random Media Test is 100 complete Media Test pass number 2 of 2 complete RANDOM MEDIA TEST UTILITY COMPLETED 4 50 PM HARRAY Section 17 External Exerciser 7 45 sonsoubeiq pue seninf ISOH NIKEARRY NIKEARRY is an ODE Offline Diagnostic Environment based offline utility that is designed to give HP field and manufacturing engineers low level control of the disk array Low level control is provided by a set of functions that gives the user access to several mode pages Acc
118. recovery process management cluster configuration kernel and device configuration networks and communications management and system auditing and security SAM can be used to configure and troubleshoot the disk array Disk array maintenance is available in HP UX 10 0 and later not in HP UX 9 04 9 05 9 07 View Status of an Array Example SAM gt Disks and File Systems gt Disk Devices Actions Disk Array Maintenance gt View Array Status Information 8 90 Troubleshooting with SAM Disk Devices Hardware Volume Total Path Use Group Mbytes Description 52 4 0 Unused 2048 Seagate ST12400 SCSI Disk Drive 52 5 0 LVM vg00 1292 HP C24 4 SCSI Disk Drive 52 6 0 Unused 645 HP C24 39 SCSI Disk Drive 8 5 zee e zm HP C1300 Disk Array Controller 8 5 1 Unused 39 HP C1300 LUN RAID 5 3 disks 8 6 x Ius ass HP C1300 Disk Array Controller 8 6 0 Unused 955 HP C1300 LUN RAID 5 3 disks Figure 8 24 Disk Devices Screen Troubleshooting with SAM 8 91 Bunooyseajqnoly View Array Status Information Controller Information X AO LUN 0 Model Number 0x1CAD A1 LUN 0 Firmware Revision 1 23 0 A2 LUN 0 A3 LUN 0 Selected LUN Information A4 LUN 0 LUN Number 0 BO LUN 1 Bind Type RAID 5 B2 LUN 1 Number of Disks 5 B1 LUN 1 Owner Controller A Peer B3 Status OK B4 G Selected Disk Information Status Bound but Unassigne
119. second LUN for example newfs dev rdsk cb00dfllis0 DGC C2300WDR1 1 9GB Creating a File System on a Series 700 System 8 103 Bunoouse qno4 8 104 Creating a File System on a Series 700 System REMOVAL AND REPLACEMENT Illustrated Parts Breakdown for the Model 10 Illustrated Parts Breakdown for the Model 20 Model 10 20 Disk Module Removal Replacement Model 10 20 Fan Pack Removal Replacement Model 10 20 SP Removal Replacement Model 10 20 Power Supply Unit Removal Replacement Model 10 20 BBU Removal Replacement Model 10 20 SIMM Removal Replacement Model 10 Deskside Component Removal Model 10 Rackmount Component Removal Model 20 Deskside Component Removal Model 20 Rackmount Component Removal 9 1 jueuie2e deaH pue jeAoway Illustrated Parts Breakdown for the Model 10 Figure 9 1 Parts Breakdown for the Model 10 Array Illustrated Parts Breakdown for the Model 10 9 2 Table 9 1 Model 10 Disk Array CE Replaceable Parts see Figure 9 1 Figure New HP Exchange index part number part number Description 1 005 41327 A3231 69007 Model 10 midplane 2 005 039649 N A LED cable 3 N A N A External chassis 4 005 041006 N A Fan cable 5 N A N A Internal AC harness bracket 6 005 041698 A3231 69008 Interconnect PCB IPCB 7 005 041068 N A Internal synchronous SE SCSI 2 cable 8 113 000552 N A On off switch 9 005 041008
120. select OK to implement them and close the window To abort and close the window without altering current settings select Cancel instead Set Up TTY Connection This action applies to model 10 and model 20 arrays and leads to a dialog window where the serial link for SAM to use while managing the disk arrays is specified An example dialog window looks like the following SAM gt Disks and File Systems gt Disk Devices Actions Disk Array Maintenance gt TTY Management Use this dialog window to configure the serial link between the host system and the disk array SAM uses this serial connection to communicate with the disk array s SP The connection is not used for normal data I O transfers between the host system and the disk array However by configuring one serial connection to each SP on the array SAM can help to perform the following operations from the Actions pulldown menu e Bind LUNs e Unbind LUNs e Change the caching parameters of specific LUNs e View the status of each disk module and LUN in the array At the top of the window is a list of hardware paths identified by SAM as RS 232 ports with the arrays connected to them To add anew TTY to the list or modify an existing one use the text boxes to define the new values for the Disk Array Hardware Path SCSI interface and the hardware path for the RS 232 interface plus the MUX port number To select from recognized valid values use the push
121. select firmware file to be used Use List commands to display current information Use Update commands to start update controller disk a Controller updates should have peer SP if any enabled b Disk updates must have peer SP if any disabled RIILT M B Figure 7 8 Firmware Update Tool Top Level Menu Table 7 11 Firmware Update Tool Menu Items and Descriptions Menu item Option Property Description File Select File filename Specifies the microcode or firmware file to be used Save As filename Saves a file using a user specified name Print N A Prints a file Exit N A Exits the Firmware Update Tool Support Tools Manager STM 7 79 sonsoubeiq pue sali ISOH Table 7 11 Firmware Update Tool Menu Items and Descriptions cont d Listing Microcode and Firmware Revisions and restart controllers Menu item Option Property Description List Current N A Lists SP microcode and disk module firmware Revisions revisions currently in use Figure 7 9 File Selection N A Reports the file previously specified using the Select File option on the File menu Instructions N A Lists basic instructions for Firmware Update Tool use Update Controller Download Downloads microcode to the SCSI disk array s Firmware update PROM database drives burns PROM code contained in the download file into the SP s PROMs and restarts the SP using the new microcode
122. specified Grid Manager does not allow more memory than either SP has to be specified If less than this amount is specified the remaining SP memory is unused Valid cache sizes are 0 8 16 32 and 64 Mbytes 64 Mbytes available in the model 20 only A cache size of 0 prevents the disk array from enabling caching on power up 6 58 Setting up Disk Array Caching for Microcode Revision 7 28 Setting up Disk Array Caching for Microcode Revision 8 xx 9 xx The disk array supports caching if it has the following modules Two SPs each with at least 8 Mbytes of memory A fully charged BBU Disk modules in slots A0 A1 A2 A3 and A4 model 10 or AO BO C0 DO and EO model 20 Follow the procedures in this section when Setting up a new disk array that supports caching Set up disk array caching before making newly bound logical disk units accessible to the operating system With disk array caching set up a logical disk unit with either the read or write cache enabled can use caching as soon as the unit is accessible to the operating system Upgrading the system to support caching Set up caching after installing the prerequisite hardware To use caching an array must have two SPs Also the amount of cache on each SP must be mirrored equal amounts of cache on each SP A minimum of 8 Mbytes of SIMM memory on each SP is required for caching The following table shows the combinations of SIMMs that can be used for caching Orders
123. terminal Figure 4 1 Model 10 20 Basic Configuration Basic Configuration for Model 10 and Model 20 Disk Arrays 4 5 009 seues 0006 dH uoneunByuo25 Table 4 2 Basic Configuration Error Recovery Failing Continue after component failure What happens and how to recover Disk module Yes With RAID of any level other than 0 or an individual disk applications continue running System Administrator or service provider replaces module SP No Disk array fails System Administrator or service provider replaces SP and the System Administrator restarts operating System Fan pack Yes Applications continue running System Administrator or service provider replaces module Power supply Yes If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply The System Administrator restarts the operating system if necessary SCSI 2 adapter No I O operations to array disk units fail Authorized service provider replaces adapter and System Administrator restarts operating system and applications SCSI 2 cable No I O operations to array disk units fail Authorized service provider replaces cable and System Administrator restarts operating system and applications 4 6 Basic Configuration for Model 10 and Model 20 Disk Arrays Two SP Configuration The two SP configuration shown in Figure 4 2 has on
124. the Disk Array Model 10 disk array Model 20 disk array Figure 1 1 Model 10 and Model 20 Disk Arrays Disk Array Chassis Model 10 The model 10 disk array chassis Figure 1 2 contains the compartments for disk modules SPs power supply units and BBU The fan pack is attached to the rear of the chassis The About the Disk Array 1 3 uonduioseg jonpold disk modules and BBU face front The SP s power supply units and fan pack are accessible from the back Each disk module has a slot ID that corresponds to an internal synchronous SE SCSI 2 bus The model 10 has two internal synchronous SE SCSI 2 buses identified by letters A and B Five disk modules can be installed on each bus Bus slot IDs and the order in which disk modules should be installed into the disk array are e Bus A disk modules AO Al A2 A3 A4 e Bus B disk modules BO B1 B2 B3 B4 AO z A1 Deskside Model Front View gt A2 Internal ES As jzs Bus A A4 Database Drives AO A1 A3 BO BO B1 Internal Cache Vault Drives E B BusB AO A1 A2 A3 Ad B3 E B o o o o g Internal Bus B Internal Bus A QULA Rackmount Model Front View A4 BO B1 B2 B3 Database Drives OO
125. the SP that lost default ownership of the LUN s This forces the SP to reboot and reloads LUN configuration information After rebooting check to see if the SP has regained default ownership of its LUN s If default ownership has not been re established use Grid Manager to reassign default ownership of the SP s LUNs to the other SP by following the menu path listed above Changing default ownership causes the SP s to reboot again After rebooting check to see if one SP is the default owner of all LUNs on the disk array Next use Grid Manager to reassign default ownership of the LUN s back to the original SP After rebooting check to see if the original SP has regained ownership of its LUNs If it has not unmount all filesets on all disk arrays connected to the same host adapter and power cycle the disk array to reboot both SPs Default ownership should be regained Then remount all filesets and enable cache If default ownership has not been re established the replacement SP is suspect and should be replaced Troubleshooting with Grid Manager 8 89 punoouse qnouj Troubleshooting with SAM SAM is the System Administration Manager program It is a self guided menu driven tool that can be used to manage system s without requiring specialized knowledge of operating system internals SAM is used to perform tasks such as user and account management disk and file system management peripheral devices management system backup and
126. the disk array must be running the same revision of microcode SP microcode is always loaded from disk when the SP boots or when the array is powered on The current at time of publication microcode revisions and corresponding PROM codes are 6 34 Presentation Utility Screen Revision 7 16 PROM 7 97 4 Gbyte HH half height disk support for the SP on model 10 and model 20 only Revision 7 23 PROM 7 97 2 Gbyte disk support for the SP only Revision 8 20 PROM 7 99 for the SP only Revision 8 27 PROM 8 02 4 Gbyte LP low profile and 9 Gbyte HH disk support for the SP on model 10 and model 20 only Revision 8 51 PROM 1 49 4 Gbyte HH disk support for the SP620 SCSI SP only Revision 8 56 PROM 1 56 for the SP620 SCSI SP only Revision 8 58 PROM 1 57 4 Gbyte LP and 9 Gbyte HH disk support for the SP620 SCSI SP only Revision 9 26 07 PROM 8 02 4 Gbyte LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP on model 20 only Revision 9 56 07 PROM 1 73 4 Gbyte LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP620 SCSI SP only Statistics Logging Gives the state of statistics logging either disabled or enabled Statistics logging can be enabled or disabled with Grid Manager Peer SP Indicates the presence or absence of another SP Possible values are PRESENT or REMOVED which means not present Requests Completed Gives the number of host SCSI 2 adapter I O requests to the SP since t
127. the side Model 10 Deskside Component Removal 9 29 jueuie2e deH pue jeAoway Side cover Side cover Figure 9 18 Removing the Outer Covers Model 10 Deskside Component Removal 9 30 Midplane Figure 9 19 and Figure 9 20 CAUTION For these procedures the disk array must be powered off and unplugged It also is recommended that ESD procedures be followed Use ESD kit P N 5182 4119 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Remove the outer covers Figure 9 18 3 Remove the following a b SPs or fillers Figure 9 9 f Fan pack Figure 9 6 Power supplies or fillers Figure 9 11 Interconnect PCB Figure 9 21 All disk modules or fillers Figure 9 3 Be sure to label each disk module with the compartment number it was removed from BBU if present Figure 9 13 4 Turn the disk array over onto its top Model 10 Deskside Component Removal 9 31 jueuie2e deH pue jeAoway WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array a Remove the wheel base i Remove the retaining screws from the front of the wheel base ii Slide wheel base 1 inch toward back and remove 5 Remove the inner metal cover 11 screws Figure 9 19 6 Remove the inner chassis a Remove the six retaining screws from the top Figure 9 20 b L
128. typing the new profile name GRIDMGR into the Profile name field Type the word terminal into the Type field and press the ADD PROFILE function key NMMGR 3000 responds by displaying a standard terminal profile screen Configuring RS 232 Ports on HP 3000 Series 900 Systems 6 7 Jobeue N pus which must be modified to create the new Grid Manager profile Figure 6 2 shows the NMMGR 3000 terminal profile screen for the GridMgr iX RS 232 ports Terminal type 10 10 18 or 24 or terminal type file name Line speed 9600 300 1200 2400 4800 9600 or 19200 bps Record width 80 1 2048 bytes Modem type 0 0 none 1 US 2 European Modem behavior 0 0 none 1 DCE 2 DTE 3 DCE high Parity NONE None Even Odd 0 s or 1 s Speed Parity Sensing N Y N Allow HELLO logon N Y N Reset HP terminal N Y N NLIO device N Y N Hardware handshaking N Y N If configuring device classes enter names below If additional names are required press Go To CLASSES otherwise press Prior Screen Device class GRIDMGR names I I i File NMCONFIG PUB SYS Figure 6 2 NMMGR 3000 Terminal Profile for Disk Array Grid Manager RS 232 Ports 6 Move the cursor to the Allow HELLO logon field and change the contents from Y to N Then move the cursor to the first Device class names field and type the new device class name GRIDMGR 7 Pressthe SAvE DATA function key to
129. units for a model 20 One BBU optional Up to 16 Mbytes of storage processor cache for the model 10 up to 64 Mbytes for the model 20 Disk Drive Modules A disk drive module also called a disk module or disk drive consists of a single ended SE SCSI 2 disk drive also called a disk mechanism a power regulator board internal cabling and a plastic carrier The carrier has a handle for inserting and removing the module from the disk array chassis The disk module is customer replaceable and hot replaceable which means it can be replaced with the disk array powered on No components within disk modules can be replaced and there are no jumpers to set Three status LEDs on the disk modules indicate the following Ready green Lights while the disk module is powered up and ready for use Busy green Lights while the drive is in use for example during formatting or user I O operations Fault amber Lights when the SP shuts down a disk module that fails and briefly after replacement while the disk mechanism spins up to speed About Disk Array Components 1 9 uonduioseg jonpold A label attached to the carrier s side shows the disk module s HP part number and capacity A disk module s capacity can also be determined using the Grid Manager or HARRAY diagnostic utility Another label specifying the slot into which the disk was bound at the factory is also attached to the disk module Bound drives must remain in the slots
130. want to quit using Grid Manager type either a Control Y or a Control K at any of the w B Grid Manager Enter Choice prompts for input Kk kk eee Se kk kk ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck ck sk sk sk sk sk sk kx kx kx Type a Carriage Return when you are ready to proceed gt gt gt gt ERROR FOPEN NIKEPORT amp 0484 0144 4096 106 amp failed with CCL DEVICE UNAVAILABLE FSERR 55 Figure 6 6 Access Denied Ldev Already in Use by Another Session Initiating Grid Manager on MPEAX 6 15 Jobeue N pus In the unlikely event that any other error occurs GridMgr iX displays an error message and then aborts operation The error message wording is not useful to the system user but should be reported to the HP service representative who will help debug the problem 6 16 Initiating Grid Manager on MPE AX Displaying the Grid Manager Main Menu When using Grid Manager the first time after powering up the disk array the Main menu appears Otherwise the screen that appears is the one displayed the last time Grid Manager was exited If the Grid Manager Main menu does not appear or to return to the Main menu from another menu level type caret and press RETURN until it does xterm P esI ID 15 GridHgr Hain Henu S N 9621954 1 Presentation Utility 2 Bind Logical Units 3 Unbind Logical Units 4 View Haintenance Information 5 Change Parameters
131. when internal parts of the array must be accessed 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 All of the CRUs must be removed because the disk array is too heavy to lift out of the rack if they are not removed Remove the following a b f Fan pack Figure 9 6 SP Figure 9 9 Power supplies Figure 9 11 BBU located under the power supply units Figure 9 13 All disk modules Figure 9 3 Be sure to label each disk module with the compartment number it was removed from Front panel Grasp the cover by its sides and gently pull off 4 Remove the four retaining screws holding the chassis to the rack 9 88 Model 20 Rackmount Component Removal WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array 5 Remove the empty chassis from the rack Model 20 Rackmount Component Removal 9 89 u w sejd y pue jenow y 000000000000 Fan pack Distribution panel les 0000000000000000000000 o0000000000000 T Array chassis Torx 25 retaining screws 2 of 4 shown EEEN Figure 9 52 Removing the Model 20 Array from the Rack 9 90 Model 20 Rackmount Component Removal Midplane Figure 9 53 Figure 9 54 and Figure 9 55 1 Power off the disk array and disconnect all cables
132. with SIMMs for the SP620 SCSI controller For caching each SP620 SCSI SP must have a minimum of 8 Mbytes total capacity two SIMMs each x 4 Mbytes The SP controls the disk modules in the disk array through internal synchronous single ended SE SCSI 2 buses A model 10 SP has two internal SE SCSI 2 buses each supporting five disk modules A model 20 SP has five internal SE SCSI 2 buses each supporting four disk modules Either SP can control any physical disk unit or logical unit number LUN within the disk array but only one SP can control a LUN at a time When using HP UX 10 x or later and the host system is configured with a primary path and an alternate path using LVM Logical Volume Manager if one SP cannot access the physical disk units it controls because of a failure the other SP can automatically take control of them This feature is called auto trespass It is supported on HP UX 10 x or later SP microcode is stored on the disk array s database drives and is loaded from one of the database drives onto the SP when an SP is enabled or when the disk array is powered up On a model 10 the disk modules in slots AO A1 A3 or BO are the database drives On a model 20 the disk modules in slots AO BO CO or A3 are the database drives that contain the SP microcode An enable disable switch is located on the SP bulkhead The SP switch must be disabled when a failed SP is removed and replaced This switch does not exist on the SP620
133. x and Later Multi Bus Initiator Configuration HP UX 10 x and Later The multi bus initiator configuration shown in Figure 4 6 provides the highest availability Each host has two SCSI 2 adapters each of which connects by a separate external Fast Wide differential SCSI 2 bus to a separate SP in the array Since this configuration protects against a SCSI 2 bus or cable failure it provides higher availability than the multi initiator configuration It is for operations requiring the highest level of availability For better performance with this configuration some physical disk units can be bound on one SP the rest can be bound on the other SP The SP that binds a physical disk unit is the default owner The path through the SP that owns a physical disk unit is the primary path and the path through the other SP is the alternate path The alternate path is available if a component in the primary path fails and the host has been configured with primary and alternate paths SCSI 2 bus SCSI 2 bus Host computer 1 Disk array 1 Disk array 2 Host computer 2 SCSI 2 SPA SPA SCSI 2 adapter1 adapter1 SCSI 2 E SCSI 2 adapter2 ll spp l sPB adapter2 I i J a J System TE System disk pem System console console Figure 4 6 Model 10 20 Multi Bus Multi Initiator
134. 00 Systems The MPE iX operating system performs I O operations to Ldevs Each DTC RS 232 I O port must have an Ldev number NMMGR 3000 is used to assign Ldev numbers to the DTC Data Communications and Terminal Controller RS 232 ports Ldev numbers assigned to the disk array s RS 232 ports are separate from Ldev numbers used for the disk array s SCSI I O ports When the HP 3000 Series 900 accesses the disk array to perform disk I O it uses the Ldev number corresponding to the disk array s SCSI I O path as configured in SYSGEN When the HP 3000 Series 900 accesses a disk array through GridMgr X it uses the Ldev number corresponding to the DTC RS 232 port as configured in NMMGR 3000 Configuring the DTC RS 232 ports used to access the disk array is identical to configuring normal user terminals except that a terminal profile slightly different from the typical TR10D96 terminal profile is used Refer to the HP 3000 Series 900 Configuring Systems for Terminals Printers and Other Serial Devices part number 32022 61000 manual for information about configuring asynchronous terminal devices attached to DTCs for HP 3000 Series 900 systems To Create a Terminal Profile 1 Run the NMMGR 3000 utility 2 Open the NMCONFIG PUB SYS configuration file 3 Press the DTS function key to enter the DTS subsystem configurator The Host Configuration screen displays 4 Press the Go To PROFILES function key 5 Create anew terminal profile by
135. 00 cece eee ee R 9 Power Consumption eeeeeeeeeee he R 10 DC Power Specifications lsseeeeeeeeeeeee eh R 10 Environmental Specifications 00 cece eee ee eens R 10 Pneu R 12 Safety Certifications v 25 ves us ee uere VER Be Tere Eg R 12 Replacement Power Cord Part Numbers by Country R 12 EMC Compliance cse eea o ea e a p eee ene n eee ence e Ee R 13 Replaceable Parts Model 10 0 cece eee eee een eens R 14 Replaceable Parts Model 20 00 cece cece eect een eens R17 Hewlett Packard Software License Terms 00 c eee e eee eens R 20 iii U9 U09 Regulatory Statements i r esep seu pron e eee ete eee ne R 24 VCCI Class 1 For Japan Only 0 eee eee ce eee eens R 24 1 Product Description Aboutihe Disk Array i fads ee ped eh bee eee See ed sl eda PETS ned 1 2 Disk Array Chassis Model 10 0 00 cece cece eee eee eee 1 3 Disk Array Chassis Model 20 0 c eee eee eee eee ene eee 1 5 About Disk Array Components 2 0 cece eee ene n ences 1 9 Disk Drive Modules 00 0 cece eet ee een eee eens 1 9 Storage Control Processor l l 1 10 Fanm Backe S025 bts eee eed udrede e eae Eure EP itt 1 13 Power Supply Units eege ere ace oak Be Ree aoe ig Bee oes 1 14 Battery Backup Unit 0 0 00 eee cece eee re 1 15 SGSE2 Adapter seo try een LEE UE eI Pe ERG EW Dee rs 1 17 Disk Array H
136. 000000000000000 2040 02 04 EO 0x602 Drive Enabled 0x0000000000000000 2039 03 DO 0x602 Drive Enabled 0x0000000000000000 2038 02 04 E0 0x906 Unit Shutdown 0x0000000000000000 2037 03 DO 0x906 Unit Shutdown 0x0000000000000000 Enter selection F orward Backward Quit Note Each SP has its own Unsolicited Event Log Be sure to access the appropriate SP or both SPs when examining these logs NIKEARRY 7 63 sonsouDeig pue seni ISOH UPDATE UPDATE will search for the appropriate devices and download controller microcode or disk firmware to these devices automatically without having to specify the device by index number Example NIKEARRY gt update KKKKKKKKKKK KK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KK KKK WARNING 7 Hittin NIKEARRY must have exclusive access to all NIKEs you wish to update m If you are in a multihost environment such as Switchover amp ServiceGuard make sure all other hosts are powered down before continuing id KKKKKKKKKK ke kk e e Sk e e ke e e Sk Sk S S kk kk ke Sk kk kk kk kk kk kk ke ke Sk Sk Sk Sk kk Sk kk Sk Sk kk ck ck ck ck kv Press Return to continue kkkkxkxkxkxkxkkkkkkkkkkkkkkkkxkxkxkkkkkkkkkxkxkxkkkkkkkkkkxxkx k FIRMWARE FILES REVISED 01APR98 ett eee eee ee ee ee ee ee ee ee ee ck oko ote cee eee eee ee ee ee ee ee ee k k k ee ee k k k ee k k k ee k k k k k k k NIKE Firmware Files Found ck ck ck ck ck ck ck ck ck c
137. 02 Memory exhausted An attempt to allocate a memory resource that is required always to be available failed 0x00000007 SCSI interrupt phase mismatch Caused by a hardware malfunction 0x00000008 Invalid trap The microprocessor executed an invalid trap procedure for example illegal instruction detected 8 76 Unsolicited Event Log Table 8 17 Microcode Panic Codes Not a Complete List cont d Value 0x00000008 Description Detection of power line disturbance or software problem In earlier microcode revisions these panics resulted from detecting power line disturbances A protection mechanism ensures the integrity of data transfers within the array If power is determined to be unstable or falls outside tolerance levels the array initiates a power up sequence and continues processing In later microcode revisions a software problem was discovered that violates the hardware s internal protection system This should not be considered a hardware fault The SP should not be replaced The following variations of panic 0x00000008 with the extended status of 3 have also been fixed An uninitialized pointer during a synch command Aborting a command that has originally generated a retry to a disk that has to be delayed because of the abort The resources for the command were freed so that when the retry finished it had no available resources to continue Anuninitialized pointer within the microcode used durin
138. 1 and 9 xx 0x00E0000C Incorrect state reported by peer Reference the 0x0080007E code 8 84 Unsolicited Event Log Table 8 17 Microcode Panic Codes Not a Complete List cont d Value Description OxOOF20009 Uninitialized NovRAM The contents of the nonvolatile RAM on the SP indicate that the NovRAM does not contain valid information The SP automatically initializes the NovRAM to a valid state after this error is reported Unsolicited Event Log 8 85 Bunooyseajqnoly Troubleshooting with Grid Manager Checking Status with the Presentation Utility To check array status start the Grid Manager utility to display the Presentation Utility screen If the Presentation Utility screen does not appear type caret until the Grid Manager Main menu appears then choose the Presentation Utility option A typical Presentation Utility screen for the model 20 deskside disk array is shown in Figure 8 23 The number and arrangement of disk modules displayed on the screen varies for the model 10 and model 20 and deskside and rackmount disk arrays e xterm i x I Hicrocode Revision IRDYIRDYIEMAL 21 1 IRI I 09 55 01 l 1 ol I 0 Statistics Logging DISABLED o 4 Peer SP PRESENT Requests Completed 71 Cache States Write DISABLED Read DISABLED Enter for HELP to GO TO HAIN HENU Figure 8 23 Typical Model 20 Deskside Present
139. 1 0 group requires a maximum of three I O operations per mirrored image If high throughput for a specific physical disk is desired use a RAID 0 group This configuration requires at least three disk modules An individual disk unit needs only one I O operation per read or write operation Two SPs Load Balancing and Device Names If the disk array has two SPs it is possible to choose which disks to bind on each SP This allows the load on the SCSI 2 adapters SPs and internal SCSI 2 buses to be balanced The SP on which a physical disk unit is bound is the default owner of that physical disk unit and determines the device name of the disk unit The path through that SP is the primary path and the path through the other SP is the alternate path In the more complex configurations it is important that a target disk not normally be accessed through more than one SP This causes continuous changes in disk ownership which causes a significant performance degradation In HP UX 10 x the second SP should be used as an alternate path in the event of an SP failure Completing the Configuration Worksheet Choosing and planning appropriate disk configurations is very important The configuration worksheet Figure 1 21 and Figure 1 22 is a useful tool for planning hardware configuration LUNs RAID levels and LVM structure 1 42 Planning Disk Unit Configurations Host Name Array Name Host Adapter Path
140. 10 Deskside Cabinet 0 0 0 0 0 ccc ccc eee en 8 3 Model 20 Deskside Cabinet ccc cc eee en 8 4 Model 10 Rackmount Chassis eee en 8 5 Model 20 Rackmount Chassis leen 8 6 Front View Disk Array LEDs sseeeeeeeee IIIA 8 7 Rear View Disk Array LEDs 0 0c cece eee eect teens 8 8 Power Up Troubleshooting 0 cece eee eee ene eens 8 10 Power Up Sequence for the SP 2 0 eee eee ee teens 8 10 Power Up Sequence for the SP620 SCSI SP 0 0 cee eee eee 8 12 xii Troubleshooting the Power Up Sequence 02 02 ee ee eee eee 8 16 Troubleshooting the Disk Array seeeeseeeeee eee 8 18 General Troubleshooting Information 00 0 eee eee eee 8 18 Disk Troubleshooting eeeeeeeeeeee eee eee ences 8 20 Disk Troubleshooting Overview 2 e eee ee eee ee eee nee 8 20 Identifying Logical Disk Units LUNs and Disk Module Failures 8 21 Obtaining Additional Disk Failure Information with Grid Manager 8 24 Hot Spare versus Rebuild 2 0 cece eee eee eee nee 8 24 SP Troubleshooting 0 2 0 0 0 cece ccc eee eee e eee n eens 8 27 SP Failures on HP UX 9 0K 0 eee eee ence 8 27 Changing Paths and Ownership if an SP Fails HP UX 9 04 8 27 OP Heat Sinks 562i elite ov Sede oe ee eee DE CUPS 8 30 SIMM Troubleshooting for the SP Only 0 2 0 e eee ee
141. 12 PM The Random Media Test is 50 complete 4 12 PM The Random Media Test is 75 complete 4 12 PM The Random Media Test is 100 complete RANDOM MEDIA TEST UTILITY COMPLETED 4 12 PM 7 40 HARRAY Section 17 External Exerciser RUL RETRIEVE UNSOLICITED LOGS Non exclusive non destructive Displays events that have occurred on a FRU The logs are in chronological order the last log displayed is the most recent event logged against the device Example HARRAY gt RUL RETREIVE UNSOLICITED LOGS RUL UTILITY 4 13 PM The Unsolicited Logs are empty at this time RETREIVE UNSOLICITED LOGS RUL UTILITY COMPLETED 4 13 PM HARRAY Section 17 External Exerciser 7 41 sonsouDeig pue seni ISOH SHOW MECHS Non exclusive non destructive Lists configuration information for the system under test Example HARRAY gt SHOW MECHS Firmware Revision 08 20 00 Prom Revision Level 7 99 0 Physical Device accessed 24 14 0 Current SP accessed A SCSI Back end Address 14 Peer SP is Present Hit CR to continue or E to Exit A B ENA ENA ENA ENA ENA ENA ENA ENA ENA UNB qo wq AO 4 CX HE GE ZE 3223 0 1 2 3 4 0 1 2 3 4 7 42 HARRAY Section 17 External Exerciser FAN FAN VSC VSC BBU Ug mU OK OK OK OK OK SUSPEND Non exclusive
142. 20 17227025 512 8 820 2 DGC C2300WDR5 69665280 512 35 669 69665279 512 35 669 3 DGC C2300WDR5 16411392 512 98 403 16411391 512 8 403 4 DGC C2300WDR1 4073139 512 2 085 4073138 512 2 085 5 DGC C2300WDR10 16411392 512 8 403 16411391 512 8 403 User Information Disk Information Dsk VendorID Product ID Cap lt blks gt BSz GBytes Last Addr BSz GBytes A IBM DGHSOSY CLAROD 17227044 512 8 820 17429579 520 9 063 SII EIRP Figure 7 35 Info Capacity Screen Info Inquiry hpterm m xl Expert Tool Inquiry for device C2300 Array on path 8 4 9 W salt rose hp com 15 43 214 34 Expert Tool Inquiry for device C2300 Array on path 8 4 9 Array Date 04 29 98 SN 9621954 System Date 04 29 98 Array Time 11 05 01 FH 9 25 00 8 02 00 System Time 11 14 07 Array SP A 9 0 MN 7624 SP System Namei salt Lun VendorID ProductID PRvL 0 DGC C2300WDR5 HP02 1 DGC C2300WDR1 HPO2 2 DGC C2300WDR5 HP02 3 DGC C2300WDR5 HP02 4 DGC C2300WDR1 HPO2 5 DGC C2300WDR10 HPO2 Tsk YendorID ProductID PRvL RO IBM DGHSOSY CLAROS HP1A BO IBM DGHSOSY CLAROS HP1A CO IBM DGHSOSY CLAROS HP1A SII Figure 7 36 Info Inquiry Screen Support Tools Manager STM sonsouBeiq pue seni ISOH T O Read Disks The following steps are required to view disk information 1 Select I O Read Disks from the Expert Tools main menu Select disks from the dialog box that appears as shown in Figure 7 37 2 Selec
143. 2B Bad interrupt An unexpected or incorrect interrupt was received from a SCSI protocol chip or drive 0x2C Too much data Too many bytes were sent or requested by a drive during an I O operation Ox2D Bad bus phase A drive violated the SCSI bus protocol Ox2E Bad message A drive violated the SCSI bus protocol Ox2F Bad reselection A drive violated the SCSI bus protocol 0x30 Parity error A SCSI bus parity error was detected 0x31 Message parity error A SCSI bus parity error was detected 0x32 Status parity error A SCSI bus parity error was detected 0x33 Bad status A drive violated the SCSI bus protocol 0x34 Invalid CDB An invalid command was issued This error code should never be seen 0x35 Unsupported function This error code should never be seen 0x36 No device driver This error code should never be seen 0x37 Unsupported IOCTL option This error code should never be seen 0x38 Unexpected disconnect A drive violated the SCSI bus protocol 0x39 User reset The device was reset This may be caused by a previous error Ox3A Protocol time out A drive violated the SCSI bus protocol Ox3B Recovered error cannot remap The device reported a Recovered error that required ECC to recover the data 01 18 but it did not report the bad LBA in the sense data The microcode was unable to attempt a remap Unsolicited Event Log 8 73 Buljooysajqnoly Table 8 15 Extended Status Codes cont d Extended status code
144. 32768 32768 1 Note Last block to verify must be gt first block to verify Figure 7 45 Tests Verify Disks Select Blocks Screen Support Tools Manager STM 7 121 sonsouBeiq pue saninn ISOH Tests Verify LUNs 1 Select Test Verify LUNs from the Expert Tools main menu A dialogue box to select a type of verification appears Figure 7 46 2 Select the desired action and click Yes The Select Blocks dialogue box appears Figure 7 47 3 Define the appropriate block parameters and click OK When the system completes the verify LUNs process the Expert Tools main window displays the information e hpterm Ot xi Expert Tool Verify Luns for device C2300 Array on path 874 9 This command will Verify via the SCSI Yerify command Luns on Array Serial 9621954 Select desired action dp Verify ALL luns ALL blocks as listed in next selection gt Verify SELECTED luns ALL blocks Lun_O Blocks thru 68908031 Lun_i Blocks thru 17227025 Lun 2 Blocks thru 69665279 Lun_3 Blocks thru 16411391 Lun_4 Blocks thru 4073138 Lun_5 Blocks thru 16411391 X Verify SELECTED luns SPECIFIED blocks Lun O Blocks lt you will be asked Lun 1 Blocks lt you will be asked gt Lun 2 Blocks lt you will be asked Lun 3 Blocks lt you will be asked Lun_4 Blocks lt you will be asked Lun 5 Blocks lt you will be asked gt v more w
145. 42H cables are used to connect the host system to the disk array 3 Power up and reboot the Series 700 system The port s should be configured as a terminal hardwired and their baud rate should be 9600 To prevent these ports from hanging the getty process should not be restarted with respawn To determine the required MUX and port information 1 Type ioscan f to identify the lu number of the MUX Example ioscan output for a MUX in an 827S system Connect Asynchronous Console Cables 2 37 uonejjeisuj Class LU H W path Driver H W Status S W Status tty 0 56 mux2 ok 0xd00 ok Example ioscan output for a port on a 735 system tty 0 2 0 4 asio0 CLAIMED INTERFACE Built in RS 232C tty 1 2 0 5 asio0 CLAIMED INTERFACE Built in RS 232C Type Issf dev tty0 for lu 0 Example output showing terminal ports for lu 0 mux2 lu 0 port 0 hardwired address 56 dev tty0pO mux2 lu 0 port 1 hardwired address 56 dev tty0pl mux2 lu 0 port 2 hardwired address 56 dev tty0p2 This output shows that ports 0 through 3 are configured as terminal hardwired ports If a port has been configured as a modem port the special file would be identified for example as ttyd0p6 for lu 0 port 6 Example output showing modem ports for lu 0 mux2 lu 0 port 4 callin address 56 dev tty0p6 Two examples of Issf commands and output for a 700 system lssf dev ttyd0 asio0 card instance 0 hardwired at address 2 0 4 dev tty0p0
146. 6 View Cache Statistics Enter or lt number gt for HELP Figure 6 7 Grid Manager Main Menu Certain Grid Manager operations are restricted and can be performed in FE Field Engineer mode only While in FE mode the Grid Manager Main menu displays FE in the top right corner and contains the additional menu selection 7 Back End Exerciser In general FE mode can be entered and exited from screens that display menu options such as the Grid Manager Main Menu screen but not from screens that display disk array data such as the View Unsolicited Event Log screen To access FE mode from a screen that Displaying the Grid Manager Main Menu 6 17 Jobeue N pus allows it type CONTROL P followed by in upper case FE To return to user mode from a similar screen type CONTROL P followed by in upper case USER Grid Manager has a help feature that is accessed by typing or n where n is the number of a menu choice on the current menu 6 18 Displaying the Grid Manager Main Menu Grid Manager Menus The Grid Manager menus and screens are organized as multiple levels in a hierarchical fashion At the top is the Grid Manager Main Menu All menus are numbered To move to a lower level menu type the menu s selection number To return to a higher level type caret followed by RETURN until the desired level is reached Following are the menu hierarchies for the different versions of the controller microcode For ease of reading a
147. 6 and Figure 9 37 Lssuus 9 63 On Off Switch Figure 9 38 00 cee cece eee eh 9 66 AC Line Filter Figure 9 39 0 0 ee eee eee ene eee 9 68 Model 20 Deskside Component Removal eese 9 70 Outer Covers Figure 9 40 sseeeeeeeeeeeee eh 9 70 Midplane Figure 9 41 Figure 9 42 and Figure 9 43 9 72 RIP PCA Figure 9 44 1 0 0 cece eh 9 75 LEDs LED Cable Figure 9 45 0 cece cece eee eee eee 9 76 DC Power Distribution Cable Figure 9 46 0 0 2 e eee eee 9 78 U9 U09 Fan Cable Figure 9 47 feccircncan ett ad tt alan a aa iau 9 80 Address and Internal SCSI 2 Cables Figure 9 48 04 9 82 Internal AC Harness Figure 9 49 and Figure 9 50 9 84 On Off Switch Figure 9 51 20 0 0 2c cee eh 9 87 Model 20 Rackmount Component Removal 00 02 c eee eens 9 88 Remove the Array from the Rack Figure 9 52 02 005 9 88 Midplane Figure 9 53 Figure 9 54 and Figure 9 55 4 9 91 RIP PCA Figure 9 56 sseseesseeelellel I 9 94 LEDs LED Cable Figure 9 57 0 cece eee cece eee eens 9 96 DC Power Distribution Cable Figure 9 58 0 2 2 2 e ee eee 9 98 Fan Cable Figure 9 59 0 cece ccc eee eh 9 100 Address and Internal SCSI 2 Cables Figure 9 60 6 9 101 Internal AC Harness Figure 9 61 and Figur
148. 67 HARRAY command 7 14 in STM 7 85 7 114 Storage System Options Package 6 65 B basic configuration HP 3000 Series 900 5 5 HP 9000 Series 700 3 6 HP 9000 Series 800 4 5 BBU battery backup unit about the 1 15 battery disposal R 11 removal replacement 9 21 status on Presentation screen 6 39 test time in STM 7 85 7 114 weight R 9 BDP on Presentation screen SP B s dirty pages 6 40 BIND Nikearry command 7 50 binding a LUN using SAM 8 96 LUNs 6 43 LUNs in STM 7 85 7 114 physical disk using SAM 7 127 block stripe size 1 24 INDEX block stripes 1 24 block striping 1 25 1 27 bound but not assigned 8 87 bound but unassigned 8 68 bus slot ID model 10 1 4 model 20 1 5 C cache changing parameters using SAM 8 94 configuration 8 38 disabling in STM 7 85 7 114 enabling in STM 7 85 7 114 specifying sizes 6 56 state on Presentation screen 6 36 cache vault drives 1 48 6 45 caching 1 11 1 19 enabling disabling 6 54 6 62 microcode rev 7 23 6 53 microcode rev 8 20 6 59 CAPACITY HARRAY command 7 15 changing SP parameters 6 65 chassis model 10 1 8 model 20 1 5 checking status 8 86 CLEAR LOGS HARRAY command 7 16 compatibility with host systems R 2 configuration setting up Hardware Event Monitor 7 132 configuration management worksheet bound BIN 6 44 xepu completing 1 42 unbound UNB 6 44 configuring MUX ports 6 3 RS 232 ports on HP 3000 6 7 SCSI LUNs for MPE AX 5 13 connect asynchronou
149. 8 1 26 mirroring 1 24 model 10 deskside component removal 9 29 rackmount component removal 9 50 model 20 deskside component removal 9 70 rackmount component removal 9 88 models and options R 5 MODESENSE Nikearry command 7 58 moving the Presentation screen grid cursor 6 41 MPE X 2 3 5 2 mstm 7 68 mstm interface of STM 7 99 7 126 multi bus configuration HP 3000 Series 900 5 11 HP 9000 Series 700 3 10 HP 9000 Series 800 4 11 multi bus initiator configuration HP 9000 Series 800 4 15 multi initiator configuration HP 9000 Series 800 4 13 MUX install on a Series 800 2 36 N NIKEARRY 7 46 FRU mapping 7 55 functional specifications 7 46 operational commands 7 47 quick start instruction 7 48 Non Vol RAM Error Log display 8 52 non operating environment R 11 number of host computers to connect with 2 25 0 ODE 7 46 on off switch removal 9 46 9 66 9 87 9 106 Operating environment R 10 Operating systems compatible with disk array R 2 Operational commands NIKEARRY 7 47 outer cover removal 9 70 P parity 1 18 1 25 1 30 1 32 1 35 1 42 password Grid Manager adding 6 71 changing 6 72 removing 6 72 patch media options R 6 peer SP on Presentation screen 6 35 physical volumes creating 7 129 power AC input specifications R 9 DC specifications R 10 power block diagram 8 40 power consumption R 10 power cord connecting 2 42 part numbers R 12 power supply unit 1 14 removal replacement 9 19 status on Presentation screen
150. 8 4 8 5 8 6 8 7 8 8 8 9 8 10 8 11 8 12 8 13 8 14 8 15 8 16 8 17 VO gt Read Disks Output Screen 0 0 cece eee n ees 7 117 VO Read LUNS Screen 6 cs ke oe eb ek See eee eae See Ra 7 118 VO 2 Read LUNs Select Blocks Screen 0 000 ccc eee 7 119 VO gt Read LUNs Output Screen sese 7 119 Test Unit Ready Screen sre slo RU DERE pie Ig gp 4A Iu ui 1 120 Tests Verify Disks Screen essel eee 7 121 Tests Verify Disks Select Blocks Screen 0 0 cece eee eee 7 121 Tests Verify LUNs Screen 0 cece cee eee ee 7 122 Tests Verify LUNs Select Blocks Screen 0 0 c eee e eee ences 7 123 Utilities System Configuration System Cache Screen 1 124 Utilities System Configuration System Memory Screen 1 125 Utilities System Configuration Host I F Settings Screen 1 126 Model 10 FRU Locations 0 ccc cece eh hn 8 3 Model 20 Deskside Disk Array 2 cece cee eee eens 8 4 Model 10 Rackmount FRU Locations 0 cc eee eee eee 8 5 Model 20 Rackmount FRU Locations 0 cece cece eee eee eee 8 6 Front LED Locations iesise 0 ccc ccc cee eee hh n 8 7 Rear LED Locations ocv ae eee e was th Rede RI eS a a ales 8 8 Model 10 Mapping Microcode 7 23 1 1 0 0 cece eee eee eens 8 21 Model 20 Mapping Microcode 7 23 1 1 0 0 cee eee eee eee 8 22 Mo
151. 86 CRU Format Failed The format of the drive aborted for some reason Further information may be provided in other Unsolicited Event Log messages 0x87 CRU Idle Powered Down CRU was found to have powered down while idle 0x88 CRU Idle Dead CRU died while in idle state 0x89 CRU Bind Powered Down The drive powered down while unbound 0x8A CRU Bind Dead CRU died while unbound 0x8B CRU Wait for Assign Powered Down CRU powered down while in the RDY state 0x8C Shutdown Shutdown while waiting to assign a LUN Usually seen with event code 0xa07 CRU died in RDY state 8 74 Unsolicited Event Log Table 8 15 Extended Status Codes cont d Extended status code Description 0x8D CRU Rebuild Powered Down CRU powered down during a rebuild Ox8E CRU Rebuild Dead CRU died during a rebuild Ox8F CRU Enabled Powered Down DRU powered down in ENA state 0x90 CRU Enabled Dead CRU died in ENA state Microcode Panics When a microcode panic occurs two 32 bit values are written to the Unsolicited Event Log The first 32 bit value identifies the type of panic The four high order hexadecimal digits of the panic type indicate in general which firmware module reported the panic If a microcode panic occurs use the checklist at the end of chapter 2 to collect relevant information and report it to your WTEC representative Be sure to reference the two 32 bit values in the Unsolicited Event Log Table 8 16 shows th
152. 9 48 Power off the disk array and disconnect all cables The fans continue to run for a few seconds Remove the outer covers Figure 9 18 Remove the following a Fan pack Figure 9 6 b Power supplies or fillers Figure 9 11 c SPs or fillers Figure 9 9 d IPCB Figure 9 21 e All disk modules or fillers Figure 9 3 Be sure to label the disk modules as they are removed from the disk array f BBU Figure 9 13 Turn the disk array over on its top a Remove the wheel base i Remove the retaining screws from the front of the wheel base ii Slide wheel base 1 inch toward back and remove Remove the inner right metal panel 11 screws Figure 9 19 Remove the inner chassis Lay the chassis onto its left side a Remove the six retaining screws from the top b Lay the chassis onto the covered side c Remove the seven retaining screws from the bottom d Lift the inner chassis up and out of the system i Rotate upward the rear of the inner chassis about 3 inches ii Lift up and out Model 10 Deskside Component Removal 7 Disconnect all spade lug wires on the AC line filter PCA 8 Remove three screws and lift the AC line filter PCA away from the mounting studs To wiring harness AC line filter Retaining screws 1 hidden Figure 9 28 Removing the AC Filter Model 10 Model 10 Deskside Component Removal 9 49 jueuie2e deH pue jeAoway Model 10 Rackmount Component Removal
153. 9009 sses S99S9999999 Ei Sng X Torx 25 retaining screws 4 2 of 4 shown 9 9 9999999599999999 S SSSS 59 Figure 9 29 Removing the Array from the Rack Model 10 Rackmount Component Removal 9 51 jueuie2e daH pue jeAoway Midplane Figure 9 30 and Figure 9 31 Note For these procedures the disk array must be powered off and unplugged Also it is recommended that ESD procedures be followed Use ESD kit P N 5182 4119 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 Remove the following a b C Fan pack Figure 9 6 SPs or fillers Figure 9 9 Power supplies or fillers Figure 9 11 Front panel Figure 9 18 BBU located in the lower front by pressing in on the two tabs at the sides of the unit Figure 9 13 All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array 4 Remove the four retaining screws Figure 9 29 WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array 5 Remove the empty chassis from the rack Figure 9 29 9 52 Model 10 Rackmount Component Removal 6 Remove 11 screws and remove the top panel of the chassis Figure 9 30 7 Remove the inner chassis Figure 9 31 a Remove the retaining screws from each side of t
154. A3 and A4 model 10 or AO BO C0 DO and EO model 20 Follow the procedures in this section when Setting up a new disk array that supports caching Set up disk array caching before making newly bound logical disk units accessible to the operating system With disk array caching set up a logical disk unit with either the read or write cache enabled can use caching as soon as the unit is accessible to the operating system Upgrading the system to support caching Set up caching after installing the prerequisite hardware To use caching an array must have two SPs Also the amount of cache on each SP must be mirrored equal amounts of cache on each SP A minimum of 8 Mbytes of SIMM memory on each SP is required for caching The following table shows the combinations of SIMMs that can be used for caching Orders for cache SIMMs part numbers A3234A or A3235A must be made according to Table 6 6 Non cache SPs have two 1 Mbyte SIMMs installed which must be removed and replaced with the SIMM combinations from the table in order to use caching Table 6 6 SIMM Sizes for Model 10 and Model 20 Model 10 Model 20 8 Mbyte 4 x 4 Mbyte 8 Mbyte 4 x 4 Mbyte four A3235A SIMMs 16 Mbyte 8 x 4 Mbyte 16 Mbyte 8 x 4 Mbyte eight A3235A SIMMs Setting up Disk Array Caching for Microcode Revision 7 23 6 53 Jobeue N pus Table 6 6 SIMM Sizes for Model 10 and Model 20 cont d Model 10 Model 20 4 or 8 of A3235A 32 M
155. AID 3 Group Parallel Access Array Displays dialogue for making a RAID 3 LUN 3 RAID 1 Mirrored Pair Displays dialogue for making a RAID 1 LUN FE 4 RAID 0 Group Non redundant Individual Access Array Displays dialogue for making a RAID 0 LUN 5 RAID 1 0 Group Mirrored RAID 0 Group Displays dialogue for making a RAID 1 0 LUN FE 6 Individual Disk Unit Displays dialogue for making an independent mode disk 7 Hot Spare Displays dialogue for making a hot spare 3 Unbind Logical Units Displays dialogue for unbinding a LUN 6 28 Grid Manager Menus 4 View Maintenance Information 1 Display Unsolicited Event Log Displays the Unsolicited Event Logs starting at entry number 1 Must use FE mode to see all entries USER mode only shows 0x9xx and OxAxx errors Allows selection of a filter for either error code or FRU 2 Display Faulted CRUs Displays detailed information on the cause of CRU failures 3 Display Verify Results Displays results of LUN the verification test launched using the menu path 5 Change Parameters 1 Change Logical Unit Parameters 6 Change Unit Verify Configuration 4 Display LUN Configuration Summary Displays LUN numbers ownership RAID type cache states and number of disks in LUN FE 5 Display Disk Failure Information Displays detailed information about disk failures FE 6 Display Non Vol RAM Unsolicited Error Log Displays non volatile RAM error information 5 Change Parameter
156. ARRAY or NIKEARRY DOWNLOAD commands or the STM Updating Microcode and Firmware sections in chapter 7 for information about how to download new firmware to the SP s CAUTION The upgrade process can be undone but not easily Contact your WTEC 8 36 representative for the restore procedure which is documented in a Nike Note if a failure or major problem is encountered and an older version of microcode must be restored When it has been verified that microcode revision 8 20 is running on the SPs use the HARRAY DOWNLOAD procedure to upgrade to microcode revision 8 56 Contact your WTEC representative for information about where to obtain the microcode See the HARRAY or NIKEARRY DOWNLOAD commands or the STM Updating Microcode and Firmware sections in chapter 7 for information on downloading new firmware to the SP s Power down the array and remove the old SP s Install a minimum of 8 Mbytes of SIMM on the new SP620 SCSI SPs Use the SIMMs from either the old SPs or new SIMMs Install the new SP620 SCSI SPs into the array and power up the array The array should boot successfully Upgrading from an SP to a SP620 SCSI SP If garbage displays on the Grid Manager screen after booting the array make sure the parity of the controller serial port is set to NONE 8 bit Upgrading from an SP to a SP620 SCSI SP 8 37 Bunoouse qnoJj Cache Configuration Instructions On all versions of 8 xx microcode for both the SP and SP620 SCS
157. ATA function key If more cards on this DTC have GridMgr iX RS 232 ports to be configured press the PRIOR SCREEN function key to return to the DTC Configuration screen and repeat steps 6 through 8 for the other cards on this DTC If additional DTCs have other GridMgr iX RS 232 ports to configure press the PRIOR SCREEN function key twice to return to the DTC Selection Configuration screen Repeat steps 4 through 9 for the other DTCs When all cards in all DTCs having GridMgr iX RS 232 port connections have been configured continue with the following steps 1 Move the cursor to the Command line at the top of the DTC Card Configuration screen type validate and press RETURN The Validate Configuration File screen displays Press the VALIDATE DTS LINK function key to validate the DTCs Check the screen to make sure no errors are found during the DTS subsystem validation process When the validation completes successfully exit the NMMGR 3000 by moving the cursor up to the Command line typing exit and pressing RETURN Configuring RS 232 Ports on HP 3000 Series 900 Systems 6 9 Jobeue N pus Initiating Grid Manager on HP UX Before logging into Grid Manager it is necessary to change the configuration of the HP terminal being used to a termmode of VT100 or EM100 The disk array requires these types of emulation modes If running Grid Manager from an HP VUE window use an xterm window by executing the xterm amp command If the
158. BLED ENABLED ENABLED Allow PQ 3 DISABLED ENABLED DISABLED No spaces in PID DISABLED ENABLED DISABLED Spindle Synch DISABLED ENABLED ENABLED Trespass Logging DISABLED ENABLED ENABLED 4 Save Options as Defaults Saves the current disk array configuration as the default Grid Manager Menus 6 31 Jobeue N pus 5 Restore Default Options Restores the disk array configuration previously saved using Save Options as Defaults 6 View Cache Statistics Displays dialogue for displaying cache statistics FE 7 Back End Exerciser Displays dialogue and instructions for running the back end exerciser FE Must be in FE mode to access these screens or functionality To enter FE mode type COnTROL P followed by in upper case FE To return to user mode type CONTROL P followed by in upper case USER 6 32 Grid Manager Menus Presentation Utility Screen The Presentation Utility screen is an annotated grid Figure 6 7 shows a typical screen To display the Presentation Utility screen select the Presentation Utility option of the Grid Manager Main menu Hicrocode Revision 09 55 01 Statistics Logging DISABLED Peer SP PRESENT Requests Completed 71 Cache States Write DISABLED Enter for HELP to GO TO HAIN HENU Read DISABLED Enter Choice Figure 6 8 Typical Presentation Utility Screen SP Status The top portion of the Presentation Utility screen displays the current SP SP SCSI ID serial
159. BOBO epo Bee Bpaa Babe Gi BB BOR BHO BR gage Bons pui jiii Bub Bi ania Deme cue pun sup anon anon Dene guod anon ono ano aman maa mdp i DA Pd re re me me e SEHEZZEHEZZ Figure 7 15 I O Read Disks Output Screen 7 88 Support Tools Manager STM T O Read LUNs The following steps are required to view LUN information 1 Select I O Read LUNs from the Expert Tools main menu Select LUNs from the dialog box that appears Figure 7 16 2 Select the desired action and click Yes The Select Blocks dialogue box appears Figure 7 17 3 Define the appropriate block parameters and click OK When the system completes the LUN read the output window appears Figure 7 18 with information about the LUNs and blocks specified d lid Heas i wai bi derar ep ee fe r E Thi command will feed blecke sectors of data rn Sitesi lb Limi irs Array aria d H5 75335 85H Select deairsd accion Cx Road ALL fing ALL blocks as Brimi in get solection o Head SELECT luna ALL blacks Blocks B they BEGBH3I lu Blocks B thru LE 11I38l a Blacke B shru BRNA Blarka B chru GEARAN Blarka B chru Hobe Blacks yeu will be asked b Block you will bs asked Blacks usu will ba axkad Blackes you will ba aakad BMlarka dijou will be sakad Figure 7 16 I O Read LUNs Screen Support Tools Manager STM 7 89 sonsoubeiq pue seni ISOH fal Lunt j i d
160. CA Packar Hewlett Packard High Availability SCSI Disk Array Service Manual High Availability SCSI Disk Array Service Manual U HEWLETT PACKARD Edition E1299 Order No A3232 90010 Printed in U S A Notice Hewlett Packard Company 1994 1995 1996 1997 1998 1999 All rights reserved Hewlett Packard Company makes no warranty of any kind with regard to this material including but not limited to the implied warranties of merchant ability and fitness for a particular purpose Hewlett Packard shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing performance or use of this material This document contains proprietary information which is protected by copyright No part of this document may be photocopied reproduced or translated into another language without the prior written consent of Hewlett Packard The informa tion contained in this document is subject to change without notice Print History Edition 1 December 1994 Edition 2 February 1995 Edition 3 September 1995 Edition 4 November 1996 Edition 5 January 1996 Edition 6 January 1997 Edition 7 June 1998 Edition 8 August 1998 Edition 9 December 1999 Warranty If you have any questions about the warranty for this product contact your dealer or local Hewlett Packard sales representative Safety Instructions To protect against personal injury and or product damag
161. CRU Powered Down 04 23 98 14 19 01 SP A 0x9008 Fault Cache Disabling 04 23 38 14 19 01 CO Oxf007 CRU Powered Down 04 23 98 11 44 42 AO 0x9001 Hard SCSI Bus Error Hit any Key to Continue ll Figure 8 19 Non Vol RAM Error Log 8 52 Viewing Maintenance Information Unsolicited Event Log Displaying the Disk Array Unsolicited Event Log Each SP maintains a log of event messages in its memory These events include hard errors and start ups and shut downs involving disk modules fans SPs power supply units and the BBU Periodically the SP writes this log to disk to maintain it when SP power is off The log can hold more than 2 000 event messages and has a filter feature that allows events to be selected by device or error message To display the Unsolicited Event Log when running microcode revision 8 56 and earlier choose option 4 View Unsolicited Event Log on the Grid Manager Main menu For microcode revisions 9 xx select 4 View Maintenance Information on the Grid Manager Main menu then select 1 Display Unsolicited Event Log SP B Display Unsolicited Event Log SP SCSI ID 1 Event Date CRU Event message Extended Status 1 08 26 94 13 59 10 VSCA 0x904 VSC Shutdown removed 0x00 2 08 26 94 15 01 02 0xa07 CRU Powered down 0x00 3 08 26 94 15 01 23 0x906 Unit shutdown 0x00 4 08 26 94 15 08 09 0x906 Unit shutdown 0x00 Figure 8 20 Unsolicited Event Log Display The event messages are in order by time elapsed sinc
162. Cables 2 41 uonejelsu Connect the Power Cord 1 Make sure the disk array s power switch is off then connect the power cord to the disk array as shown in Figure 2 13 2 Plug the other end of the power cord into an AC power outlet Rear of storage system Figure 2 13 Connecting the Power Cord and Enabling Power Typical 2 42 Connect the Power Cord Power Up the Disk Array 1 Turn on the disk array s power by setting the POWER switch on the back of the unit to ON The green POWER LED on the front of the disk array will light up and the fans will rotate 2 Ifthe busy light on none of the drive modules lights up make sure the SP is enabled as follows a Move the fan pack latch to the UNLOCK position b Open the fan pack c Move the SP power switch to the ENABLE position d Close the fan pack within two minutes by reversing steps A and B 3 Power up the host computer and have the system administrator reboot the host s Refer to the host computer s operation manual for information on how to power up the host computer Determine Terminal Connectivity Verify that the Grid Manager utility can be accessed Grid Manager runs continuously when the disk array is powered up Grid Manager output should display on the terminal as soon as terminal emulation or a dedicated hard terminal has been set up There is no need to explicitly start or stop the
163. Configuration Multi Bus Initiator Configuration HP UX 10 x and Later 4 15 008 seues 0006 dH uoneanbiuo2 Table 4 7 Multi Bus Initiator Error Recovery Failing component Disk module Continue after failure Yes What happens and how to recover With RAID of any level other than 0 or an individual disk aApplications continue running System Administrator or service provider replaces module Provides the same features as the basic configuration but with better performance and higher availability With RAID of any level other than 0 or an individual disk applications can continue to run after the failure of a single disk module SP Yes If one SP fails ownership of the failed SP s LUNs can transfer automatically to the remaining operational SP If a primary and an alternate path have been configured in LVM LVM switches automatically to the alternate path These operations are transparent to the user Fan pack Yes Applications continue running System Administrator or service provider replaces module Power supply Yes If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply The System Administrator restarts the operating system if necessary SCSI 2 adapter cable 4 16 Yes I O operations fail to all physical disk units configured on SP attached to a failed SCSI 2 adapter cab
164. Duuo25 Disk Array Support Requirements for HP UX Release 9 05 9 07 The high availability disk array is supported on HP UX operating system releases 9 05 and 9 07 when patches are installed See your WTEC representative for the latest patch information Installing and Reconfiguring on S700 Systems When installing a disk array on a Series 700 system first determine which version of the HP UX operating system is running on the host system s If HP UX 10 01 or greater is running install the disk array according to the Series 800 installation procedures except use a 24542H cable instead of a 17355D cable for any serial port connection from the disk array to the CPU See chapter 2 for installation procedures If running HP UX 9 05 or 9 07 the disk array must be reconfigured HP UX 9 05 and 9 07 impose a 4 Gbyte file system size limit If using 2 Gbyte disk modules in a model 10 disk array the array can be configured in three disk RAID 5 two disk RAID 1 or four disk RAID 1 0 configurations The minimum configuration is three 2 Gbyte disk modules bound in a RAID 5 group which is the factory default and requires no special action at installation If more than the minimum number of disk modules or a model 20 disk array is installed the disk array must be reconfigured to an acceptable RAID configuration If 4 Gbyte disk modules are used only RAID 1 groups can be configured The 4 Gbyte file system size limit applies to 4 Gbyte and 2 Gbyte disk mo
165. E3 About the Disk Array Deskside Model Front View Database Drives AO BO CO A3 Cache Vault Drives AO BO CO DO EO A A0 j B1 BO c CO D1 DO Ei EO A3 A2 B3 B2 c3 C2 D3 D2 E E3 E2 Oo o O O O 1 AO BO COI DO EO A2 B2 C2 D2 E2 AA B1 C1 D1 ET1 AS B3 C3 D3 E3 Rackmount Model Front View Database Drives AO BO CO A3 Cache Vault Drives AO BO CO DO EO Figure 1 4 Model 20 Disk Array Chassis About the Disk Array 1 7 uonduioseg jonpold SCSI 2 RS 232 RS 232 SCSI 2 CuO Optional Storage Control SCSI A SCSI B Storage Control processor SCSI C SP processor SP SCSI D SCSI E UU U UU UU OO UU UU UU UU Battery Backup Unit optional Power supply Power supply Optional power supply E e eR Figure 1 5 Model 20 Internal Hardware Architecture 1 8 About the Disk Array About Disk Array Components The disk array consists of the following components Three to 10 disk modules for a model 10 five to 20 disk modules for a model 20 One or two SPs One fan pack One or two power supply units for a model 10 two or three power supply
166. ER 0 FRU Slot Number A0 FRU NUMBER 6 FRU Slot Number A1 FRU NUMBER 12 FRU Slot Number A2 FRU NUMBER 18 FRU Slot Number A3 Enter FRU NUMBER for testing or 99 to Cancel gt 18 MECH UTILITY 4 20 PM Sending Read Capacity Command to FRU Number 18 READ CAPACITY INFORMATION Maximum Block Address 4123241 Block Size 520 Drive Capacity 2144085840 7 32 HARRAY Section 17 External Exerciser Read Capacity Command SUCCESSFUL MECH UTILITY COMPLETED 4 20 PM Example 4 MECH command with option 3 HARRAY gt mech The following commands are available at the mech level 0 Test Unit Ready TUR 1 Inquiry 2 Read Capacity 3 Read 4 Read Media Test RO MT 5 Verify via scsi commands 99 Exit Enter number 0 5 or 99 to Cancel 4 gt 3 The following mechs FRU s are available for test FRU NUMBER 0 FRU Slot Number A0 FRU NUMBER 6 FRU Slot Number A1 FRU NUMBER 12 FRU Slot Number A2 FRU NUMBER 18 FRU Slot Number A3 Enter FRU NUMBER for testing or 99 MECH UTILITY 4 21 PM to Cancel 6 A Read Block Command will be issued to FRU Number 6 Enter Block Address where the Read shall begin 0 4121402 gt 2 Reading one block of data beginning at block address 2 Read Command was SUCCESSFUL MECH UTILITY COMPLETED 4 21 PM Example 5 MECH command with option 4 sub option 1 HARRAY mech The following commands are available at the mech level 0 Test Unit Ready T
167. ET Nikearry command 7 61 RETRIEVE UNSOLICITED LOGS HARRAY command 7 41 REVISION HARRAY command 7 39 RIP PCA removal 9 75 9 94 RIP PCB function of 8 42 LEDs 8 43 troubleshooting 8 42 RO MT HARRAY command 7 40 S safety certifications R 12 SAM 7 127 8 90 bind aLUN 8 96 unbind a LUN 8 93 view status of an array 8 90 SCSI cable model 10 deskside removal 9 41 model 10 rackmount removal 9 62 options R 6 SCSI IDs for SP 2 29 SCSI 2 adapter 1 17 SCSI 2 bus connector locations 2 23 internal model 10 1 4 internal model 20 1 5 length 2 21 sense key code qualifier values 8 23 serial number array 7 85 7 114 setting SCSI IDs 2 29 setting the enable disable switch 9 15 setting up disk array caching microcode rev 7 23 6 53 microcode rev 8 20 6 59 setting up the disk array on a MUX port Series 700 2 37 Series 800 2 36 setting up the disk array on an HP 3000 DTC 2 38 SHOW MECHS HARRAY command 7 42 SIMM about 8 31 disk array failure 8 15 for model 10 9 25 for model 20 9 25 locations on SP 9 25 locations on SP6xx SCSI SP 9 25 removal replacement 9 24 troubleshooting SP 8 31 SP6xx SCSI SP 8 32 single host 2 25 software mirror 1 24 1 26 1 32 SP aboutthe 1 10 changing parameters 6 65 enable disable switch 1 11 identifier on Presentation screen 6 33 removal replacement 9 14 SCSI ID on Presentation screen 6 34 troubleshooting 8 27 upgrading to a SP6xx SCSI 8 35 SP storage control processor disk array failure 8 15 microcod
168. ETURN Enter choice Acceptable choices are f 1 p n a c or Figure 8 21 Apply Filters Menu 2 Choose the type of filter The FRU Filter prompt to specify the filter appears 3 To choose a FRU filter enter the FRU s ID to choose an error code filter enter the error code The Apply Filters menu reappears Figure 8 21 4 Atthe prompt type caret Unsolicited Event Log 8 55 punoouse qnouj The Unsolicited Event Log appears listing only the messages specified by the filter If an error code filter is specified while an FRU filter exists or vice versa the program uses both filters the filters are ANDed Note Filter s are effective until deleted or until disk array power goes down If the log listing seems short one or more filters may be active In this case delete all filters as explained next To delete all filters of one type 1 At the Event Log prompt type the letter A The Apply Filters menu appears Choose the type of filter FRU or error code to be deleted A prompt to specify the filter appears 2 At the prompt press RETURN All filters of the specified type are deleted SP Event Codes The following tables contain the list of event codes to be found in the Event Code field EVENT field within an Unsolicited Sense Data block They describe abnormal events on a FRU basis and fall within three categories e Informational codes Require no action on the part of the user The
169. FAULT LED lit Disk Module Disk module has encountered a fault Drive is ready READY LED lit or FAULT amber condition and cannot be accessed busy BUSY LED lit Fan pack Fan pack detected a fault with one of Normal operation REPLACE amber the fans SP Normal operation Array is off or failing to provide 24 READY green VDC to the 24 VDC bus or the SP has a fault FAULT LED lit SP SERVICE red SP has just been powered on and is running power on self tests or has encountered a fault condition and is no longer configured in the array Normal operation Power Supply Unit READY green Normal operation Array is off no AC power reaching the power supply or the power supply has a fault FAULT LED lit Power Supply Unit REPLACE amber Power supply has encountered a fault condition and is no longer configured in the array Normal operation BBU READY green model 20 only Normal operation Normal operation BBU CHARGING green model 20 only Batteries in BBU are being charged Normal operation BBU REPLACE amber BBU has encountered a fault and is no longer operational Normal operation Rear View Disk Array LEDs 8 9 Buljooysajqnoly Power Up Troubleshooting Power Up Sequence for the SP When the disk array is powered up each SP cycles through a series of power up tests The power up tests for the SP display the following Copy
170. FRU FRU amp y a E 1 18 19 20 21 22 Ai Bi Ci Di Ei AG BS C3 D3 ES Figure 7 1 Disk FRU Map Model 20 Deskside A FRU FRU FRU FRU FRU B FRU FRU FRU FRU FRU Oo 6 12 18 24 1 7 13 189 25 o 1 2 3 4 0 1 2 3 4 Figure 7 2 Disk FRU Map Model 10 Rackmount On a model 10 disk array to bind the first three FRUs on the B Bus type 1 7 13 in response to the following prompt Enter 1 to 16 FRUs between 0 and 29 to be bound 0 4 0 29 0 1 7 13 0 4 0 29 q The disk modules in slots BO B1 and B2 would be bound together NIKEARRY 7 55 sonsouDeig pue seni ISOH Table 7 4 FRU Map All Disk Array Components at All Bus Locations Bus Quadrant A B C D E 0 0 1 2 3 4 1 6 7 8 9 10 2 12 13 14 15 16 3 18 19 20 21 22 FRUINFO The FRUINFO command displays a FRU Information Page for the selected disk FRU to the console FRUINFO takes two inputs as parameters a LUN index and a FRU number If no parameters are given the user is prompted for the LUN index and FRU number When FRUINFO is issued with one parameter it is taken as the FRU number and the LUN index is taken from the TESTUNIT environment variable Example NIKEARRY gt fruinfo Enter test array cr 0 5 0 Enter FRU number cr 0 29 1 The Disk FRU Information Page contains Disk FRU Number 1 0x1 Logical Unit Number 0 FRU LUN Type Disk Array RAID 5 S
171. GB 7 16 7 97 7 16 7 97 8 51 1 49 A3396A 4 2 GB 7 16 7 97 7 16 7 97 8 51 1 49 A3541A 8 8 GB 8 27 8 02 8 27 8 02 8 58 1 57 A3666A 4 2 GB 8 27 8 02 8 27 8 02 8 58 1 57 A5132A 8 8 GB Not supported 9 25 00 8 02 9 55 01 1 73 A5133A 17 8 GB Not supported 9 25 00 8 02 9 55 01 1 73 A5169A 4 2 GB Not supported 9 25 00 8 02 9 55 01 1 73 Disk Troubleshooting Disk Troubleshooting Overview Verify the disk module is properly seated in the appropriate slot before running further disk diagnostics See Chapter 9 Removal and Replacement for procedures Before replacing a faulted disk determine how the disk is configured within the array Disk s can be configured in RAID 0 RAID 1 RAID 1 0 or RAID 5 modes and also as a hot spare or individual disk on HP UX Disk s can be configured in RAID 1 or RAID 5 modes on MPE X Before replacing the failed disk module determine what action to take after the disk module is replaced and obtain information about the failure Unsolicited Event Logs and Presentation Utility screen If the disk is a part of a RAID 1 RAID 1 0 or RAID 5 configuration a rebuild occurs automatically when the failed disk is replaced by a new disk and no additional action is required If the failed disk is part of a RAID 0 group or an individual disk the data must be restored to the RAID 0 group or to an individual disk If the failed disk is a hot spare no additional action is required after the disk is replaced 8 20 Trouble
172. GERM Errag on parh W 4 B Pirre Dara BD I a8 Be eat fadam Dara Bie id Se Fera Tiss 17 2c oF Pee 5075 00 1 00 Ha Ties 13 Bec te Fere SF BOOCTNES Fam Sabla Fase ail bero beKaremd dum Lue liara Dmm dimlin b s ijies VermburiD Preker O TEX LTOGNEREBNCS rebel air i CTIA BALE EJF CEN Rare HI CENERI EISEN SIF CEA 3 CEI DI Co Nan d d ra kii TWO uM ITs IH Figure 7 11 Info Capacity Screen Info gt Inquiry Tk 43 318 Ls Fuperr Teed regiinrg lar desire GERM Array on parh H O4 Bo Ferr Dara WD un Zh o STRE aries Dara 26 Ere Ties i Fe LS 43 72 00 aia Ties 13 Prej F BINE Fam yale hiamh bun VerdarbD Frihet 10 LEF LH Ina Cra tad Couma CEE Coa Frisur 10 Figure 7 12 Info Inquiry Screen 7 86 Support Tools Manager STM I O Read Disks The following steps are required to view disk information 1 Select I O gt Read Disks from the Expert Tools main menu Select disks from the dialog box that appears as shown in Figure 7 13 2 Select the desired action and click Yes The Select Blocks dialogue box appears Figure 7 14 3 Definethe appropriate block parameters and click OK When the system completes the disk read the output window appears with information about the disks and blocks specified Figure 7 15 nap on aib Bd E This command will Hu sd Becks sectors of daia om eit On Array Boria M mn rnan
173. Host Interface Options for Microcode Revision 8 20 Microcode revision 8 20 includes some changeable Host Interface Options that must remain in their default states Changing these parameters from their default values can create problems The parameters settings should be the following Addressing model LUN Target initiated negotiation enabled Substitute busy for qfull disabled Mode page 8 enabled Recovered error log reporting enabled Allow non mirrored cache disabled Allow non mirrored cache is a new parameter which is not supported It must always be disabled If the disk array contains only one SP this parameter allows cache to be enabled 6 68 Changing Storage System Parameters However caching cannot be mirrored because a single SP is present If the SP fails the data in cache may not be posted onto the disk module which may result in data integrity problems Default Host Interface Options for Microcode Revision 9 xx The microcode includes some changeable Host Interface Options that must remain in their default states Changing these parameters from their default values can create problems The parameters should be set as shown in Figure 6 17 m xterm SP A SP SCSI ID 15 S N 9621954 Change Host Interface ptions This menu allows you to define the behavior of this SP s interface to host systems Incorrect use of these options can result in incompatibilities which may make host systems unable to connect to this SP The
174. I SPs the cache can be configured with various capacities of read and write cache A rule of thumb is to allocate about twice as much write cache as read cache For example if a SP620 SCSI SP has 64 Mbytes of memory the required 4 Mbytes for buffer space should be allocated leaving 60 Mbytes for cache Allocate 40 Mbytes for write cache and 20 Mbytes for read cache Different applications require different combinations of read and write cache The best fit for the specific system and applications needs to be determined by the customer The procedure for configuring cache follows 1 Make sure caching is disabled To disable the cache from the Grid Manager Main menu follow the menu path 5 Change Parameters 2 Change Storage System Parameters 1 Change Storage System Caching Parameters a Type Y to the Change caching parameters N prompt b Accept the default on all remaining prompts by pressing RETURN 2 Change the memory allocation To configure memory follow the menu path from the Grid Manager Main menu 5 Change Parameters gt 2 Change Storage System Parameters 2 Change System Memory Configuration The memory configuration will display a Type Y to the Change memory configuration N prompt b Allocate 2 Mbytes for buffer space on the SP or 4 Mbytes for buffer space on the SP620 SCSI SP A rule of thumb is to allocate about twice as much memory for write cache than for read cache with the remaining memory
175. Look for the HARAYMON process usr diag bin HARAYMON On HP UX 10 x HARAYMON is not a daemon process and cannot be seen with the ps ef command On HP UX 10 x HARAYMON should appear as an executable file when the following is typed Il usr sbin diag HARAYMON HARRAY the disk array diagnostic and HARAYMON the disk array monitor can be installed on HP UX 9 04 from the patch tape that comes with the disk array The patch tape label for HP UX 9 04 reads for example HP UX 9 04 Patch s800 For the HP9000 S800 T OE Update FMT DATE CODE 09N4 P N B3108 13872 REV A B9 04 Contact your HP WTEC Representative for the latest patch tape and firmware revision numbers HARRAY and HARAYMON are available for HP UX 9 05 and 9 07 on patch file PHSS 7587 Install this patch file onto the HP UX operating system To install HARRAY and HARAYMON 1 Type cd usr diag bin 2 Type update Select all filesets on the patch or update source media Load all filesets 3 Check tmp update log for update errors by typing more tmp update log 7 6 High Availability Disk Array Monitor HARAYMON 4 If running HP UX 9 x check to see if HARAYMON is running by typing ps ef grep HARAYMON Look for the HARAYMON process usr diag bin HARAYMON Check to see if the HARRAY diagnostic was installed by typing ioscan fC disk Choose a PDEV associated with the disk array Look for NIKE Nike or nike to find one Type SUPLICEN On the next line type the
176. MM on the SP are interchangeable with SIMMs on the SP620 SCSI controller However each SP620 SCSI SP must have a minimum of 8 Mbytes SIMM capacity two SIMMs each x 4 Mbyte Install replacement or add additional SIMMs by performing the following steps Note SIMMs always must be installed in pairs and must be equal in capacity 1 Observe antistatic precautions CAUTION It is recommended that the ESD wrist strap be grounded to the disk array chassis or suitable ground and strapped to your wrist when removing or handling SIMMs and place them on the ESD conductive sheet when they are removed from their packaging Use ESD kit P N 5182 4119 2 Remove the SP Figure 9 9 a 9 24 Disable cache to force the unposted data to be written to the disk modules If this is not done the unposted data will be lost See Disabling and Enabling Disk Array Caching in chapter 6 for information about disabling caching Open the fan pack Disable the enable disable switch on the SP The SP620 SCSI SP does not have an enable disable switch Grid Manager is used to disable it See Enabling Disabling the SP620 SCSI Storage Processor in chapter 6 for information Pull both ejectors until they unlock and the SP starts out of the chassis Pull the SP out of the chassis making sure not to touch any electronic components or circuits Place the SP on a static free work surface Model 10 20 SIMM Removal Replacement 3
177. Model 20 64 dBA 48 dBA Safety Certifications UL listed CSA certified TUV certified with GS mark Replacement Power Cord Part Numbers by Country 8120 1351 UK 8120 1369 Australia 8120 1689 Europe 8120 1751 North America 120 VAC 8120 3996 North America 240 VAC 8120 2104 Switzerland 8120 2956 Denmark 8120 4211 South Africa 8120 4753 Japan 8120 1625 Cabinet IEC320 C13 to C14 R 12 Reference EMC Compliance US FCC Canadian DOC EN55022 Class A VCC1 Level 1 IEC801 2 IEC801 3 and IEC801 4 Reference R 13 o2ue49JoH Replaceable Parts Model 10 Figure R 1 Parts Breakdown for the Model 10 Disk Array Reference R 14 Table R 1 Replaceable Parts Model 10 Figure index New HP part Exchange part number number number Description 1 005 41327 A3231 69007 Model 10 midplane 2 005 039649 N A LED cable 3 N A N A External chassis 4 005 041006 N A Fan cable 5 N A N A Internal AC harness bracket 6 005 041698 A3231 69008 Interconnect PCB IPCB 7 005 041068 N A Internal synchronous SE SCSI 2 cable 8 113 000552 N A On off switch 9 005 041008 N A Power supply AC cable 10 N A N A Internal chassis N A 5063 5396 A3389 69021 4 Gbyte disk module N A 5063 5398 A3232 69121 2 Gbyte disk module N A 5063 5393 A3666 69021 4 Gbyte LP disk module N A 5063 5394 A3541 69021 9 Gbyte HH disk module N A A3231 60005 A3231 69005 Battery bac
178. N A Power supply AC cable 10 N A N A Internal chassis N A 5063 5396 A3389 69021 4 Gbyte disk module N A 5063 5398 A3232 69121 2 Gbyte disk module N A 5063 5393 A3666 69021 4 Gbyte LP disk module N A 5063 5394 A3541 69021 9 Gbyte HH disk module N A A3231 60005 A3231 69005 Battery backup unit BBU N A A3231 60003 A3231 69103 Storage processor SP without SIMMs N A A3231 60004 A3231 69104 Power supply unit VSC N A A3231 60006 A3231 69006 Fan pack Illustrated Parts Breakdown for the Model 10 u w vsejd y pue jenow y Table 9 1 Model 10 Disk Array CE Replaceable Parts see Figure 9 1 cont d N A 5062 9398 A3232 69022 various vendor numbers including MT12D136G 7 MH1M36ADJ 7 HB56D136BR HB56D136BV 7B MH1M36BDJ 7 THM361020ASG 70 MCM36100ASG70 KMM5361000AG KMM5361000BG 7 KMM5361000CG 7 MH1M36BYJ 7 KMM5361003CG 7 KMM5361003BG 7 TM124MBK36B 70 KMM5361203AWG 7 SM5361000WG 7DG 4 Mbyte SIMM N A 5062 9399 A3232 69020 numerous vendor numbers including HB56D436BR 7A MH4M36AXJ 7 MH4M36BXJ 7 HB56D436B 7RB SCM421186SG70 SM5364000G 7DG KMM5364100AG 7 SM536044002P3G7 KMM5364103AKG 7 16 Mbyte SIMM N A 008 040801 N A Model 10 AC line filter PCB N A 123 002825 N A Model 10 20 caster N A 5063 5397 N A Model 10 20 disk slot filler panel N A 005 041690 N A Model 10 rackmount front bezel N A 9 4 005 041691 N A
179. N A Saves the contents of main STM window to a file Print N A Prints the contents of main STM window Exit N A Exits the Expert Tool Logs View Unsolicited Log Event Codes All Displays Unsolicited Event Log information Event Codes gt 0x800 for the SP to which the command is issued Note that the log can be saved to a file and Event Codes gt 0x900 further filtering can then be done using Event Codes gt 0xA00 HP UX commands such as grep Clear Unsolicited Log N A Clears all Unsolicited Event Log information for the SP to which the command is issued Info Capacity N A Displays capacity information for LUNs and disk modules on the SCSI disk array See Figure 7 11 Inquiry N A Displays inquiry information for LUNs and disk modules on the SCSI disk array See Figure 7 12 1 0 Read Disks N A Reads and optionally displays data from specified disk modules on the SCSI disk array See Figure 7 13 Read LUNs N A Reads and optionally displays data from specified LUNs on the SCSI disk array See Figure 7 16 Tests Test Unit Ready N A Performs unit ready testing of LUNs and disk modules See Figure 7 19 Verify Disks N A Performs verification of accessible disk modules See Figure 7 20 Verify LUNs N A Performs verification of accessible LUNs See Figure 7 22 7 84 Support Tools Manager STM Table 7 12 Expert Tool Menus and Descriptions cont d Menu items Utilities Option Property Description Bind LUN N
180. O OS SOOOOOOOOC OSS fe SG GG OOOOCOOCOOOOOOOOOCOOCOOCO OOOOOOOOOOCOOOOOOOOOO h Figure 1 10 Power Supply Unit Battery Backup Unit The BBU is required for disk array caching During a power outage the BBU allows the SP to write cache to disk before the disk array shuts down The BBU can provide up to 60 seconds of DC power which is sufficient for the SP to empty its cache by writing data from the cache to a reserved area on a cache vault drive To conserve power the SP shuts down all unnecessary disk modules while emptying the cache Only cache vault drives remain active when cache is dumping Once cache data is safely stored on the cache vault disks the SP disables the BBU to preserve the life of its About Disk Array Components 1 15 uonduioseg jonpold internal power cells When power returns the BBU recharges its cells automatically It can fully recharge in less than one hour If it does not recharge in this time the BBU faults and must be replaced Cache cannot enable until the BBU is fully charged See Disk Array Caching later in this chapter for more information The model 10 disk array uses the HP A3233A BBU the model 20 uses the HP A3254A BBU BBUs are not interchangeable between the model 10 and model 20 The batteries in replacement BBUs may discharge to a fairly low level when they are in storage It can take up to two hours to charge these batteries If they do not charge within 120 minutes the BBU wi
181. O Presentation Utility PROM Revision OK OK A ENA ENA BIN BIN BIN B ENA ENA BIN BIN OK 0 0 1 1 L OK Figure 8 9 Model 10 Mapping Microcode 8 20 8 22 Troubleshooting the Disk Array Microcode Revision 07 23 Statistics Logging ENABLED Peer SP REMOVED Requests Completed 6K Cache State DISABLED Microcode Revision 8 20 Statistics Logging ENABLED Requests Completed 6K Cache States Write ENABLED SP SCSI ID 0 Presentation Utility PROM Revision Microcode Revision 7 99 8 20 FAN A OK Statistics Logging FAN B OK ENABLED ENA ENA BIN BIN BIN ENA BIN BIN 0 0 1 1 1 OK OK 2 2 3 Requests Completed 6K Cache States Write ENABLED Read DISABLED Figure 8 10 Model 20 Mapping Microcode 8 20 To determine how the failed disk is configured use the up down left and right arrow keys to highlight the failed disk slot and press RETURN This displays information similar to the following RAID 5 information Performance Statistics for disk module B4 Description 2 1 Gb SCSI Disk State REMOVED but bound Unit number hex 2 Unit type RAID 5 Group Individual access array Sense key additional sense code and additional sense code qualifier information can be found under the Performance Statistics for disk module XX screen Table 8 7 contains a list of the commonly reported sense key additional
182. OO OOOO OOO OOO OOOO AOA TAIBO Cache Vault Drives AO A1 A2 A3 A4 Figure 1 2 Model 10 Disk Array Chassis 1 4 About the Disk Array ptional Control Storage processor Control SP processor I I I I I I I I Cache 1 SP I I I I I I I I ra raTra1ra Lr gr gr gg g g Lr gr gr gg gg dg LJ LJ LJ LJ ra ra rara Lr gr gr gg g g Lr gr gr gg g dg LJ LJ LJ LJ Power supply Optional power supply l l Figure 1 3 Model 10 Internal Hardware Architecture Disk Array Chassis Model 20 The model 20 disk array chassis Figure 1 4 contains the compartments for disk modules SPs power supply units and BBU The fan pack is attached to the rear of the chassis The disk modules face front The SP s power supply units BBU and fan pack are accessible from the back Each disk module slot has a slot identification that corresponds to an internal synchronous SE SCSI 2 bus The model 20 disk array has five internal synchronous SE SCSI 2 buses identified by letters A through E Four disk modules can be installed on each bus Bus slot IDs are About the Disk Array 1 5 uonduioseg jonpold 1 6 Bus A disk modules AO A1 A2 A3 Bus B disk modules BO B1 B2 B3 Bus C disk modules CO C1 C2 C3 Bus D disk modules DO D1 D2 D3 Bus E disk modules E0 E1 E2
183. P A and SP B 3 9 m 13 ft The SP to SP external jumper piggyback cable is 0 5 m 19 7 in and is not included in the lengths shown in Table 2 3 If this cable is used this measurement must be added to the cable length to determine the total length of the SCSI 2 bus Determine Disk Array SCSI 2 Bus Lengths 2 21 uonejjeisuj Table 2 4 Cables and Terminators 2 22 Determine Disk Array SCSI 2 Bus Lengths Product or part number Description Length 8120 6147 both 68 pin high density connectors with thumb screws on both ends 0 5m models SP to SP piggyback cable A1658 62024 Differential terminator N A A1658 62018 68 pin high density connectors with thumb screws on both ends 0 9 m A1658 62020 68 pin high density connectors with thumb screws on both ends 2 5m A1658 62021 68 pin high density connectors with thumb screws on both ends 5 0m A1658 62022 68 pin high density connectors with thumb screws on both ends 10 0m 5061 5003 HP RS 232 cable 3m 17355D HP MUX box to RS 232 array cable 1 per SP Series 800 3m also connects a disk array s RS 232 port to a DTC RS 232 port on the HP 3000 Series 900 24542H HP Core I O Connector to RS 232 1 per SP Series 700 3m 5181 8686 2 0m V cable 68 pin HD male 2m Determine the Locations of SCSI 2 Bus Connectors The disk array can have one or two SPs SP A and SP B If the disk array has one SP it must be installed in slot SP A The external Fast Wid
184. P must be mirrored that is each SP must have equal amounts of cache Orders for cache SIMMs part numbers A3234A or A3235A must be made according to the following Model 10 20 SIMM Removal Replacement 9 25 jueuie2e deH pue jeAoway Model 10 Model 20 8 Mbyte 4x4 Mbyte 8 Mbyte 4x 4 Mbyte 4 A3235A SIMMs 16 Mbyte 8x4 Mbyte 16 Mbyte 8 x 4 Mbyte 8 A3235A SIMMs 4 or 8 of A3235A 32 Mbyte 4 x 16 Mbyte 4 A3234A SIMMs 64 Mbyte 8 x 16 Mbyte 8 A3234A SIMMs _ SIMM 1 Socket _ SIMM2 i Socket SIMM3 lt Socket 4 SIMM4 gt Edge of SP board Figure 9 14 SIMM Locations for the SP 9 26 Model 10 20 SIMM Removal Replacement Second Pair of Connectors nnnannnnn nnnnnnnnn nnnnnnnnn m Retaining clips Figure 9 16 SIMM Removal Model 10 20 SIMM Removal Replacement 9 27 jueuie2e deH pue jeAoway Figure 9 17 SIMM Installation on an SP 9 28 Model 10 20 SIMM Removal Replacement Model 10 Deskside Component Removal The deskside models have front and side covers The covers should be removed only to access to the internal parts of the disk array Outer Covers Figure 9 18 1 Remove the front panel a Lift from the bottom b Pull up and away 2 Remove the top cover a Remove the two front screws b Slide top 1 inch toward rear c Lift up and away 3 Remove the sides a Lift side up approximately 1 inch to release the hubs from the keyhole slots b Remove
185. Ps PS PS Ps Ps Ps Ps PS Ps Ps PS PS Ps PS Ps PS PS Ps PS PS PS PS PS PS PS PS PSP DISKS Fru Vendor Product ID AprxCap Serial Number Rev RO IBM DGHSOSY CLAROS 8 8gb 130054DC HP1A BO IBM DGHSOSY CLRROS 8 89b 13004708 HP1A C IBM DGHSOSY CLRROS 8 89b 13008582 HP1A DO IBM DGHSOSY CLAROS 8 89b 13005460 HP1A EO IBM DGHSOSY CLAROS 8 8gb 1300D30B HP1A Al IBM DCHSOSF CLAROS 8 8gb 58051838 HP24 Bi IBM DCHSOSF CLAROS 8 8gb 13199905 HP24 C1 IBM DGHS18Y CLRR18 17 8gb 13006831 HP1A D1 IBM DGHSi8Y CLRR18 17 8gb 13007072 HP1A E1 IBM DGHSi8Y CLRR18 17 8gb 13007350 HP1A A2 SEAGATE ST15150N 4 2gb 00218399 HPO3 B2 SEAGATE ST15150N 4 2gb 01668550 HPO3 C2 SEAGATE ST15150N 4 2gb 01446683 HPOS I2 SEAGATE ST32550N 2 1gb 00882323 HPOS E2 SEAGATE ST32550N 2 1gb 04352880 HPOS A3 IBM DCHSO4F CLARO4 4 2gb 68083144 HP24 Weoovovuvuvuu vu vu AAAA AAAA AAAA AAAA AAAA AAAA AAAA AAAA AAAA Figure 7 30 Information Tool Log Screen Disks Section For more information on a warning incomplete or failure status refer to the tool Activity Log Access this screen from xstm s main menu by selecting Tools Information gt Activity Log 7 106 Support Tools Manager STM amp hpterm CHE x Information Tool Activity Log for C2300 Array on path 874 9 W salt rose hp com 15 43 214 34 Information Tool Activity Log for C2300 Array on path 874 9 Log creation time Wed Apr 29 10 53 46 1998 Wed Apr 29 10 59 46 1998 Information tool
186. Remove the outer covers Figure 9 18 Remove the following a Fan pack Figure 9 6 b Power supplies or fillers Figure 9 11 c SPs or fillers Figure 9 9 d IPCB Figure 9 21 e All disk modules or fillers Figure 9 3 f BBU Figure 9 13 Turn the disk array over onto its top a Remove the wheel base i Remove the retaining screws from the front of the wheel base ii Slide wheel base 1 inch toward back and remove Remove the inner right metal panel 11 screws Figure 9 19 Remove the inner chassis Lay the chassis onto its left side a Remove the six retaining screws from the top b Lay the chassis onto the covered side c Remove the seven retaining screws from the bottom d Lift the inner chassis up and out of the system i Rotate upward the rear of the inner chassis about 3 inches ii Lift up and out Cut the cable tie securing the AC cable to the chassis Model 10 Deskside Component Removal 9 43 jueuie2e deH pue jeAoway 8 Disconnect the associated spade lug wires one green neutral wire and remove the AC cable Bracket retaining screw Oneat X each end AC harness Figure 9 25 Removing the AC Harness Bracket 9 44 Model 10 Deskside Component Removal Figure 9 26 Removing the AC Harness Plugs Model 10 Deskside Component Removal 9 45 jueuie2e deH pue jeAoway On Off Switch Figure 9 27 1 Power off the disk array and disc
187. Removed the failed SP Remove the functional SP install it in the slot from which the failed SP was removed and re enable it Unmount all logical volumes file systems associated with the path through the failed SP umount dev vg0n lvolx Deactivate all volume groups associated with the path through the failed SP vgchange a n dev vg0n Using LVM remove from the system tables all information about the volume groups associated with the path through the failed SP vgexport m mapfile dev vg0n Disable the functional SP remove it from the failed SP s slot install it in its original slot and re enable it Using Grid Manager change the SP ownership of the failed SP s LUNs to the ownership of the remaining functional SP by choosing the following menu options 5 Change Parameters 1 Change Physical Unit Parameters 1 Change Default Ownership of LUN Troubleshooting the Disk Array 10 Power cycle the disk array 11 Reboot the host system 12 Recreate the directories and nodes for the volume group s associated with the path through the failed SP mkdir dev vg0n mknod dev vg0n group c 64 0x0n0000 vgimport m mapfile dev vgO0n dev dsk cnd0s2 vgchange a y dev vg0n mount dev vgO0n 1voll mount directory When a new SP becomes available 1 Install the new SP in the failed SP s original slot 2 Unmount all LUNs whose ownership is to be changed to the ownership of the replacement SP 3 Deactivate all vol
188. SCSI SP The SP620 SCSI SP is enabled and disabled using the Grid Manager utility The enable disable feature does not need to be used when a failed SP620 SCSI SP is removed and replaced See Enabling Disabling the SP620 SCSI Storage Processor in chapter 6 for more information For information about upgrading from an SP to the higher performance SP620 SCSI SP see Upgrading from an SP to a SP620 SCSI SP in chapter 8 About Disk Array Components 1 11 uonduioseg jonpold ambe x 4 4 memory slots Q Q Ready light Q green CM Q Enable Disable Cie switch Sette AQ QA Latch IN DN EN Figure 1 7 Storage Control Processor Printed circuit Memory board modules Service light amber EM Ready light cay tont f a T Figure 1 8 Model SP620 SCSI Storage Control Processor 1 12 About Disk Array Components Fan Pack The fan pack is located on the rear of the chassis It pulls air through the disk array for front to back air flow The fan pack contains three fans in the model 10 and six fans arranged in two groups of three in the model 20 One fan from each group can fail and the disk array will still run because the other fans will increase their speed automatically from 18 VDC to 26 VDC to continue to cool the array However if a second fan from either group fails the disk array will shut down A failed fan pack must be replaced as soon as possib
189. SI2 Bus Lengths 00 0 0 eee eee ee 2 21 Determine the Locations of SCSI 2 Bus Connectors 265 2 23 Connect the Disk Array s to the Host Computer 00 00 e eee 2 25 Connect the SCSI 2 Cables to a Single Host 0 0 2 0 00 005 2 25 Connect the SCSI 2 Cables in Two Host Configurations 2 27 Set SCSI IDs for SPs in the Disk Array s 0 ee eee cee ee eee 2 29 Connect Asynchronous Console Cables 00 0 cece eee eee eee 2 33 Direct Connect Terminal Configuration 00 02 eee eee eee 2 34 Terminal Emulation through a MUX on HP UX 2 0 5 2 35 Set up the Disk Array on a Series 800 MUX Port 0204 2 36 Install MUX Card on a Series 800 System T500 or K Series 2 36 Set Up the Disk Array on a Series 700 MUX Port 205 2 37 Set Up the Disk Array on an HP 3000 Series 900 DTC 2 38 Verify Array Connectivity llle 2 39 Connect the Power Cord 00 cece eee een e 2 42 Power Up the Disk Array 0 ec eee III 2 43 Determine Terminal Connectivity 00 eee eee eee ee 2 43 Disk Array Installation Troubleshooting Checklist Revision 1 3 2 44 3 Configuration HP 9000 Series 700 Disk Array Support Requirements for HP UX Release 9 05 9 07 3 2 Installing and Reconfiguring on S700 Systems 0 0 00 e eee
190. SOH WRT MT Exclusive destructive Example HARRAY gt wtr mt RANDOM MEDIA TEST UTILITY 4 47 PM E US Enter If you continue this command WILL overwrite data on the disk Before proceeding ENSURE that disk does NOT contain needed data Continue Y N N y The test is performed by executing a sequence of reads or writes The length of each transaction can be Randomly selected lengths User specified length the transaction length type R U R r The sources of the data pattern to be used are U User specified pattern EB Random pattern Which pattern source would you like U R R u Enter the pattern in hex abcd Do you want to continue on errors Y N N y Enter the loop count 1 255 1 1 RANDOM MEDIA TEST beginning MON APR 29 1996 4 47 PM RANDOM Test Pass Count 1 4 47 PM The Random Media Test is 25 complete 4 47 PM The Random Media Test is 50 complete 4 48 PM The Random Media Test is 75 complete 4 48 PM The Random Media Test is 100 complete RANDOM MEDIA TEST UTILITY COMPLETED 4 48 PM WTR MT command with another set of options HARRAY gt wtr mt RANDOM MEDIA TEST UTILITY 4 48 PM 7 44 HARRAY Section 17 External Exerciser If you continue this command WILL overwrite data on the disk Before proceeding ENSURE that disk does NOT contain needed data Continue Y N N y The test is performed by executing a sequence of
191. Tests Verify Disks Screen 0 0 cece eee eee 7 92 Tests Verify Disks Select Blocks Screen 00 cece eee ee eee 7 93 Tests Verify LUNs Screen 02 0 0 cece cece nh 7 94 Tests Verify LUNs Select Blocks Screen 0 0 0 eee eee ee ee eee 7 95 Utilities System Configuration System Cache Screen 1 96 Utilities System Configuration System Memory Screen 7 97 Utilities System Configuration gt Host I F Settings Screen 7 98 mstm Device Map and Menu 0 eee eee eee eee ene eens 7 101 Information Tool Log Screen Map Section 00 0 0 e eee ee eee 7 102 Information Tool Log Screen Configuration Section 7 104 Information Tool Log Screen Disks Section 002 2 e ee 7 106 Activity Log Screen 5 os ese see aa eas ER pec IG ee pees 7 107 Firmware Update Tool Top Level Menu 0 cece eee eens 7 108 Current Revisions Display 0 cece eee eee teen eee 7 110 Expert Tool Top Level Menu seeseeeeeee II 7 112 Info 2 Capacity Screen 0 cece eh eee 7 115 Info 2 Inquiry Screen 0 cece ehh 7 115 I O Read Disks Screen sseeeeeeeeeeeeeee eh 1 116 VO Read Disks Select Blocks Screen 0 00 ccc eee ene 7 117 7 39 7 40 7 41 7 42 7 43 7 44 7 45 7 46 7 47 7 48 7 49 7 50 8 1 8 2 8 3
192. The SP has aborted a Rebuild Equalize operation 0x616 Flare Installed Extended status is revision information This message is logged once when new flare code is installed Unsolicited Event Log 8 57 punoouse qnouj Table 8 10 Informational Codes cont d Event code Description 0x617 Code Installed to Drive Extended status is revision information This message is logged once for each updated drive when new drive code is installed Note that the revision information for drives consists of 4 ASCII characters for example code 0123 will appear as 0x30313233 0x621 Background Verify Started A background check on the correctness of parity information has begun on the unit of which this drive is a member 0x622 Background Verify Complete Disk array has completed background checkpoint verification of the accuracy and completeness of the parity check data in a RAID 5 RAID 3 RAID 1 or RAID 1 0 LUN 0x623 Fan Pack Disable Door Open Fans disabled because the fan pack door was opened The disk array shuts down if the fans are not operational within two minutes 0x624 Fan Pack Enable Door Close Fans were enabled after the fan pack door was closed 0x630 Fan Installed Disk array has detected that a fan has been installed or replaced 0x631 VSC Installed Disk array has detected that a VSC has been installed or replaced 0x632 AC Box Installed An AC Box has been ins
193. UL Read Disks command CANCELED per user request Read of DISK RB beginning blocks 36865 thru 36991 126 cmd max Read of DISK RB completed blocks 36865 thru 36991 B errors detected Read Disks command SUCCESSFUL 1 disks processed errors detected Read of LUN beginning blocks thru 127 127 cmd max Read of LUN B completed blocks thru 127 B errors detected Read Luns command SUCCESSFUL 1 luns processed errors detected Test Unit Ready command SUCCESSFUL Verify Verify Verify Verify Verify Verify A Info I 0 Tests Utilities Help of DISK AG beginning blocks thru 32768 32768 cmd max of DISK AB completed blocks thru 32768 errors detected Disks command SUCCESSFUL 1 disks processed errors detected of LUN beginning blocks thru 65535 65535 cmd max of LUN completed blocks thru 65535 B errors detected Luns command SUCCESSFUL 1 luns processed errors detected p Interrupt Abort Figure 7 10 Expert Tool Top Level Menu Support Tools Manager STM 7 83 sonsoubeiq pue seninf ISOH The properties for each menu are documented in Table 7 12 The section following the table contains example screens showing the most common Expert Tool features Table 7 12 Expert Tool Menus and Descriptions Menu items Option Property Description File Save As
194. UMBER 18 FRU Slot Number A3 Enter FRU NUMBER for testing or 99 to Cancel 6 MECH UTILITY 4 20 PM Sending Test Unit Ready Command to FRU Number 6 Test Unit Ready was SUCCESSFUL MECH UTILITY COMPLETED 4 20 PM Example 2 MECH command with option 1 HARRAY gt mech The following commands are available at the mech level 0 Test Unit Ready TUR 1 Inquiry 2 Read Capacity 3 Read 4 Read Media Test RO MT HARRAY Section 17 External Exerciser 7 31 sonsouDeig pue seni ISOH 5 Verify via scsi commands 99 Exit Enter number 0 5 or 99 to Cancel 4 gt 1 The following mechs FRU s are available for test FRU NUMBER 0 FRU Slot Number A0 FRU NUMBER 6 FRU Slot Number A1 FRU NUMBER 12 FRU Slot Number A2 FRU NUMBER 18 FRU Slot Number A3 Enter FRU NUMBER for testing or 99 to Cancel gt 12 MECH UTILITY 4 20 PM Sending Inquiry Command to FRU Number 12 INQUIRY INFORMATION Vendor Identification SEAGATE Product Identification ST12550N Product Revision HP03 Inquiry Command SUCCESSFUL MECH UTILITY COMPLETED 4 20 PM Example 3 MECH command with option 2 HARRAY gt mech The following commands are available at the mech level 0 Test Unit Ready TUR 1 Inquiry 2 Read Capacity 3 Read 4 Read Media Test RO MT 5 Verify via scsi commands 99 Exit Enter number 0 5 or 99 to Cancel 4 gt 2 The following mechs FRU s are available for test FRU NUMB
195. UR 1 Inquiry 2 Read Capacity 3 Read 4 Read Media Test RO MT 5 Verify via scsi commands 99 Exit Enter number 0 5 or 99 to Cancel 4 gt 4 The following mechs FRU s are available for test FRU NUMBER 0 FRU Slot Number A0 FRU NUMBER 6 FRU Slot Number A1 FRU NUMBER 12 FRU Slot Number A2 FRU NUMBER 18 FRU Slot Number A3 HARRAY Section 17 External Exerciser 7 33 sonsouBeiq pue seninf ISOH Enter FRU NUMBER for testing or 99 to Cancel gt 12 MECH UTILITY 4 21 PM A Read Only Media Test will be performed on FRU 12 Types of Media Tests 1 Selected Area 2 Random Area Enter Test Type 1 2 2 1 Enter Starting Block Address 0 4123241 0 500 Enter Ending Block Address 500 4123241 4123241 505 Transfer Length Types 1 Random Transfer Length 2 User Transfer Length Enter Transfer Type 1 2 1 gt 1 Enter Loop Count 1 256 1 gt 1 MECH Read Only Media Test in progress please wait MECH Loop Count 1 Completed MECH Read Only Media Test Successful Example 6 MECH command with option 4 sub option 2 HARRAY gt mech The following commands are available at the mech level 0 Test Unit Ready TUR 1 Inquiry 2 Read Capacity 3 Read 4 Read Media Test RO MT 5 Verify via scsi commands 99 Exit Enter number 0 5 or 99 to Cancel 4 gt 4 The following mechs FRU s are available for test FRU NUMBER 0 FRU Slot Number A0 FRU NUMBER 6 FRU Slo
196. VDC model 20 which is supplied to the power bus on the midplane Two power supply units can be configured in the model 10 Up to three power supply 8 40 Power Block Diagram units can be configured in the model 20 depending on the power load of the disk drive modules The power supplies can be removed without removing power e The disk array s major power bus is on the midplane The midplane distributes 24 VDC power from the power supply units to the disk modules power regulator board the disk modules are connected directly to the midplane and to the fan pack by a cable e The fan pack provides cooling for the array It receives 18 VDC power from a cable connected to the midplane The fan pack provides a front to rear airflow The model 10 fan pack has three fans two are needed to cool the array The model 20 fan pack contains six fans five are needed to cool the array The fan pack can operate and cool the array even if one of the fans fail If one fan fails the other two fans in the model 10 and the other five fans in the model 20 speed up and provide the array with the needed cooling CAUTION When replacing the fan pack the array can continue normal operation for about two minutes before it reports an overtemperature fault and starts to shut itself down Maintenance performed on the disk array with the fan pack opened or when replacing the fan pack must be performed within this two minute margin e The power regulator
197. When completed the available memory should be 0 Mbytes 8 38 Cache Configuration Instructions Re enable cache by following this menu path from the Grid Manager Main menu 5 Change Parameters 2 Change Storage System Parameters 1 Change Storage System Caching Parameters a Type Y to the Change caching parameters N prompt b Type E for enable to the next prompt to enable caching Enable LUN caching by following this path from the Grid Manager Main menu 5 Change Parameters 1 Change Logical Unit Parameters gt 4 Change Unit Caching Parameters a Enter the LUN number at the Physical unit to change hex prompt b If caching is not enabled type Y to the Change caching parameters N prompt c Type Y to the Reset caching parameters to default values Y prompt Repeat this step for all LUNs that are bound on the array Cache Configuration Instructions 8 39 punoouse qnouj Power Block Diagram midplane power supply AC power _AC A in in regulator power I Hi Biy E power supply C Figure 8 12 Array Power Block Diagram Disk array power consists of the AC power switch the power supply units the midplane the fan pack and the regulator board of the disk drive modules e The AC power switch is used to turn the AC power on and off to the array e The power supply units take the AC power and convert it into 18 24 VDC model 10 and 18 24 48
198. a to the second SP The second SP would panic with a 0x00000010 code as it tried to dump cache due to the panic of its peer The result was a failure to properly dump cache LUNs that had data in cache at the time of the panic became accessible Fixed in microcode revision 8 20 0x00000011 Instruction RAM parity error The hardware detected an unrecovered error in the instruction RAM area Note that the instruction RAM is not located on the SIMMs 0x00000012 Data RAM parity error The hardware has detected an unrecoverable error in the data RAM area This corresponds to the SIMMs and can occur if the SIMMs are not seated properly or have been handled without following proper ESD procedures 0x00000015 Memory exhausted 0x00000016 Memory exhausted 0x00100006 Degraded LUN experiencing heavy I O Reported on a very large degraded RAID 5 LUN usually with more than eight disks under heavy I O conditions Fixed in microcode revision 8 20 See also panic 0x00100000 0x00100017 Multiple trespasses Could be caused by multiple trespasses taking place during a rebuild The trespass disabled the SP which caused the rebuild to abort After this happens several times the array panics Avoid trespass during a rebuild Fixed in microcode revisions 9 xx 0x00220011 Cache dump conflicts with trespass If a write cache fault dump conflicted with an attempted trespass of a dirty unit the microcode would interm
199. abase Sunc Error 0x03 5 04 24 97 16 46 08 IB Oxa08 Database Sunc Error 0x03 B 04 24 97 16 46 08 AS Oxa08 Database Sync Error 0x03 7 04 24 97 16 48 31 IB Oxa08 Database Sync Error 0x03 8 04 24 97 16 48 31 AS Oxa08 Database Sunc Error 0x03 9 04 24 97 17 03 10 IB Oxa08 Database Sunc Error 0x03 10 04 24 97 17 03 10 A3 Oxa08 Database Sync Error 0x03 11 04 24 97 17 05 34 D3 Oxa08 Database Sync Error 0x03 12 04 24 97 17 05 35 A3 Oxa08 Database Sync Error 0x03 13 04 24 97 17 19 16 03 Oxa08 Database Sync Error 0x03 14 04 24 97 17 19 32 A3 Oxa08 Database Sync Error 0x03 Enter for HELP to GO TO HAIN HENU Enter Choice j sa There are S2 more lines in this display Figure 8 14 Grid Manager Unsolicited Event Log To navigate the Unsolicited Event Log type the following single letter commands Key Action Norn Scrolls to the next screen of entries Porp Scrolls to the previous screen of entries Forf Displays the first screen of entries Lorl Displays the last screen of entries Aora Defines a filter to select event types by FRU or error code For help type A then A filter stays in effect until changed Corc Clears all entries from the Unsolicited Event Log Caution The host computer relies on entries to report subsystem errors clearing the Log may remove recent entries before they reach the host Viewing Maintenance Information 8 45 Bunoouse
200. ach SP maintains a mirror image of the other SP s caches The mirror image of the cache data is stored on the cache vault drives If one SP fails the cache data is available from the other SP s mirror image Caching can enhance disk array performance and save time in two ways Read request If a read request is issued for data that is already in the read cache the disk array avoids accessing the disk modules to retrieve the data Write request If a write request modifies data in the write cache the SP can keep the modified data in the cache and write it back to disk at the most expedient time instead of immediately after the write Disk Array High Availability Features 1 19 uonduioseg jonpold In the event of an unexpected disk array shutdown the disk array switches over to BBU power When drawing power from the BBU the disk array has about 60 seconds to power down in an orderly way During this time the disk array dumps its cache that is it writes any data in its cache memory to the cache vault drives Cache is also dumped by the peer SP if an SP fails Cache is mirrored so storage system caching is disabled until the failed SP is replaced The cache image is not dumped however when both SPs are removed from the disk array simultaneously If the cache image is lost the data integrity on the disk array is compromised To avoid this problem never remove both SPs from the disk array simultaneously In addition to disk array
201. ache Parameters At any prompt enter to GO BACK TO PREVIOUS MENU or for help Current storage system cache parameters Cache State ENABLED Cache Page Size 2 KB SP Memory 16 MB Cache size 16 MB Change caching parameters N Figure 6 13 Change Storage System Cache Parameters Screen Enable To change the cache state from the current state type Y for yes to the Change caching parameters prompt The following prompt will appear Cache State Enable Disable Enable Answer E or D to specify the other state If changing the cache state from enabled to disabled Grid Manager displays the previous menu Cache parameters can be changed only when caching is disabled This may take a few seconds since the cache is not disabled until all modified pages are written to disk To determine whether caching is disabled display the Presentation Utility screen and wait for the Cache State display to change to DISABLED When DISABLED appears other disk array caching parameters can be changed as described in the next section Setting up Disk Array Caching for Microcode Revision 7 23 6 55 Jobeue N pus To set up a new disk array or upgrade the disk array to support caching and disk array caching is disabled continue to Specifying Cache Page Size and Cache Size which follows If changing the cache state from disabled to enabled Determine whether disk array caching is enabled display the Presentation Utility screen an
202. addition to the above the disk drives in slots A2 and A4 model 10 and DO and EO model 20 cannot be bound as hot spares if cache is installed because they are cache vault drives Hot spares should never be assigned LUN number 0 Hot spares are assigned a LUN number but are not recognized as an addressable LUN by the host Note The operating system recognizes LUNs 0 through 7 only To Bind a Logical Disk Unit 1 Select the SP to which the disks are to be bound by accessing the array through the appropriate RS 232 port SP A or SP B 2 Access the Grid Manager Main menu Then type caret and press RETURN until the Grid Manager Main menu appears 3 From the Grid Manager Main menu choose the Bind Logical Units option Grid Manager displays a Binding Recommendations screen Press any key to continue to the Bind Logical Units screen as shown in Figure 6 10 4 Choose one of the binding options and press RETURN Creating Logical Disk Units LUNs Jobeue N pus e xterm SP A SP SCSI ID 15 Bind Logical Units Henu S N 9621954 1 RAIDS Group Individual Access Array 2 RAID 3 Group Parallel Access Array 3 RAID Hirrored Pair 5 RAID 1 0 Group Mirrored RAID 0 Group 7 Hot Spare Figure 6 10 Bind Logical Units Menu Microcode revision 9 xx 5 The first available unbound UNB disk is highlighted as shown in Figure 6 11 Press RETURN to select the disk for binding If configuring a hot spare or
203. ages to the LUNs and the UDP number returns to 096 When disk array caching is enabled and SP A SP B accesses a LUN with caching enabled the ADP BDP number should change 6 40 Presentation Utility Screen Moving the Grid Cursor To select the modules for binding move the grid cursor to the boxes representing the modules to be bound Later when the disk array is running normally the Presentation Utility can be used to display read write statistics on any disk module by moving the cursor to the box representing the module and clicking on the module Table 6 3 demonstrates how to move the grid cursor If the arrow keys do not access the desired box use the space bar to move the cursor through the screen Table 6 3 Cursor Movement Keys Key H h or left arrow Function Moves cursor left one box J j or down arrow Moves cursor down one box K k or up arrow Moves cursor up one box L or right arrow Moves cursor right one box Space bar Moves the cursor right one box From the rightmost box in any row except the bottom row moves the cursor to the leftmost box in the next row From the rightmost box in the bottom row returns the cursor to the top left box Also moves the cursor from one unbound disk module to the next unbound disk module when binding and from one bound LUN to the next bound LUN when unbinding Grid Manager Procedures The remainder of this chapter descri
204. al Figure 5 1 Model 10 20 Basic Configuration Basic Configuration for Model 10 and Model 20 Disk Arrays 5 5 006 Seles 000 dH UOHeunByuoD Table 5 2 Basic Configuration Error Recovery Failing Continue after component failure What happens and how to recover Disk module Yes With RAID of any level supported by MPE ix applications continue running System Administrator or service provider replaces module SP No Disk array fails System Administrator or service provider replaces SP and the System Administrator restarts operating system Fan pack Yes Applications continue running System Administrator or service provider replaces module Power supply Yes If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply The System Administrator restarts the operating system if necessary SCSI 2 adapter No I O operations to array disk units fail Authorized service provider replaces adapter and System Administrator restarts operating system and applications SCSI 2 cable No I O operations to array disk units fail Authorized service provider replaces cable and System Administrator restarts operating system and applications 5 6 Basic Configuration for Model 10 and Model 20 Disk Arrays Two SP Configuration The two SP configuration shown in Figure 5 2 has one host with one SCSI 2 adap
205. al disks are grouped into a storage unit by binding disk modules together into a physical disk unit Three of the RAID groups defined in this section use disk striping two use mirroring Planning Disk Unit Configurations 1 23 uonduioseg jonpold CAUTION Disk modules having different storage capacities or different model numbers should not be bound into the same RAID group Disk Striping Disk striping is a technique in which data is read from and written to uniformly sized segments block stripes across the multiple disk modules in a RAID group simultaneously and independently By allowing multiple sets of read write heads to work on the same task at once disk striping can enhance performance The amount of information read from or written to each module makes up the block stripe element size for example 128 sectors The stripe size is the number of data disks in a RAID group multiplied by the stripe element size For example if the stripe element size is 128 sectors in a five disk RAID 5 group the stripe size is 640 sectors or 128 x 5 More information on stripe size is provided in RAID 5 Group Individual Access Array later in this chapter Disk striping can enhance performance by allowing multiple sets of read write heads to work on the same task at once Hardware disk striping can be implemented by configuring the disk modules as a RAID 0 group RAID 1 0 group or a RAID 5 group Mirroring Mirroring maintains a duplicate c
206. an Shows Two LUN 0s for Disk Array on HP UX 9 0x When ioscan is run on HP UX 9 0x two LUN Os for the disk array will show if two SPs are installed This is a function of HP UX 9 0x and is not related to auto trespass being on or off For example ioscan fC disk Class LU H W Path Driver H W Status S W Status disk 0 8 5 0 scsi3 target disc3 ok 0x20650202 ok disk 1 8 6 0 scsi3 target disc3 ok 0x650202 ok disk 2 8 Gre scsi3 target disc3 ok 0x650202 ok Changing Storage System Parameters 6 67 Jobeue N pus LUN 0 is assigned two LU values LU numbers 0 and 1 If the diskinfo command is used on LU 0 and LU 1 it can be seen that one of the LUs has a capacity of 0 Kbytes The LU that has 0 Kbytes capacity does not own the LUN as the following two diskinfo examples illustrate Example 1 diskinfo dev rdsk c0d0s2 SCSI describe of dev rdsk c0d0s2 vendor DGC product id C2300WDR5 type direct access size 0 Kbytes This LU does not own the LUN bytes per sector 512 Example 2 diskinfo dev rdsk cld0s2 SCSI describe of dev rdsk c0d0s2 vendor DGC product id C2300WDR5 type direct access size 8146175 Kbytes This LU owns the LUN bytes per sector 512 Diskinfo shows that LU 0 has 0 Kbytes in this RAID 5 configuration Diskinfo shows that LU 1 has a size of 8146175 Kbytes This shows that the SP having SCSI address 6 H W path 8 6 0 owns LUN 0 Select H W path 8 6 0 when building a new volume group in SAM Default
207. an one disk at a time RAID 0 offers enhanced performance via simultaneous I O to different modules but does not intrinsically offer high availability For high availability the striped disks can be software mirrored RAID 1 0 RAID 1 0 mirrored RAID 0 Four six eight ten twelve fourteen or sixteen disk modules bound as mirrored RAID 0 groups These disk modules make up two mirror images of two three four or five modules each A RAID 1 0 group combines the speed advantage of RAID 0 with the redundancy advantage of mirroring RAID 1 Even numbers of mirrored disk modules RAID 5 Three to sixteen disk modules that use disk striping in which the hardware writes to or reads from multiple modules simultaneously with high availability provided by parity information on each module The ideal number of disk modules in a RAID 5 group is five replace under power hot pluggable The disk array provides replace under power capability which means that a FRU for example a disk module or a fan pack can be replaced without powering down the disk G 4 Glossary array If redundant components are present applications continue to run while the failed component is replaced SCSI or SCSI 2 bus A bus that complies with the SCSI small computer system interface standard In the context of a disk array the SCSI 2 bus connects the SCSI 2 adapter in the host to an SP in the storage system chassis The SP s in a model 10 disk array div
208. anged If SP A is the default owner of LUN 0 and SP B is the default owner of LUN 1 the primary path was configured to LUN 0 through SP B and the alternate path was configured to LUN 1 through SP A Configuring the primary and alternate paths in this manner would force an auto trespass when a LUN is accessed If 0x906 errors occur check how the primary and alternate paths are configured under LVM Make sure the array is not forced to do an auto trespass every time a LUN is accessed The main cause of 0x906 errors is incorrect configuration of the LUNs The most common error is that the LUNs are bound but split between the two SPs in the array Then the array LUNs are configured either manually or via SAM and they all end up associated on the primary path with only one SP None of the LUNs become associated with the second SP Therefore every time the second SP path is used which is how the LUN was bound it can create an auto trespass and a 0x906 error This shows up most often during system boot autoconfig polling the bus ioscan vgscan and inside SAM which also uses ioscan to do its bidding Sometimes Recovered Error Log Reporting is disabled to do away with most 0x906 errors This works to eliminate 0x906 errors and can be turned on and off without causing a reboot of the SPs However if the function is left turned off all error reporting also is turned off In other words HARAYMON is not able to report errors as it was designed to do
209. are available on the Series 700 Note There are port restrictions on certain computers If using a J2085A and the MDP port of a CPU ports 0 and 7 are reserved on an 8x7 CPU ports 0 1 and 7 are reserved on a K Series CPU Additionally a Series 700 CPU has no expansion capability for a MUX and may have neither serial port available for use by the 24542H cable 2 Modify the usr lib uucp Devices file to specify the device files to be used and a baud rate of 9600 Note Ifthe uucp fileset has not been installed on the system it must be installed from an HP UX 9 04 update tape cd usr lib uucp Go to the uucp directory cp Devices Devices old Make a backup copy of the Devices file vi Devices Use the vi editor to edit the Devices file 3 Fora Series 800 add a line to the Devices file for each MUX port that is directly connected to the array via the 17355D cable For example if cables 17355D have been connected to ports 1 and 2 on the MUX box and the LU number for the MUX is 0 the following lines would be added to the Devices file Direct ttyOpl 9600 direct Direct tty0p2 9600 direct 4 ForaSeries 700 add a line to the Devices file for each serial port that is directly connected to the array via the 24542H cable For example if cables 24542H have been Configuring MUX Ports on HP 9000 Systems 6 3 Jobeue N pus 6 4 connected to the serial ports 1 and 2 or RS 232 A and RS 232 B the following lines woul
210. arica CI lr DNE LS IEEE JI Por Lun Tl nn Array Sera d 90 75325 n5n Firat Hock in res 0 ESDmi k 1 1 Laat block in read BISSOHBIIT Tocka per read coved 127 uz Hote Laat block in reed must ba e trat black ti raed ali pase hg aim z 15 43 714 Espert Teo lt Read Luns for devies C2300 Array ne path Bid g Array ava 0E7579 DH d5 7535 B5E Cope Dare BJ es oe Hrray Timas L153432 25 FH 8 55 81 1 73 08 Byrta lira 13 15 Bf Array SF A 1730 SPEI bjere Fan malt LUH B sara black d GI zI 5 a I a DIIISG Dlz3455703 200 338 3 3 3 BOR ea ee DID A o Bia iuasiuas garavasaes ama j vaninsmras DI4EARIRIRI ee s BARIRIBIRE PDAKAEARIRI ee Ome p omn apan amn amn ama Dam rmm ZEEESHBEZiur Figure 7 18 I O Read LUNs Output Screen 7 90 Support Tools Manager STM Tests Test Unit Ready amp Expert Tool Test Unit Ready for device C2300 Array on path 874 8 salt rose hp com 15 43 214 34 Expert Tool Test Unit Ready for device C2300 Array on path 8 4 8 Array Date 82 25 38 SN 95 7535 656 System Date 82 24 38 Array Time 17 42 59 FH 3 55 81 1 73 88 System Time 13 15 36 Array SP A 7385 SP620 System Name salt Note The array does NOT support Test Unit Ready commands to Hot Spare luns The command will fail 85 25 81 and Status will be set to EXCEPTION The array does NOT support Test Unit Ready commands to luns which are inaccessi
211. arness Bracket 0 00 cece cence eee ences 9 104 Removing the AC Harness Plugs 0 00 eee cece eee eee nee 9 105 On Off Switch and Power Plug Assembly 00 0 ee eee eee 9 106 TABLES 1 1 2 1 2 2 2 3 2 4 2 5 2 6 3 1 3 2 3 3 3 4 4 1 4 2 4 3 4 4 4 5 4 6 4 7 5 1 5 2 5 3 5 5 6 1 Replaceable Parts Model 10 0 0 ccc ee cee nee eens R 15 Replaceable Parts Model 20 0 ccc cece en R 18 Relative Performance Compared to a Single Disk 0065 1 41 llmqnmGCabinet i 2fkRlged9ix sf arbres has ea iat dul 2 10 1 6 m Cabinet iila Chisel a lb er erede oe GAN oes 2 11 Internal SCSI 2 Bus Lengths for Model 10 and Model 20 2 21 Cables and Terminators 1 0 20 0 eee eee ee 2 22 SCSI ID Switch Settings 0 0 0 cee e 2 31 MUX Components usce Regest nce ach ec eh REO x de es d e 2 36 Array Configurations scesi aeaea aa E a m I E a 3 5 Basic Configuration Error Recovery 02 e cee ee eee eee ee eens 3 7 Two SP Configuration Error Recovery 2 see e eee eee eee eee 3 9 Multi Bus Configuration Error Recovery 0020s eee cece eee eens 3 11 Array Configurations nessi ieat ena aeaa peda aaae po Ih Ir 4 2 Basic Configuration Error Recovery 0 2 eee e eee ee eee eee eens 4 6 Two SP Configuration Error Recovery eee ee eee eee tenes 4 8 Two Adapter Two SP Configuration Error
212. ata is stored in the disk array s cache memory where it is available for the next read request from the system without having to read from the disk again Troubleshooting with SAM 8 95 punoouse qnouj Enabling the read cache improves performance in applications that perform repeated successive reads such as when copying editing or processing files through a program with output going to a separate destination If applications that access the LUN being configured make mostly short random reads and if there is a low probability that the applications need to access the read cache for more data the advantage of enabling the read cache is less pronounced Bind a LUN Use this dialog window to bind one or more disks to a LUN Interpretation of the information in the screen see example in Figure 8 28 is as follows Used Device is part of an existing LUN that may or may not be in use Free Device is not part of an existing LUN Empty No disk is installed at that location N A The disk cannot be used for the selected bind type Example SAM gt Disks and File Systems gt Disk Devices gt Actions Disk Array Maintenance gt Bind LUN 8 96 Troubleshooting with SAM Bind LUN LUN Number 1 Maximum Rebuild Time 4 hours Select between 3 and 16 free disks AO Used Bind Type RAID 5 gt Al Used A2 Used Sectors per Stripe Element 128 A3 Used BO Free Enable Write Cache
213. ate the firmware file list File list created KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KK KKK KKK KKK ko ko kokok FIRMWARE FILES REVISED O1APR98 ck ck ck ck ck kc kk ck ck ck ck ck ck KK KKK ck ck ck KKK ck ck ck ck ck KKK KKK KKK KKK KKK KKK kk ck kk kk e ke e e Se e e e S e kk Sk kk kk kk kk Sk kk kk kk kk Sk kk kk kk Sk Sk Sk kk kk kk ck ck ck ck ck ck kv NIKE Firmware Files Found d KKKKKKKKKKKKK e S e Sk kk kk kk kk kk Sk kk kk kk ck kk kk kk kk kk kk kk kk Sk Sk Sk Sk Sk kk ke kk Sk Sk ck Sk ck ck Sk Sk kA A amp File name Intended Product Size SP 829 NIKE SP Controller Sauna Firmware 1310720 SP 925 00 NIKE SP Controller Sauna Firmware 1310720 SP 861 20 NIKE SP620 Controller Phoenix Firmware 1572864 SP 955 01 NIKE SP620 Controller Phoenix Firmware 1572864 NIKE2GHHO03 NIKE Disk FRU Firmware 269824 NIKE2GLP08 NIKE Disk FRU Firmware 269824 NIKE4GHH03 NIKE Disk FRU Firmware 269824 NIKE4GLP53 NIKE Disk FRU Firmware 335872 NIKE9GHH53 NIKE Disk FRU Firmware 335872 Legend File name name of the firmware file Intended Product firmware file s intended product Size exact byte size of the firmware image Enter the firmware file name will display files Default will quit SP 955 01 NIKEARRY 7 53 sonsouDeig pue seni ISOH KKKKKKKKKKK e e KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KK x NIKE Controllers Found E kk kk See e ke kk kk kk kk kk kk Sk kk kk kk Sk kk kk kk Sk
214. ation Auto Trespass Sets the auto trespass state on or off BBU Test Time Specifies the day and time the BBU self test is to run Date Time Sets SCSI disk array date and time to match the host system Host I F Options Sets or changes the Host interface settings for the selected SCSI disk array See Figure 7 50 Serial Number Sets or changes the disk array s serial number System Cache Specifies system cache states See Figure 7 48 System Memory Specifies system memory configuration values See Figure 7 49 System Type Selected Expert Tool Screens Controls certain operating characteristics of the array HP supported types are 2 and A The following screen captures Figure 7 35 through Figure 7 50 are examples of selected Expert Tool screens 7 114 Support Tools Manager STM Info Capacity amp hpterm ioj xj Expert Tool Capacity for device C2300 Array on path 874 9 A salt rose hp com 15 43 214 34 Expert Tool Capacity for device C2300 Array on path 8 4 9 Array Date 04 29 98 SN 9621954 System Date 04 29 98 Array Time 11 04 08 FH 9 25 00 8 02 00 System Time 11 13 15 Array SP A 9 0 MN 7624 SP System Namei salt User Information Lun Information Lun VendorID ProductID Cap lt blks gt BSz GBytes Last Addr BSz GBytes 0 DGC C2300WDR5 68908032 512 35 281 68908031 512 35 281 1 DGC C2300WDR1 17227026 512 8 8
215. ation Utility Screen 8 86 Troubleshooting with Grid Manager Bound but Not Assigned Status after Replacing an SP When a disk is bound to an SP it shows an ENABLED state under the Performance Statistics screen for the selected disk The ENABLED state indicates the disk is currently owned by the SP To view the disk state in Grid Manager highlight a disk module in the Presentation Screen and press RETURN The following output will display Performance Statistics For Disk Module A2 Description 2 1 Gbyte SCSI Disk State ENABLE Unit Number Unit Type D Looking from the owning SP hex 0 RAID 5 Group Individual Access Array Stripe Element Size 128 Sectors aximum Rebuild Time 4 Hours Read Requests 4246546 Write Requests 324804 Kbytes Read 388832 Kbytes Written 2563443 Average Disk Service Time 14 ms Stripe Boundary Crossing 360 If viewed through the non owning SP the state for a disk is BOUND BUT NOT ASSIGNED as in the following Troubleshooting with Grid Manager 8 87 punoouse qnouj Performance Statistics For Disk Module A2 Description 2 1 Gbyte SCSI Disk State BOUND BUT NOT ASSIGNED Looking from the non owning SP Unit Number hex 0 Unit Type RAID 5 Group Individual Access Array Stripe Element Size 128 Sectors aximum Rebuild Time 4 Hours Read Requests 4246546 Write Requests 324804 Kbytes Read 388832 Kbytes Writt
216. ay by Da eed a rhe eee caches ea sm digahlad Change sysium memory values bo fhai vpu iiid Figure 7 25 Utilities System Configuration System Memory Screen Support Tools Manager STM 7 97 sonsoubeiq pue seni ISOH Utilities 9 System Configuration Host I F Settings lhis commend eid pot m hne the Heri Intar jare Betting tor Array Farisi d PSK Tai beri issu Rhian rin 5 Enabled oot Disaini Suber thay tor Dean Full gt Enabled d ie ved vid Heis Page kDa Super Enabled v be eid il Recond Error Log Reporting Erabud ie Dead iow Pun Minor Cache at Embi ET EE C1 CA Hele Figure 7 26 Utilities System Configuration Host I F Settings Screen 7 98 Support Tools Manager STM mstm Interface The main mstm screen provides a list of the system devices The main menu appears at the top of the screen Select a device to access the diagnostic tools described below Access these tools through the Tools pull down menu on the device map screen These tools along with an indication of whether a license is required to run each are listed in Table 7 13 To set the license select System License Install HP Only from the mstm menu bar Enter the license password in the dialog box and click OK Table 7 13 Available Support Tools Support tool license required Tool type Description No Yes Information Provi
217. ay divides the multiplexed external Fast Wide differential SCSI 2 bus traffic from the host into two internal synchronous SE SCSI 2 buses identified by the letters A and B The SP in a model 20 disk array divides the multiplexed external Fast Wide differential SCSI 2 bus traffic from the host into five internal synchronous SE SCSI 2 buses identified by the letters A B C D and E Each internal SCSI 2 bus supports multiple physical disk units LUNs storage system caching See disk array caching striping disk striping The arrangement of a physical disk unit LUN such that reads and writes can occur with multiple disk modules simultaneously and independently By allowing multiple sets of read write heads to work on the same task at once disk striping can enhance performance Hardware disk striping is used by RAID 5 RAID 0 and RAID 1 0 groups The size of the stripe is the stripe element size area on each disk multiplied by the number of data disks in the group The stripe element size for supported RAID levels can be selected with Grid Manager unbind See bind Unsolicited Event Log A time ordered list of messages about disk array events maintained by the SP at the array level G 6 Glossary A AC line filter removal 9 48 9 68 AC power specifications R 9 acoustics R 12 ADP on Presentation screen SP A s dirty pages 6 40 anti tip legs 2 7 auto assign in STM 7 85 7 114 auto trespass R 3 1 11 1 22 checking 6
218. ay the chassis on the covered side c Remove the seven retaining screws from the bottom Figure 9 20 d Disconnect the LED cable from the midplane e Lift the inner chassis up and out of the system i Rotate upward the rear of the inner chassis about 3 inches ii Lift up and out 7 Remove the midplane The midplane is part of the inner chassis a Remove 13 retaining screws and remove the midplane PCA from the inner chassis Note To install the inner chassis in the system lift it over the lip at the rear of the system 9 32 Model 10 Deskside Component Removal Inner metal cover Figure 9 19 Removing the Inner Metal Cover Model 10 Deskside Component Removal 9 33 jueuie2e deH pue jeAoway 4 Top Six retaining Screws Bottom Seven retaining screws Figure 9 20 Removing the Inner Chassis 9 34 Model 10 Deskside Component Removal Interconnect Printed Circuit Board IPCB Figure 9 21 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds Access the rear of the disk array Disconnect the external SCSI 2 cables Use a small screwdriver if necessary a Remove two screws and lift the SCSI connector and cable away from the IPCB bulkhead Remove the two retaining screws that secure the IPCB to the chassis Pull both ejectors until they unlock and the IPCB starts out of the chassis Pull the IPCB out of the chassis making su
219. bes use of the Grid Manager utility to control a variety of disk array functions The procedures and Grid Manager screens are from the latest microcode available at the time of publication Presentation Utility Screen 6 41 Jobeue N pus Enabling Disabling the SP620 SCSI Storage Processor The SP has an enable disable switch The higher performance SP620 SCSI storage processor does not have an enable disable switch It must be enabled and disabled using Grid Manager The procedure for disabling and enabling the SP620 SCSI SP follows 1 Display the Grid Manager Presentation Utility screen 2 From the Presentation screen type caret and press RETURN to go to the Main menu 3 Follow the menu path 5 Change Parameters 3 Change SP Parameters 5 Soft Disable Enable SP amp xterm SP A SP SCSI ID 15 S N 9621954 Soft Disable Enable SP This option will shut down write caching in the system block any new I 0 and then disable this controller You will be prompted for input to restart this controller Do you wish to continue Y N N Y Mrite cache is disabled Stopping this controller This controller is disabled Do you wish to restart this controller Y N NIYR Figure 6 9 Soft Disable Enable SP 4 Type Y at then first prompt to disable the SP620 SCSI SP 6 42 Enabling Disabling the SP620 SCSI Storage Processor To re enable the SP620 SCSI SP follow steps 1 through 5 above Type Y at the second
220. ble by lifting up and out through the top of the bracket See arrow 4 in Figure 9 57 Ss 2 LED Bezels 7 gt LEDs zs 7 4 LED Bracket Q LED Cable Connector Midplane Connector Figure 9 57 Removing the LEDs LED Cable Model 20 Rackmount Component Removal 9 97 u w vejd y pue jeAoway DC Power Distribution Cable Figure 9 58 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 Remove the following a b Fan pack Figure 9 6 SP Figure 9 9 Power supplies Figure 9 11 BBU located under the power supply units Figure 9 13 All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array 4 Remove the front panel Figure 9 40 5 Remove the 4 retaining screws that hold the chassis to the rack Figure 9 52 WARNING The disk array is very heavy Use extreme care when moving or SR a p 9 98 lifting it At least two persons are needed to lift a disk array Remove the empty chassis from the rack Figure 9 52 Remove the top metal panel 11 screws Figure 9 53 Remove the midplane Figure 9 55 Remove the retaining screws from the four connectors two screws per connector Model 20 Rackmount Component Removal 10 Remove the harness Back of midplane mounting bracket Connector J27 DC power distribution cable
221. ble to this SP example lun owned by peer SP and auto trespass is off The command will fail 85 84 88 and Status will be set to EXCEPTION Status SenseKey Sense Data w location of Sense Key ASC and ASCO y EXCEPTION 05 25 01 78 B8 85 BB B amp B BB BB BA BB BB BB BB 25 81 BB BA B8 HA Figure 7 19 Test Unit Ready Screen Support Tools Manager STM 7 91 sonsoubeiq pue seninf ISOH Tests Verify Disks The Verify Disks and Verify LUNs functions are set up the same as the Read functions but the information is displayed in the Expert Tools main window Figure 7 10 1 bcs Select Tests Verify Disks from the Expert Tools main menu Select the type of verification from the dialog box that appears Figure 7 20 Select the desired action and click Yes The Select Blocks dialogue box appears Figure 7 21 Define the appropriate block parameters and click OK When the system completes the verify Disks process the Expert Tools main window displays the information mami Vr Wdy Dah ia This cemar ei rib vis ihe BEH Yeeibp rorem H Diski deri Amey erie WPI Sein dusalrasd sman Ace Verity ALL dioke ALL blocka sa liriwd in not selection Vary SELECTED dimks ALL bierke 1722 172530997 romae TBEEBT 728008987 Brash Blihe B thee OBIE lL bs arbei il1 ba soko 11 b sake s l he asked Ll be anke LI ba amk Iu py wont cz Verty the disks as sekected apecitied Figu
222. buttons to the left of each text box to obtain a new window containing a list of available values Troubleshooting with SAM 8 99 punoouse qnouj When satisfied with the three values select one of the items in the list of RS 232 configurations if that entry is being modified then select Modify to change the current values to the new values in the text boxes To add a new device configuration to the list select Add When satisfied with all of the configuration settings select OK to implement them and close the window To abort and close the window without altering current settings select Cancel instead 8 100 Troubleshooting with SAM Creating a File System on a Series 700 System On an HP 9000 Series 700 system running HP UX 9 05 or HP UX 9 07 if LUNs are configured so that they are all identical in size SAM or the HP UX newfs and mkfs commands can be used to create file systems on the LUNs If the LUNs are configured to be a mix of different sizes only newfs and mkfs can be used to create the file systems SAM cannot be used if the LUNS are configured to be different sizes Also on S700 systems running HP UX 9 05 or HP UX 9 07 with patch PHCO 7461 installed if multiple LUNs of varying sizes are configured and diskinfo is run each LUN with the same RAID level will report the same vendor and product id even though the capacity of the LUNs may differ either because of disk capacity or the number of disks used in the RAID binding pro
223. byte 4 x 16 Mbyte four A3234A SIMMs 64 Mbyte 8 x 16 Mbyte eight A3234A SIMMs To set up disk array caching use Grid Manager to specify the cache page size and cache size then change the cache state from disable to enable Specify the cache page size and cache size only if disk array caching is disabled Before specifying these parameters for the first time they have these default values Cache Page Size 2 Kbytes Cache Size 0 Disabling and Enabling Disk Array Caching Because of cache mirroring the maximum cache size is 16 MBytes in the model 10 disk array and 64 Mbytes in the model 20 On power up the disk array enables caching if the prerequisite hardware is working and the cache size parameter is a valid non zero value Determine whether disk array caching is disabled or enabled by looking at the Cache State status on the Presentation Utility screen Disk array caching can be disabled or enabled without affecting the information stored on the logical disk units However disk array caching must be disabled before Grid Manager allows a change to the cache page size and the cache size parameters or when removing an SP To Disable or Enable Disk Array Caching 1 From the Grid Manager Main menu follow the menu path 5 Change Parameters gt 2 Change Storage System Parameters 1 Change Storage System Cache Parameters 6 54 Setting up Disk Array Caching for Microcode Revision 7 23 Change Storage System C
224. cache The problem might be BBU is not ready not present and fully charged One or more vault disks is missing or being rebuilt Fan fault occurred or SP failed 0x909 Cache Dump Failed A fault caused the disk array to try to dump the vault The cache dump failed because two or more vault disks are missing or have failed 0x90A Assign Fails Cache Dirty Unit The SP cannot assign this unit because it is marked as having cached data outstanding and the SP has no access to the data Ox90B Controller Cannot Initialize Cache Some hardware or software failure occurred when the SP attempted to initialize a new cache image 0x90C Cache Image Exceeds Memory Size This SP is incapable of loading the saved cache data image due to insufficient cache memory size 0x90D BBU Removed The BBU in this disk array has been removed Ox90E BBU Disabled The BBU in this disk array has failed its internal selftest Ox90F Cache Recovered With Errors A non mirrored cache recovery failed to recover the cache pages for some but not all cached LUNs It does not apply to a disk array with two SPs 0x910 Cache Recovery Failed A non mirrored cache failed to recover information for all cached LUNs It does not apply to a disk array with two SPs 0x920 Hard Media Error This disk has reported a media defect with could not be cleared by the SP See Hard Media Error versus Hot Spare on p 67 of
225. cedure The following example of diskinfo output shows a LUN bound at RAID 1 with a capacity of 4 Gbytes etc diskinfo dev rdsk cb00df10s0 SCSI describe of dev rdsk cb00df10s0 vendor DGC product id C2300WDR1 type direct access size 4102875 Kbytes bytes per sector 512 The next example of diskinfo output shows another LUN bound at RAID 1 with a capacity of 2 Gbytes etc diskinfo dev rdsk cb00dfl1s0 SCSI describe of dev rdsk cb00dflls0 vendor DGC product id C2300WDR1 type direct access size 2036569 Kbytes bytes per sector 512 The vendor and product id fields are concatenated together in the etc disktab file to identify different disk drive types and capacities This information traditionally has been Creating a File System on a Series 700 System 8 101 punoouse qnojj used for creating file systems with the newfs command If this is done now as is shown in the next example either the newfs command will fail or the full capacity of the disks will not be used etc newfs dev rdsk cb00dflis0 DGC_C2300WDR1 where DGC is the vendor C2300WDR1 is the product id To avoid these problems it is strongly recommended that LUNs with different capacity configurations not be mixed If however LUNs of different size must be configured the newfs command must be run using the correct etc disktab entry To find the correct entry diskinfo must be run on the desired LUN For example etc
226. cessary SCSI 2 adapter No 9 04 I O operations fail to all physical disk units configured on SP cable No 10 x attached to a failed SCSI 2 adapter cable The System Administrator can transfer control to the other SP and adapter by changing host s and array configuration Can regain access to LUNs associated with failed SP host adapter or SCSI 2 cable by making changes manually in Grid Manager and LVM 4 12 Multi Bus Configuration Multi Initiator Configuration HP UX 10 x and Later The multi initiator configuration shown in Figure 4 5 consists of more than one host sharing one external Fast Wide differential SCSI 2 bus Each host has a SCSI 2 adapter connected to one of the two SPs in one array Each host has its own SP which defines that host s disks The disks remain primary to that host Host computer 1 Host computer 2 SCSI 2 SCSI 2 bus goats adapter adapter Disk array SPA C SPB Figure 4 5 Model 10 20 Multi Initiator Configuration Multi Initiator Configuration HP UX 10 x and Later 4 13 009 seues 0006 dH uoneanbiuo2 Table 4 6 Multi Initiator Error Recovery Failing component Disk module Continue after failure Yes What happens and how to recover With RAID of any level other than 0 or an individual disk applications continue running System Administrator
227. che is not disabled until all modified pages are written to disk To determine whether caching has been disabled display the Presentation Utility screen and wait for the Cache State display to change to DISABLED When DISABLED appears other disk array caching parameters can be changed To set up anew disk array or upgrade the disk array to support caching and disk array caching has been disabled follow the instructions in Specifying Cache Page Size and Cache Size on p 56 of this chapter Specifying Memory Configuration for Microcode Revision 8 xx 9 xx 6 63 Jobeue N pus If Changing the Cache State from Disabled to Enabled To determine whether disk array caching is enabled display the Presentation Utility screen and wait for the Cache State display to change to ENABLED If ENABLED does not appear in a few moments make sure that the following conditions are true Both SPs are powered up as indicated by the display of Peer SP Present on the Presentation Utility screen The BBU is fully charged as indicated by the display of BBU OK on the Presentation Utility screen If BBU NR appears the battery is charging and the Cache State should go to ENABLED when BBU NR changes to BBU OK Each SP has at least 8 Mbytes of memory The disk array is configured and functioning properly that is The array contains disk modules in slots A0 Al A2 A3 and A4 in the model 10 and in slots AO BO CO DO and EO in the model 20 All the di
228. cited Event Log POW n Powering up RDY n Ready bound with its secondary route via this SP The other SP is the primary route to this disk module On the other SP s display this module s box shows status ENA For a hot spare code is RDY H and means that the hot spare is in normal standby mode ready for service if a module in a RAID group fails REB n Rebuild is in progress UNB Unbound ready to be bound into logical disk unit blank No disk module or a non operational disk module in compartment slot PROM Cache and Other Components Status The leftmost column shows the PROM revision and the status of the fan packs power supplies BBU and disk array cache PROM Revision Shows the revision of the SP s PROM code in the form n nn PROM code revisions can be mixed that is one SP can run one version of PROM code 7 97 for example and the other SP can run a different version 7 99 or 1 56 for example The current at time of publication PROM codes and their corresponding microcode revisions are Presentation Utility Screen Jobeue N pus PROM 7 97 PROM 7 97 PROM 7 99 PROM 8 02 PROM 1 49 PROM 1 56 PROM 1 57 PROM 8 02 PROM 1 73 Microcode revision 7 16 Microcode revision 7 23 Microcode revision 8 20 Microcode revision 8 27 Microcode revision 8 51 Microcode revision 8 56 Microcode revision 8 58 Microcode revision 9 26 07 Microcode revision 9 56 07 4 Gbyte HH half he
229. cont d Displayed information Model 10 Model 20 Activity 4 dots 4 dots DDRAM Odd Crosstalk Test 5 dots 5 dots DDRAM Address Test 6 dots 6 dots DDRAM Not Address Test 7 dots 7 dots DDRAM Burst Initialization Test A A B B C C IDRAM Write Read Test D amp E D amp E IDRAM Code Test F F IDRAM Parity Test G G Extended DRAM Test H H Non Volatile Memory Test l l 710 SE SCSI Bus Controller Chip Test J J 710 SE SCSI Bus Controller Chip Test K 710 SE SCSI Bus Controller Chip Test L 710 SE SCSI Bus Controller Chip Test M 710 SE SCSI Bus Controller Chip Test K N 720 WD Bus Chip Test L O M P Interrupt Test N Q Starting Database Drive O R Starting Database Drive P S Starting Database Drive Q T Starting Database Drive R U S V Sizing Disk Sectors T W Booting Power Up Troubleshooting 8 11 Bunoouse qnoJj Table 8 3 Disk Array Power up Tests for the SP only cont d Displayed information Model 10 Model 20 Activity U X ms VioZ Y amp Z dee The alphabetic characters indicating that power up tests are running do not display for the SP620 SCSI SP Instead the power up tests display together with a pass or fail status following the tests The power up tests and the power up sequence for the disk array should complete within three minutes Visual indications of power up are
230. crocode rev 8 20 6 22 menus for microcode rev 8 56 6 25 menus for microcode rev 9 xx 6 28 password 6 71 Presentation Utility screen 6 33 GridMgr ix about 6 11 error conditions and messages 6 14 how to exit 6 13 6 14 program file 6 11 H HARAYMON about 7 5 error reporting 7 10 HP UX 9 0x vs HP UX 10 x 7 10 installing 7 5 messages 7 7 updating on HP UX 9 x 2 5 hard media error 8 67 Hardware Event Monitor 7 132 Hardware Event Monitoring for HP UX 7 132 hardware mirror 1 24 1 26 HARRAY installing 7 5 Section 10 7 12 Section 17 7 13 updating on HP UX 9 x 2 5 heartbeat on Presentation screen 6 34 heat output R 11 HELP HARRAY command 7 26 HELP Nikearry command 7 57 high availability array monitor HARAYMON 7 5 features 1 18 options R 5 host adapter cards installing 2 19 firmware 2 4 host I F options array 7 85 7 114 Index l 3 xepu host interface options defaults for microcode rev 8 20 6 68 6 69 hot spare 1 32 1 38 1 41 6 45 8 24 for model 10 1 33 for model 20 1 34 hot swap R 3 HP UX operating system 2 3 l ID string model 10 7 3 model 20 7 3 individual access array 1 30 1 39 individual disk unit 1 32 1 38 1 40 Information Tool of mstm 7 99 7 107 Information Tool of xstm 7 69 7 78 initiating Grid Manager HP UX 6 10 MPE X 6 11 INQUIRY HARRAY command 7 28 INQUIRY Nikearry command 7 57 install a rackmount disk array in a cabinet 2 10 installation overview of 2 2 installation troubleshooti
231. cted by the SP 0x648 Termpower Low Extended status is SCSI bus number This message is logged once at the completion of each powerup for any SCSI bus which the SP detects as having an insufficient TERMPWR voltage level This event code is only used by SPs which do not themselves source termpower into the bus 0x650 FRU Signature Error The given FRU s current physical position in the matrix of disks does not match its expected position which is set at the time the FRU was bound into a LUN 0x654 Cache Dumping SP started dumping the storage system cache to the vault disks 0x657 Cache Dump Completed SP finished dumping the storage system cache to the vault disks 0x658 Cache Enabling The SP is in the process of enabling the cache for use on host data 0x659 Cache Disabling Storage system caching was disabled by the disk array or System operator The disk array disables caching if the BBU is not fully charged or an SP vault disk or fan fails see error 0x908 0x660 AC Power Fail After an AC power failure the disk array dumped the cache to the vault and turned off power to the BBU it does this to minimize drain on the BBU 0x661 BBU Sniffing Enabled This disk array is now allowed to shut down the BBU in order to test the BBU 0x662 BBU Sniffing Disabled BBU shutdown testing is disabled on this disk array 0x663 BBU Sniff Initiating The BBU on this disk array is beginning a shutdown test cycle 0x664 BBU Sniff Cache Dump
232. cution of power up diagnostic test data is written and read to all disks This indicates a data miscompare on one of the drives Ox1F Write stamp error SP microcode information imbedded in the data stored on the array was inconsistent with what was expected 0x20 Device dead The targeted disk is not available Any residual commands to that disk initiate this code 0x21 Recovered error The targeted disk returned Recovered Error sense key excluding invoking ECC and spindle sync 0x22 Recovered error with ECC The targeted disk returned Recovered Error ECC invoked SCSI sense key 0x23 No error sense key The targeted drive returned No Error SCSI sense key 0x24 Logical sector data error Same as 0x14 0x18 0x25 Host driver abort Host caused termination of back end bus activity 0x26 Host bus protocol error An unrecoverable protocol error on the host bus caused an I O to terminate 8 72 Unsolicited Event Log Table 8 15 Extended Status Codes cont d Extended status code 0x27 Description Host bus parity error The host bus parity error retry count was exceeded for an I O causing it to terminate 0x28 Spindle synchronization change This error code should never be seen 0x29 Not ready start in progress This error code should never be seen Ox2A Transfer count not zero Too few bytes were transferred by a drive during an I O operation Ox
233. d Manager starts the binding operation it changes the unbound module symbol to the LUN number specified for each module being bound To bind additional disks type caret once to return to the Bind Logical Units menu and repeat steps 4 through 10 Or to return to the Grid Manager Main menu type twice While Grid Manager is binding the unit it is possible to bind another unit or hot spare by repeating the procedure When Grid Manager finishes binding the unit it changes BIN to ENA Note To determine the progress of the binding procedure highlight one of the group s disk modules and press RETURN The Performance Statistics screen displays a percentage complete figure Creating Logical Disk Units LUNs 6 49 Jobeue N pus Table 6 5 Bind Options for Different Groups Logical disk unit type Bind options Recommendations RAID 5 Logical unit number Default Maximum rebuild time Default Sectors per stripe element Default Enable caches Both RAID 1 0 Logical unit number Default Maximum rebuild time Default Sectors per stripe element Default Enable caches Both RAID 1 Logical unit number Default Maximum rebuild time Default Enable caches Both RAID 0 Logical unit number Default Sectors per stripe element Default Enable caches Both Individual disk unit Logical unit number Default Enable caches Both Global hot spare Logical unit number Default LUN caching works only if disk a
234. d SCSI chip on an SP or a bad device 0xA05 NOVRAM Uninitialized The non volatile memory on the SP is uninitialized No host I O can occur 0xA06 Chassis Shutdown An overtemperature situation could not be corrected All drives in the chassis will be shut down The problem must be corrected before powering up again OxA07 Drive Failure The disk drive has failed and can no longer be accessed 0xA08 Data Base Sync Error The SP cannot determine the correct logical configuration of all units in this disk array Some units may be unusable 0xA09 Drive too Small A replacement drive has been inserted and does not have adequate capacity for a rebuild operation to begin OxA11 Peer Controller Removed This SP s peer controller is no longer responding to queries from this SP following a period of time where the peer was known present 8 64 Unsolicited Event Log Table 8 13 Fatal Error Codes cont d Event code Description 0xA12 Cache Memory Hard Error The disk array detected a hard error in SP memory in a disk array with non mirrored caching This error does not apply to a disk array with two SPs OxA13 CRU Type Unsupported This type of disk CRU is not supported by the current SP software The disk cannot be bound or used for host I O Rebuild Aborts A 0x641 CRU Rebuild Aborted event may occur if an auto trespass is launched against a LUN when a rebuild is in progress for
235. d Vendor ID SEAGATE Product ID ST1251ON OK Help Figure 8 25 View Status Screen The View Status screen has a series of 10 or 20 buttons arranged in one or two vertical columns with two columns of text to the right of each button Each button corresponds to a disk device in the array enclosure cabinet The text identifies the device and related information e The first column of text to the right of each button identifies which bus the disk is connected to and which bus position it occupies 8 92 Troubleshooting with SAM e The second column identifies the LUN number associated with the device if status information about the array is being viewed or the current status of the device if one or more devices is being bound to a selected LUN LUN n is the LUN number associated with the corresponding disk Disk Module Identification The first column to the right of each toggle button contains a letter followed by a number as follows C1300 model 10 disk array Two internal SE SCSI 2 buses labeled A and B Five disks per SCSI 2 bus labeled 0 through 4 C2300 model 20 disk array Five internal SE SCSI 2 buses labeled A through E Four disks per SCSI 2 bus labeled 0 through 3 Selecting a Disk Module in the Array To view the status of a particular disk module in the array or to bind a specific disk drive to a LUN select the appropriate toggle to view the corres
236. d the SP rebuilds the data and parity on it During this time however the disk array continues to function and give users access to all data including data stored on the failed module Individual Disk Unit An individual disk unit is a disk module bound to be independent of any other disk module in the disk array An individual disk unit has no inherent high availability It can be made highly available by using software mirroring with HP UX 10 x and later Note Individual disk unit is not supported on models A3549 or A3550 because it does not provide data redundancy For microcode revision 8 56 and later this RAID level can only be configured in FE mode with the Grid Manager utility See chapter 6 Grid Manager for more information about FE mode Hot Spare A hot spare is a dedicated replacement disk unit on which users cannot store information Grid Manager explained in chapter 6 is used to bind a disk module as a hot spare The capacity of a disk module bound as a hot spare must be at least as great as the capacity of the largest disk module it might replace If a hot spare is larger than the failed disk module it replaces its extra capacity is not used If any disk in a RAID 5 group RAID 1 mirrored pair or RAID 1 0 group fails the SP automatically begins rebuilding the failed disk module s structure on the hot spare When 1 32 Planning Disk Unit Configurations the SP finishes rebuilding the disk group functions as usual
237. d Descriptions 0 2 e eee ee eee eee 7 112 8 1 8 2 8 4 8 5 8 6 8 7 8 9 8 10 8 11 8 12 8 13 8 14 8 15 8 16 8 17 9 2 Table of FRUS s strndesisp say ek bed Qa eles Shee hte eel weds BAIE SE EI ee 8 2 Model 10 and 20 Disk Array LEDs sseeeeeee eee 8 9 Disk Array Power up Tests for the SP only 0 eee eee eee 8 10 Disk Array Power Up Tests for the SP620 SCSI SP only 8 13 Power up Troubleshooting 0 0 cece cee eee II 8 16 Minimum Microcode and PROM Revisions for Supported Disk Modules 8 20 Sense Key Additional Sense Code Additional Sense Code Qualifier Values Not a Complete List 21 0 cece Ie 8 24 SP Troubleshooting emer RP Pe bey Ape oN e Ce P ER Pes 8 30 SIMM Configuration seeeeeeeeeeee e 8 33 Informational Codes 2 0 0 cece cece Ih 8 57 Thresholded Codes aede EERMULC IU amd 8 61 Error Codes nee exu EFE Use em s een Rena 8 62 Fatal Error Codes onerare aes Pe id oa EE PR een Seed Beaks 8 64 Rebuild Abort Events and Messages 00 e eee eee eee eee eeneee 8 65 Extended Status Codes 2 0 0 cece eee eect EE ene eens 8 70 Microcode Panic Categories 00 cc eect eee ene eeee 8 75 Microcode Panic Codes Not a Complete List elles esee 8 76 Model 10 Disk Array CE Replaceable Parts see Figure 9 1 9 3 Model 20 Disk Array CE Replaceable Parts see
238. d as root boot disk refer to the appropriate system 1 48 manual for instructions Also see chapter 6 Grid Manager if changes to the default RAID 5 configuration of the array are required In HP UX 9 04 it will not be possible to easily fill in these columns until the hardware is actually configured with the operating system up and after running the ioscan f n Cdisk command The H W Path output from ioscan will provide the Hardware Path column information The information under ioscan s Driver output will provide the Device File column information Planning Disk Unit Configurations In HP UX 10 x and later it will not be possible to easily fill in these columns until the hardware is actually configured with the operating system up and after running the ioscan f n Cdisk command The H W Path output from ioscan will provide the Primary Path and Alternate Path column information The information under ioscan s Driver output will provide the Device File column information Normal procedures should be followed for setting up logical volumes To set up logical volumes with primary and alternate path functionality refer to the steps in LVM Logical Volume Manager which is in chapter 7 Use the remaining spaces on the worksheet to help in that configuration Planning Disk Unit Configurations 1 49 uonduioseg jonpold 1 50 Planning Disk Unit Configurations INSTALLATION Overview of Installation Check the Softwar
239. d be added to the Devices file Direct tty0pl 9600 direct Direct tty0p2 9600 direct Add a special user login for Grid Manager When using multiple MUX ports set up a special login for each port The profile setup requires separate logins It is a good practice to set up the first port as gridmgr0 the second as gridmgrl the third as gridmgr2 etc cd etc cp passwd passwd old vi passwd Go to the etc directory Make a backup copy of the passwd file Use the vi editor to edit the passwd file Add the appropriate number of logins to the passwd file This example sets up two logins to access two different SPs by way of the MUX gridmgr0 101 20 The Gridmgr users gridmgr0 bin ksh gridmgr1 101 20 The Gridmgr users gridmgrl bin ksh Create a directory for each login and add a profile to each directory Each login accesses the specified MUX port which in turn accesses Grid Manager by way of the RS 232 connection to the SP If there is one array with two SPs set up the gridmgr0 profile to access SP A and the gridmgr1 profile to access SP B Example Go to the users directory Make directory for the first login Make directory for the second login Go to the first login directory cd users mkdir gridmgr0 mkdir gridmgrl cd gridmgr0 vi profile Add the profile information shown below for the first login so it accesses the first SP Do not add the comments shown between the
240. d other activity on all hosts attached to the disk array whose firmware is being updated 3 Run SYSDIAG Remember to type SUPLICEN 4 Run HARRAY Section 17 and choose a PDEV Note Both SPs should be enabled during the download 5 Runthe DOWNLOAD command 6 Downloading new SP microcode takes approximately five minutes 7 20 HARRAY Section 17 External Exerciser 7 After the diagnostic completes the download successfully check the firmware and PROM revisions in Grid Manager When updating to microcode revision 8 56 which is for the SP620 SCSI SP only the microcode revision number will be updated when the SP620 SCSI SP is installed and the array reboots under control of the SP The current microcode revisions and their corresponding PROM codes are Example Revision 7 16 Revision 7 23 Revision 8 20 Revision 8 27 Revision 8 51 Revision 8 56 Revision 8 58 PROM 7 97 PROM 7 97 PROM 7 99 PROM 8 02 PROM 1 49 PROM 1 56 PROM 1 57 Revision 9 26 07 PROM 8 02 Revision 9 56 07 PROM 1 73 Download type 1 4 Gbyte HH half height disk support for the SP on model 10 and model 20 only 2 Gbyte disk support for the SP only for the SP only 4 Gbyte LP low profile and 9 Gbyte HH disk support for the SP on model 10 and model 20 only 4 Gbyte HH disk support for the SP620 SCSI SP only for the SP620 SCSI SP only 4 Gbyte LP and 9 Gbyte HH disk support for the SP620 SCSI SP only 4 Gby
241. d the process does not continue until the request is satisfied Grid Manager must be used to disable an SP STM does not support this functionality CAUTION Never reenable a peer SP while a disk module firmware update operation is in progress Doing so can interrupt the operation causing permanent disk module and or data loss Exclusive access is required to download disk firmware All file systems should be deactivated while updating disk firmware Expert Tool Expert Tool provides useful SCSI disk array support capabilities including e Information retrieval e Functionality testing e Failure analysis e Data integrity checking e SCSI disk array management Expert Tool offers much of the functionality available through Grid Manager 7 82 Support Tools Manager STM Access the Expert Tool window by selecting Tools Expert Tool Run This window provides six menu items in addition to Lab Use Only and Help Figure 7 10 The Lab Use Only item is intended for use only by Hewlett Packard factory engineers and is not documented here This window also provides a status of the activities performed on the selected device Expert Tool for device type C2300 Array on path 874 8 File Logs Cxx Array Expert Tool version R 82 88 Use Help to get more information Bind command selection CANCELED per user request Capacity command SUCCESSFUL Capacity command SUCCESSFUL Inquiry command SUCCESSF
242. d wait for the Cache State display to change to ENABLED If ENABLED does not appear in a few moments make sure that the following conditions are true Both SPs are powered up as indicated by the display of Peer SP Present on the Presentation Utility screen EachSP has at least 8 Mbytes of memory The disk array is configured and functioning properly that is The array contains disk modules in slots AO Al A2 A3 and A4 model 10 and in slots AO BO CO DO and EO model 20 All disk modules in the array are functioning normally that is none is being rebuilt A hot spare has not been configured in slots AO Al A2 A3 and A4 model 10 or AO BO CO DO and EO model 20 Even though disk array caching is enabled the disk array does not use caching with any logical disk unit unless the write and or read cache for that unit has been enabled It can be enabled when binding the unit or at any later time with the Change Logical Unit Parameters menu Specifying Cache Page Size and Cache Size The values that can be specified for the cache page size and cache size are Cache page size 2 4 8 or 16 Kbytes Cache size 0 8 16 32 or 64 Mbytes Before the cache size or page size value can be changed the cache state must be disabled The previous section explained how to disable it 6 56 Setting up Disk Array Caching for Microcode Revision 7 23 To specify disk array caching parameters follow these steps
243. d without affecting the information stored on the logical disk units However disk array caching must be disabled before Grid Manager allows a change to the cache page size and the cache size parameters or when removing an SP To disable or enable disk array caching 1 From the Grid Manager Main menu follow the menu path 5 Change Parameters 6 62 Specifying Memory Configuration for Microcode Revision 8 xx 9 xx 2 Change Storage System Parameters 1 Change Storage System Cache Parameters 2 To change the parameters answer the screen prompts e xterm SP A SP SCSI ID 15 Change Storage System Cache Parameters S N 9621954 fit any prompt enter to G BACK TO PREVIOUS HENU for HELP Current storage system cache parameters Write Cache type HIRRORED Read cache state DISABLED Write cache state DISABLED Peer Read cache state DISABLED BBU test time SUNDAY 01 00 Cache page size 2 KB Change caching parameters N Y Write cache state Enable Disable Disable E Cache page size in KB 2 4 8 16 2 Read cache state Enable Disable Disable E Peer Read cache state Enable Disable Disable Ej Figure 6 16 Change Storage System Cache Parameters Screen Disable 3 Answer E or D to specify the other state If Changing the Cache State from Enabled to Disabled Grid Manager displays the previous menu Cache parameters can be changed only when caching is disabled This may take a few seconds since the ca
244. daptor connector RS 232 cable tightly secured to host port OOUddUdA Shortest possible SCSI cable lengths between disk arrays and host adaptors SCSI bus length under 25 m including the disk array s internal bus length Check disk array pre power up J All FRUs reseated into the disk array J SP SIMMs securely fastened in correct SP SIMM sockets J Disk modules properly seated in their appropriate slots SP enable disable switch enabled on SP Disk Array Installation Troubleshooting Checklist Revision 1 3 2 45 uonejjeisuj Check disk array functionality and configuration Qa udguuci clclaou eaucL cL L J Date and time correctly set on all SPs No fault indicators on disk array LEDs All disk modules spin up Disk array in default factory configuration RAID 5 Reconfigure if necessary All LUNs bound and assigned Cache enabled if present Unsolicited Event Log s error free at first power up Disk array Storage Systems Option Package parameter properly set option A or 2 System IOMAP properly maps out the disk array Current SP microcode installed Current disk mechanism firmware installed Current F W SCSI host adaptor at current firmware level Different capacity disk modules not mixed in same LUN Required information for escalated sites unaua 2 46 Installation troubleshooting checklist reviewed with exceptions noted Epic number for escalation Detailed problem descriptio
245. del 10 Mapping Microcode 8 20 0 eee eee eee eee eee 8 22 Model 20 Mapping Microcode 8 20 0 cece eee cee eee eee 8 23 SEP SIMMS iuit Sg eee insta we erie es SS ape Sil ad hd aed aes 8 31 Array Power Block Diagram 0 0 cece eee ee ee ee 8 40 RIE PCBEEDS uinea E at ene bee hee ae E RUE e 8 43 Grid Manager Unsolicited Event Log lleleleleeeeeeeeeeee 8 45 Current Faults Screen 0 0 0 cee cent nee arree 8 47 First Page of a Unit Verify Results Report 0 0 c eee eee eee 8 48 LUN Configuration Summary Screen 2 cece eee eens 8 50 xxi sainbi4 8 18 8 19 8 20 8 21 8 22 8 23 8 24 8 25 8 26 8 27 8 28 8 29 9 1 9 2 9 3 9 4 9 5 9 6 9 7 9 8 9 9 9 10 9 11 9 12 9 13 9 14 9 15 9 16 9 17 xxii First Report of the Display Disk Failure Information 8 51 Non V RAM Error Logis cst oe een RE bare le Rees 8 52 Unsolicited Event Log Display 0 0 eee cece eee eens 8 53 Apply Filters Menu RR Rt CD ERR ESO eee oe ot 8 55 SCSI Jumper Cable Example 00 0 0c eee cece eee een neee 8 66 Typical Model 20 Deskside Presentation Utility Screen 8 86 Disk Devices Screen 2 unte Lee e e REA 8 01 View Status Screen x estne E enit b Adee tld vex antes Rb Exige 8 92 Unbind SCFeenzz cue ste ok iva hep Abeba Gee cy oe eed Sa aes 8 94 Cache Parameters Screen 0 0c cece eee teen e
246. des quick access to the most useful information X Firmware Update Updates SP microcode and disk module firmware X Expert Provides low level access to the SCSI disk array for X configuration and troubleshooting Information Tool The Information Tool displays SCSI disk array configuration and status information including e Auto trespass SCSI disk array and LUN e Caching parameters SP and LUN e Capacities LUN and disk module e Component status fan pack power supply units BBU etc e Date and time SCSI disk array e Disk module information by slot location vendor product identification bind slot e Firmware revisions disk module e Microcode revisions SP Support Tools Manager STM 7 99 sonsouDeig pue seni ISOH e LUN binding RAID type LUN number current default ownership associated disk modules e System serial number SCSI disk array e Hardware Path The Information Tool provides access to the following three logs and detailed descriptions of the contents of the Information Log Selecting any of the logs displays the information in anew window e Information Log e Activity Log e Failure Log To use the Information Log first build it with Tools Information Run The log displays after it is built If it doesn t display it with Tools Information gt Information Log The Information Log contains map configuration and disks information The Tools Informatio
247. dev rdsk c5t1d5 Creating Needed Directories for Volume Groups mkdir dev vg01 mkdir dev vg02 mkdir dev vg03 mkdir dev vg0l group c 64 0x010000 mkdir dev vg02 group c 64 0x020000 mkdir dev vg03 group c 64 0x030000 Creating Volume Groups In general volume groups encompass one or more physical volumes This example uses this capability to set up a path for one physical volume HP UX 9 04 vgcreate dev vg01 dev dsk c5d0S2 vgcreate dev vg02 dev dsk c5d1S2 vgcreate dev vg03 dev dsk c5d2S2 vgdisplay v dev vg01 vgdisplay v dev vg02 vgdisplay v dev vg03 In general volume groups encompass one or more physical volumes This example sets up primary and alternate paths for one physical volume HP UX 10 x vgcreate dev vg01 dev dsk c5t1d0 dev dsk cit2d0 vgcreate dev vg02 dev dsk c5t1d3 dev dsk clt2d3 vgcreate dev vg03 dev dsk c5t1d5 dev dsk cilt2d5 vgdisplay v dev vg01 vgdisplay v dev vg02 vgdisplay v dev vg03 7 130 LVM Logical Volume Manager Creating Logical Volumes The maximum size for a logical volume is 2 Gbytes In this example the smallest possible number of logical volumes have been created to take advantage of the total disk space The available disk space can be determined with the vgdisplay command shown previously In this example each physical extent is 4 Gbytes lvcreate lvcreate lvcreate lvcreate lvcreate lvcreate lvcreate lvdisplay lvdisplay lvdisplay lvdis
248. device C2300 Array on path 8 Specify the System Memory configuration values SP A Physical Memory MButes Total 64 Free 4 SP B Physical Memory MButes Total 64 Free 4 SP A System Buffer size Mbytes 2 SP B System Buffer size Mbytes 2 Optimized RAID 3 Memory MBytes Used 0 Free 0 Optimized RAID 3 Memory size MBytes 0 Write Cache size MBytes 38 SP A Read Cache size MBytes 20 SP B Read Cache size MButes 20 NOTE System Memory Configuration values can only be altered if certain cache options are disabled See the Help pages for more information WARNING Setting an SP System Buffer or Optimized RAID 3 Memory size to a value other than it s current setting will cause the SP s to restart which will v more wv X k v more wv w vy more wv ek EN an Figure 7 49 Utilities System Configuration System Memory Screen Support Tools Manager STM 7 125 sonsouBeiq pue seni ISOH Utilities System Configuration Host I F Settings e hpterm ioj x Expert Tool Host I F Options for device C2300 Array on path 8 4 3 This command will set change the Host Interface Settings fori Array Serial 9621954 Target Initiated Negotiation IP Enabled lt gt Disabled Substitute Busy for Queue Full lt gt Enabled X Disabled Mode Page 0x08 Support X Enabled lt gt Disabled Recovered Error Log Reporting lt K gt Enabled lt gt Disabled Allow Non Mi
249. diskinfo dev rdsk cb00df11s0 SCSI describe of dev rdsk cb00dflls0 vendor DGC product id C2300WDR1 type direct access size 2036569 Kbytes bytes per sector 512 Then the vendor product id and size listed in the diskinfo output must be used in the newfs command In the next example newdisk was used to create an entry for a RAID 1 LUN it found on the bus DGC_C2300WDR1_noswap in the example which is about 4 Gbytes in size as indicated by the sO field of s0 4102656 DGC_C2300WDR1_noswap DGC_C2300WDR1_noreserve DGC_C2300WD_noswap DGC _C2300WD_noreserve No swap or boot ns 32 nt 16 nc 8013 s0 4102656 b0 8192 f0 1024 se 512 rm 3600 In the next example newdisk was used to create an entry for a second RAID 1 LUN it found indicated by inserting the approximate disk capacity after the disk type name DGC_C2300WDR1_1 9GB_noswap in the example This RAID 1 LUN is about 2 Gbytes in size not 4 Gbytes as was the first LUN The additional identifier corresponds to the size value in the s0 field of the disktab entry which is s0 2036224 in the example 8 102 Creating a File System on a Series 700 System DGC_C2300WDR1_1 9GB_noswap DGC_C2300WDR1_1 9GB_noreserve DGC_C2300W D_noswap DGC_C2300WD_noreserve No swap or boot ns 32 nt 16 nc 3977 s0 2036224 b0 8192 f0 1024 se 512 rm 3600 With all this information the newfs command can now be used to create a file system that uses the full capacity of this
250. ditions and Messages GridMgr iX should run without encountering any errors Only one of two errors should ever occur The first error can occur when a user without the required capability level attempts to start GridMgr iX This has already been explained e The second error can occur when a user attempts to access one of the disk array s Grid Manager RS 232 ports at the same time someone else is accessing that port GridMgr iX gets exclusive access to the disk array s Grid Manager RS 232 port to prevent two users 6 14 Initiating Grid Manager on MPE AX from making configuration changes to a disk array device at the same time Access is granted on a first come first served basis Figure 6 6 shows the error message GridMgr iX displays when a second user attempts to access a Grid Manager RS 232 port already in use run gridmgr pub sys GridMgr iX A 01 01 C COPYRIGHT Hewlett Packard Co 1996 FRI APR 26 1996 8 30 AM There are 29 Grid Manager Disk Array Ldevs configured on this system Their Ldev numbers are 102 103 104 106 107 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 35 316 317 318 319 320 321 322 323 Which of these Disk Array Ldevs do you want to use You can select Ldev number zero to mean none of them Ldev number 106 MANAGER PUB SYS on Ldev 5 GridMgr Nike Ldev 106 User Ldev 5 KRKKKKKKKKKKKKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK k When you are done and
251. duct Model Number 7341 Revision 7 23 0 Major Minor 23 Pass 0 Number of Hard Errors 0 0x0 Maximum Outstanding Requests 1 0x1 Average Outstanding Requests 0 0x0 Busy Ticks 463 x 100ms Idle Ticks 4183218 x 100ms Number of Reads 0 0x0 7 58 NIKEARRY Host Utilities and Diagnostics 69 2 AIV AMIN S1012eg swOOT X 9876 8TPI NYA STU YFM SYOTL IX0 I senen oadezuou M STLATIIY IX0 T su3bueT enend Jo ung TP0000000000000 0 8Sc pexees s xoo gd jo wns 0X0 0 U9343T4M SX92OITH JO Tequinn X0 peeu SOOT jo zequnN 0x0 0 S1032eg peddeusy 1X0 I ronend ur sqysenbey bay 33333333X0 Be enen ur sqsoenboy xeWw 0006X0 7989 eoedg ejZeaTIg JO S30329g E9PIIXO 6S0EL eoedg Zesn go saioq3oeg zP9qd9 X0 v 00IIY QHU UO sro IDES 0X0 seTajey eqtim Jo aequnw 0X0 0 fSer423eu peeu jo zequnwN 0x0 0 SeqTaim JO azequnN X0 Speeu Jo zequnn S2103459S 6III 1Soouo1gogjgrq 1I079 DAY sug eurp eorAdgeg bay 0x0 0 S2O042328 931T2M 470S 0x0 0 S30422154 peeu 470S 0x0 0 SIO4 9314M PILH 0x0 0 STOIIW 3esoeu pIe perdeue da snqeas nua IX0 I qI 92rAed ISOS 0x0 0 1equnN sng ISOS 9X0 9 equnN NAI x 009683X0 ZSEZ6Z9T Aqroede2 zesn x IJ TTOI73U00 styuy Aq p umo 3TU x p rqeug juswubtssy ojny x 08X0 8ZT ZTS JU UW TA x s arvH Aeaav 4STd ed L p ta OG punog ebejusecozeg x GGZ prrnqew ebequeod8d x SxXSIp G I sq
252. dule is present applications continue running otherwise the disk array fails Service provider replaces power supply and System Administrator restarts the operating system if necessary SCSI 2 adapter cable No If the adapter cable fails I O operations to array disk units fail Authorized service provider replaces adapter or cable and System Administrator restarts operating System and applications Can regain access to LUNs associated with failed SP by manually making changes in Grid Manager and LVM Two SP Configuration 3 9 002 seues 0006 dH uoneunDuuo25 Multi Bus Configuration HP UX 10 x and Later The multi bus configuration shown in Figure 3 3 has two hosts each with a SCSI 2 adapter connected by an external Fast Wide differential SCSI 2 bus to an SP in the disk array The SP that binds a physical disk unit LUN is the default owner Each host uses its own disks in the disk array independently The multi bus configuration resembles two basic configuration systems side by side Table 3 4 contains error recovery procedures HP recommends that auto trespass be disabled when using multi bus configuration on a Series 700 This configuration can be used for either a model 10 or a model 20 disk array One host is connected to the array by way of SP A and the other host is connected to SP B The RS 232 cables are connected to their own array terminals In this configuration one disk array can support tw
253. dules After unpacking and setting up the disk array hardware which should be done by an HP service representative or by following the installation procedures in chapter 2 locate the following e A5061 5003 cable e Aterminal capable of EM100 mode of operation e The Grid Manager chapter of this manual chapter 6 3 2 Disk Array Support Requirements for HP UX Release 9 05 9 07 Connect one end of the 5061 5003 cable to the serial port on the back of the disk array Each SP has one port A port is located immediately below on the deskside model 10 or to the left on the rackmount model 10 of the SP A or SP B connectors or above on the deskside model 20 or to the right on the rackmount model 20 of the SP A or SP B connectors See Figure 2 12 for SP console port locations Connect the other end of the cable to the PORT 1 connection on the terminal Set up the terminal for EM100 mode of operation refer to the terminal manual for information Set the BAUD rate to 9600 and the parity to NONE 8 Once the terminal is set up power up the disk array The terminal s display screen indicates the disk array is booting by cycling through the alphabet from A to Z Once this completes a LICENSED INTERNAL CODE message displays followed by the Grid Manager Presentation Utility screen After a few minutes the Presentation Utility screen displays disk array status information For an example of Presentation Utility screen output see Figure 6 8 i
254. dvantage of mirroring Planning Disk Unit Configurations 1 27 uonduioseg jonpold Stripe Blocks i User Data Q First module of 384 511 768 895 1152 1279 1536 1663 a primary image Stripe Q O O O ee element size Second module of 512 639 O primary image Third module of 640 767 O O primary image Stripe size First module of 384 511 768 895 1152 1279 1536 1663 i O secondary image Second module of secondary image 512 639 Third module of secondary image 640 767 Figure 1 14 Distribution of User Data in a RAID 1 0 Group Figure 1 14 illustrates the distribution of user data with the default stripe element size of 128 sectors 65 536 bytes in a six module RAID 1 0 group The disk block addresses in the stripe proceed sequentially from the first mirrored modules first and fourth modules to 1 28 Planning Disk Unit Configurations the second mirrored modules second and fifth modules to the third mirrored image modules third and sixth modules then from the first mirrored modules and so on A RAID 1 0 group can survive the failure of multiple disk modules provided one module in each image pair survives For highest availability and performance the modules in each image pair must be on different internal SCSI 2 buses RAID 1 0 for a Model 10 Array When planning a RAID 1 0 group for highest availability and best performance it is necessary to have t
255. e Check Power Requirements Unpack the Disk Array Install a Rackmount Disk Array in a Cabinet Install Additional SCSI 2 Host Adapter Cards on the Host Determine Disk Array SCSI 2 Bus Lengths Determine the Locations of SCSI 2 Bus Connectors Connect the Disk Array s to the Host Computer Set SCSI IDs for SPs in the Disk Array s Connect Asynchronous Console Cables Connect the Power Cord Power Up the Disk Array Disk Array Installation Troubleshooting Checklist Revision 1 3 2 1 uonejjeisuj Overview of Installation 2 2 Check the software Check the power requirements Unpack the disk array s Unpack a deskside disk array Unpack a rackmount disk array Install a rackmount disk array in a cabinet Verify disk modules are properly seated in their appropriate slots whenever the disk array is moved Install additional SCSI 2 host adapter cards in the host s Determine SCSI 2 bus lengths Determine the locations of the SCSI 2 bus connectors Connect the disk array s to the host s Connect the SCSI 2 cables to a single host Connect the SCSI 2 cables in two host configurations Set SCSI IDs for the SPs in the disk array s Connect the asynchronous console cables Connect the power cord s Power up the disk array s Disk array installation troubleshooting checklist Overview of Installation Check the Software HP UX 9 04 Operating System If run
256. e Download operation SUCCESSFUL HARRAY Section 17 External Exerciser or 99 to exit hit CR to abort download 1 22 gt firmware controller s Updating Disk Firmware Table 7 2 Disk Drive Firmware Revisions 1 Run SYSDIAG Remember to type SUPLICEN 2 Run HARRAY section 17 on the desired PDEV Drive Description Download file Revision ST12550N 1 6 2 Gbyte Seagate Drive c2hhhp03 fdf HP03 ST32550N 1 2 Gbyte Seagate Drive c2lphp08 fdf HP08 ST15150N 1 6 4 Gbyte Seagate Drive c4 hp03 fdf HP03 DCHS04F 1 4 Gbyte IBM Drive ibm 4gb Ip hp53 wrp frm HP53 DGHSO4Y 1 4 Gbyte LP IBM Drive ibm 4gb lp hp3c wrp frm HP3C DCHSO9F 1 6 9 Gbyte HH IBM Drive ibm 9gb hh hp53 wrp frm HP53 DGHSO09Y 1 9 Gbyte LP IBM Drive ibm 9gb lp hp1a wrp frm HP1A DNESO9Y 1 9 Gbyte LP IBM Drive ibm 9gb lp hpjO wrp frm HPJO DGHS18Y 1 6 18 Gbyte HH IBM Drive ibm 18gb hh hp1a wrp frm HP1A DNES18Y 1 6 18 Gbyte LP IBM Drive ibm 18gb lp hpjO wrp frm HPJO At this point on a dual controller SP the peer SP must be disabled during the download For example if SP A is selected in the diagnostic SP B must be disabled If SP B is selected disable SP A See Enabling Disabling the SP620 SCSI Storage Processor in chapter 6 for information on how to disable the SPs m ome Run the DOWNLOAD command from the HARRAY prompt Type the firmware file name when prompted for it Re enable the pe
257. e Product ID HP A3232A RAID 5 Physical Device 48 0 1 Disk Position B2 The array and the slot number in the array containing the failed device are reported In this example 48 0 1 is the hardware path to the array that contains the failed disk module B2 is the slot number that contains the failed disk The disk module may not be part of the LUN at address 48 0 1 Fan Failure Message If HARAYMON detects a fan unit failure it displays a message similar to the following example on the console and e mails a copy to the specified user Mon Jul 11 10 23 17 MST 1994 High Availability Array Monitor Fan Unit Failure Product ID HP A3232A RAID 5 Physical Device 48 0 1 Fan Number B Controller Failure Message If HARAYMON detects a controller failure it displays a message similar to the following example on the console and e mails a copy to the specified user 7 8 High Availability Disk Array Monitor HARAYMON Mon Jul 11 10 23 17 MST 1994 High Availability Array Monitor Peer Controller Failure Product ID HP A3232A_RAID_5 Physical Device 48 0 1 Controller A The array and controller number in the disk array are reported In this example 48 0 1 is the hardware path to the disk array containing the failed controller and 48 0 1 is the hardware path to controller B which is reporting that controller A has failed Power Supply Failure Message If HARAYMON detects a power supply VSC failure it displays a message similar to th
258. e Password menu is entered Grid Manager Inventory Screens System inventory and physical disk inventory information is available from Grid Manager in FE mode To see the system inventory information go to the Grid Manager Presentation screen then type i The following information is displayed su xterm System Inventory Part Number Qty Description SCSI Disk Array Ctlr Phoenix class sni SCSI Disk Array Ctlr Phoenix class sni SCSI Disk Array RAH Hegabutes SCSI Disk Array VSC Module SCSI Disk Array Fan Module SCSI Disk Array BBU 2 1 GB SCSI Disk SEAGATE 2 1 GB SCSI Disk SEAGATE 1 1 8 3 2 1 4 4 _ Hit any key to continue Figure 6 20 System Inventory Information Grid Manager Inventory Screens 6 73 JoDeue N piis To see the physical disk inventory information go to the Grid Manager Presentation screen and type I The following information is displayed amp xterm PHYSICAL DISK INVENTORY Slot Disk Hfr Hodel No I ST32550N SERGRTE ST12550N SERGRTE ST12550N SERGRTE ST32550N SERGRTE ST32550N ST12550N HP03 ST32850N HP08 AO BO Co po EO AL BL C1 n E1 ge p e p E2 fis Bs 3 B E3 Hit any key to continue l Figure 6 21 Physical Disk Inventory Information 6 74 Grid Manager Inventory Screens HOST UTILITIES AND DIAGNOSTICS Overview IOMAP High Availability Disk Array Monitor HARAYMON HARRAY Section 10 D
259. e be sure to follow the instructions that accompany this symbol wherever it is found in this manual Service Any servicing adjustment maintenance or repair must be performed only by authorized service trained personnel Format Conventions Denotes WARNING A hazard that can cause personal injury Caution A hazard that can cause hardware or software damage or data loss Note Important information or operating instructions this font Text to be typed verbatim all commands path names file names and directory names this font Text displayed on the screen THIS FONT Keys to be pressed for example the RETURN key CONTENTS Reference Features s us eaten satin Wares teed e ete EGER NER OR des R 2 Models Options and Add On Products llle R 5 Models Available 3 3 2 ose echy dg senex bg CREER SEDES UR R 5 Models Discontinued June 1 1998 0 cece ee eee R 5 Options Available 2 c xl xd gk dx lx 4l lee R 5 Storage Capacity Options 0 cece eee e R 5 High Availability Options 0 0 eee RI R 5 Patch Media Options 0 cece eee eee e R 6 SCSI2 Cable Options eseeeeeeeee RII R 6 Upgrade and Add On Products 0 cee cece eee eee neen R6 Specifications umm RS EDNer4eeEXENSi UseseRr ENS RENS R 8 Dimensions es mI due M oth dad send SEMEN UE RAE UA R 8 Weighiz oec ec bie m eds EXE este uere e s R 8 AC Power Input Specifications 0
260. e following example on the console and e mails a copy to the specified user Mon Jul 11 10 23 17 MST 1994 High Availability Array Monitor VSC Module Failure Product ID HP A3232A_RAID_5 Physical Device 48 0 1 VSC Number A BBU Failure Message If HARAYMON detects a battery backup unit failure it displays a message similar to the following example on the console and e mails a copy to the specified user Mon Jul 11 10 23 17 MST 1994 High Availability Array Monitor Battery Backup Unit Failure Product ID HP A3232A RAID 5 Physical Device 48 0 1 High Availability Disk Array Monitor HARAYMON 7 9 sonsouDeig pue seni ISOH HARAYMON Error Reporting In the following example messages two LUNs are reporting a disk module failure in slot B2 Example 1 Mon Feb 20 10 05 12 MST 1995 High Availability Array Monitor Drive Failure Product ID HP A3232A RAID 5 Physical Device 48 0 0 Disk Position B2 Example 2 Mon Feb 20 10 05 12 MST 1995 High Availability Array Monitor Drive Failure Product ID HP A3232A RAID 5 Physical Device 48 0 1 Disk Position B2 These two error messages are reporting that both LUN 0 and LUN 1 were active when the disk drive in slot B2 failed Two separate LUNs cannot use the same disk module but because both LUNS were active at the time of failure they both reported the failure to HARAYMON The result is that HARAYMON reports the disk failure twice The failed disk module in slot B2 should be
261. e 40 Mbyte Read Cache Size 21 Mbyte 21 Mbyte 20 Mbyte 20 Mbyte Optimized RAID 3 memory size 0 Mbyte 0 Mbyte 0 Mbyte 0 Mbyte Available Memory 0 Mbyte 0 Mbyte 0 Mbyte 0 Mbyte To Specify Storage System Memory Parameters 1 From the Grid Manager Main menu follow the menu path 6 60 Specifying Memory Configuration for Microcode Revision 8 xx 9 xx 5 Change Parameters gt 2 Change Storage System Parameters 2 Change Storage System Memory Configuration 2 To change the parameters answer the screen prompts e xterm SP A SP SCSI ID 15 Change Storage System Hemory Configuration S N 9621954 fit any prompt enter to GO BACK TO PREVIOUS HENU for HELP Current system memory configuration Total Physical Hemory Systema Buffer Size Write cache size Read cache size Optimized RAID 3 memory size Available Hemoru S oona aR EEEEL cob han ECEE Write cache size in HB 14 SPA Read cache size in HB 141 SPB Read cache size in HB 141 Optimized RAID 3 memory size in HB 01 Figure 6 15 Storage System Memory Parameters Table 6 9 and Table 6 10 show the respective system memory configuration sizes for the SP and the SP620 SCSI SP Table 6 9 Change Storage System Memory Configuration JoDeue N pud Total cache Buffer Write cache Read cache 4 Mbyte 2 Mbyte 1 Mbyte 1 Mbyte 8 Mbyte 2 Mbyte 4 Mbyte 2 Mbyte 16 Mbyte 2 Mbyte 9 Mbyte 5 Mbyte 32 Mbyte 2 Mbyte 20
262. e 8 13 weight R 9 SP error codes error codes 8 56 fatal error codes 8 56 informational codes 8 56 thresholded codes 8 56 SP6xx SCSI SP enabling disabling 6 42 specifications R 8 weight BBU R 9 disk module R 9 Index l 7 xepu fan pack R 9 power supply unit R 9 SP R 9 specifying cache page size and cache size 6 56 statistics logging on Presentation screen 6 35 STM Support Tools Manager 7 68 7 126 auto assign 7 85 7 114 auto trespass 7 85 7 114 BBU test time 7 85 7 114 binding LUNs 7 85 7 114 cstm 7 68 default ownership 7 114 default SP 7 85 disabling cache 7 85 7 114 disk array set date amp time 7 85 7 114 set host I F options 7 85 7 114 set serial number 7 85 7 114 set systerm type 7 85 7 114 enabling cache 7 85 7 114 mstm 7 68 7 99 7 126 mstm Expert Tool 7 111 7 126 mstm Firmware Update Tool 7 107 7 111 mstm Information Tool 7 99 7 107 rebuild time 7 85 7 114 system caching 7 85 7 96 7 114 7 124 system memory 7 85 7 97 7 114 7 125 unbinding a LUN 7 85 7 114 xstm 7 68 7 69 7 98 xstm Expert Tool 7 82 7 98 xstm Firmware Update Tool 7 78 7 82 xstm Information Tool 7 69 7 78 storage capacity options R 5 Storage system Options Package auto trespass 6 65 storage system memory configuration 6 60 stripe size 1 24 SUSPEND HARRAY command 7 43 SYSMAP 7 2 system caching in STM 7 85 7 96 7 114 7 124 1 8 Index system memory in STM 7 85 7 97 7 114 7 125 system status on Presentation
263. e 8 31 SIMM Troubleshooting for the SP620 SCSI SP 8 32 SCSI 2 Bus Termination Power eeeeeeeeeeeeee een 8 34 Upgrading from an SP to a SP620 SCSI SP 0 2 eee eee eee 8 35 Cache Configuration Instructions 2 0 2c cece eee eee 8 38 Power Block Diagram 0 0c cece ete Ih 8 40 Disk Array Power Troubleshooting 2 0 cece ee eee eee 8 41 RIP PCB Troubleshooting Model 20 Disk Array Only 8 42 Location of RIP PCB LEDs 0 00 eee eee eee eee nee 8 43 Viewing Maintenance Information 0 c eee eee eee eee eee 8 44 Display Unsolicited Event Log 00 eee eee eee ee 8 44 Display Faulted CRUS 00 cee ec eee eee eee neee 8 46 Display Verify Results sere paanid ne cece aa e ee neee 8 47 LUN Configuration Summary 0 cee ee eects 8 49 Disk Failure Information 00 0 cece eee eee een eee 8 50 Non Vol RAM Error Log eenean onee a TEETE E e 8 52 Unsolicited Event Log 0 0 eee cece ee II 8 53 Displaying the Disk Array Unsolicited Event Log 8 53 xiii U9 U09 Controlling the Display of Event Records 0 2 cece eee eee 8 54 Filtering Event Records 0 0 e eect een eee 8 54 SP Event Codes ea a a oe sa EU Seer Shien needs 8 56 Rebuild AbOrts rect hice eth gt ea eta Deeg id eat Graded eee ed 8 65 Recovered Error Log Rep
264. e 9 62 9 103 On Off Switch Figure 9 63 ssssseeeeeeeeee ee 9 106 Glossary Index FIGURES 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 1 10 1 11 1 12 1 13 1 14 1 15 1 16 1 17 1 18 1 19 1 20 1 21 1 22 1 23 Parts Breakdown for the Model 10 Disk Array 2 20000 R 14 Parts Breakdown for the Model 20 Disk Array lesse esee R 17 Model 10 and Model 20 Disk Arrays esleeleeeeeee eee 1 3 Model 10 Disk Array Chassis 0 0 cee cece IA 1 4 Model 10 Internal Hardware Architecture leeeeeeeeeeeeeeeee 1 5 Model 20 Disk Array Chassis esses 1 7 Model 20 Internal Hardware Architecture 0 0 eee eee ee eee 1 8 Disk Mod le siaaa airn Agere eit va sea INE RAI enge hie REDE 1 10 Storage Control Processor 4 secede dog Pe sd e ERR rem rr RR exu 1 12 Model SP620 SCSI Storage Control Processor 0 c eee eee eee eee 1 12 Fan Packs for the Model 10 and Model 20 Disk Arrays 1 14 Power Supply Uniti bonsai e ck es eee ee I Ret a oe eee ee 1 15 Battery Backup Unit 00 0 ccc cece eee Ih 1 16 Distribution of User Data in a RAID 0 Group 0 00 02 eee eee eee 1 25 RAID 1 Mirrored Pair Hardware Mirrored Pair 006 1 26 Distribution of User Data in a RAID 1 0 Group 00 0 0 eee eee 1 28 Mirrored Modules in a RAID 1 0 Group in a Model 10 Disk Array
265. e Disable Periodic Error Reporting Displays dialogue for controlling periodic error reporting 3 Change Console Parameters Displays several modifiable console parameters default is 9600 baud 8 bit no parity 4 Update SP PROMs and Reboot SP Displays dialogue for updating SP PROMs 5 Soft Enable Disable Displays dialogue for enabling disabling the SP620 SCSI controller 6 Change Host Interface Options Addressing model LUN TARGET LUN Target initiated negotiation DISABLED ENABLED ENABLED Substitute busy for qfull DISABLED ENABLED DISABLED Mode page 8 DISABLED ENABLED DISABLED Recovered error log reporting DISABLED ENABLED ENABLED Allow non Mirrored Cache DISABLED ENABLED DISABLED Auto Trespass DISABLED ENABLED ENABLED 6 View Cache Statistics Displays dialogue for displaying cache statistics Grid Manager Menus 6 27 Jobeue N pus FE 7 Back End Exerciser Displays dialogue and instructions for running back end exerciser FE Must be in FE mode to access these screens or functionality To enter FE mode type CONnTROL P followed by in upper case FE To return to user mode type CONTROL P followed by in upper case USER Grid Manager Menus for Microcode Revision 9 xx Menu number from the Grid Manager Main Menu 1 Presentation Utility Displays the Presentation screen 2 Bind Logical Units 1 RAID 5 Group Individual Access Array Displays dialogue for making a RAID 5 LUN 2 R
266. e Inner Metal Cover 2 cece eee eee eee een eens 9 33 Removing the Inner Chassis eeeeeeeeeeee ee 9 34 Removing the IPCB i ceteRC P3 cet ee eee oe det exui tef us NN 9 36 Removing the LEDs LED Cable eeeeeeeee IIIA 9 38 Removing the Fan Cable eeseeeeeeeeeee Ih 9 40 Removing the SCSI 2 Cables sseeeeeeeeeeee e 9 42 Removing the AC Harness Bracket 2 0 cee cece eee een eens 9 44 Removing the AC Harness Plugs 0 cece cee eee eens 9 45 Removing the On Off Switch 0 0 2 0 eee cece ee 9 47 Removing the AC Filter Model 10 00 0 eee eee eee eee eee 9 49 Removing the Array from the Rack 0 c eee eee eee eee 9 51 Removing the Top Panel 0 eee eee eee nent eens 9 53 Removing the Inner Chassis seeeeeeeeeee eA 9 54 Removing the IPCB e saccade erbe Sees ie ee a eNFdFbeknees ida 9 56 Removing the LEDs LED Cable sseeeeeee RII 9 58 Removing the Fan Cable 0 0 eee ence een ene eens 9 61 Removing the SCSI 2 Cables 00 0 eee eee een eee nes 9 62 Removing the AC Harness Bracket 2 0 cee ee eee ee ee eee eens 9 64 Removing the AC Harness Plugs 2 0 c cece eee eee eee ene 9 65 Removing the On Off Switch 00 0 eee eee e 9 67 Removing the AC Filter Assembly 0 0 e eee eee eee eee 9 69 Removing the Outer Covers 2 0 0 cc
267. e SP microcode revisions and their corresponding PROM codes at the time of publication Refer to Service Notes for updates Revision 7 16 Revision 7 23 Revision 8 20 Revision 8 27 Revision 8 51 Revision 8 56 Revision 8 58 PROM 7 977 PROM 7 977 PROM 7 99 PROM 8 02 PROM 1 49 PROM 1 56 PROM 1 57 Revision 9 26 07 PROM 8 02 Revision 9 56 07 PROM 1 73 4 Gbyte HH half height disk support for the SP on model 10 and model 20 only 2 Gbyte disk support for the SP only for the SP only 4 Gbyte LP low profile and 9 Gbyte HH disk support for the SP on model 10 and model 20 only 4 Gbyte HH disk support for the SP620 SCSI SP only for the SP620 SCSI SP only 4 Gbyte LP and 9 Gbyte HH disk support for the SP620 SCSI SP only 4 Gbyte LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP on model 20 only 4 Gbyte LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP620 SCSI SP only Troubleshooting the Disk Array 8 19 punoouse qnouj The following table shows the minimum supported microcode and PROM revisions required for each disk module supported on the High Availability SCSI disk array Table 8 6 Minimum Microcode and PROM Revisions for Supported Disk Modules Model 20 with SP620 Model 10 Model 20 SP Disk Disk module Capacity microcode PROM Microcode PROM Microcode PROM revision revision revision revision revision revision A3240A 2
268. e Same commands to a Fast RAID 3 LUN would result in 0x00280002 or 0x00220055 panics Fixed in microcode revision 9 xx Unsolicited Event Log 8 81 punoouse qnouj Table 8 17 Microcode Panic Codes Not a Complete List cont d Value 0x0028000F Description Heavy I O with unusual stripe size Occurred while executing heavy I O to a RAID 5 LUN with an unusual stripe size setting such as 99 Fixed in microcode revision 9 xx Binding a RAID 5 LUN containing something other than five disks and a non standard stripe size could cause this when the host writes to the LUN Fixed in microcode revision 8 20 0x00300000 Invalid memory sizes Invalid memory sizes specified in mode page 0x2A could result in a panic instead of rejecting the request 0x0040300E Inaccessible LUN A power failure trespass or SP fault with a high SP cache dirty page count under a heavy I O load might cause one or more LUNs to become inaccessible Cycling power did not clear this state A limited number of data sectors could be unrecoverable Fixed in microcode revision 7 23 0x00501001 Parity error Firmware revision greater than 8 00 A SCSI controller chip detected a parity error during a transfer from data RAM Reference the 0x00000012 code 0x00501003 Chip and driver incompatibility Occasionally the microcode would panic with a 0x00501003 or 0x00503003 code This was caused by a known problem with the Symbios 53C810 SCSI pr
269. e differential SCSI 2 bus connectors for the SPs are on the back of the disk array oc Em nr Rear Views SCSI A SCSI B IN IN SCSKA G SCSHB G ont ne on sas e oDC oI gt Aas SCSIA SCSIB n eo e SCSI A SCSI B OUT OUT Figure 2 6 SCSI 2 Bus Connector Locations Model 10 Determine the Locations of SCSI 2 Bus Connectors 2 23 uonejjeisuj n olf fio eS SCSI B IN m AY SCSI B OUT eim oO o Wu gi NE dis ME Es ss T BN eo e ou rJ z x HH lt scsiain BO 2 SCSI A OUT Bi QS eo Rear View JT U Rear View SCSI A IN SCSI B IN ee SCSIA G SCSIB 7 e E Ce 8 CI SS 225 l scsiA scsiB Ss cie 2 n SCSI A OUT SCSI B OUT Figure 2 7 2 24 SCSI 2 Bus Connector Locations Model 20 Determine the Locations of SCSI 2 Bus Connectors Connect the Disk Array s to the Host Computer The procedure for connecting the disk array varies depending on the number of host computers in the configuration Note that multi bus multi initiator configurations are not supported on MPE iX Multi initiator configurati
270. e drives are A0 A3 BO and CO At least one database disk module must be present for the power up sequence to complete The SP microcode is called Flare in the power up sequence Copyright c Motorola Inc 1988 1996 All Rights Reserved Copyright c CLARiiON Storage Systems Inc 1995 1996 All Rights Reserved Licensed material property of CLARiiON Storage Systems Inc and Motorola Inc Debugger Diagnostics Release Version 1 72 Initiating tests 182378 IRQ Interrupt Request eee eee Running PASSED Initiating autoboot Fetching Flare revision numbers Database Database Database Database Drive O0 Drive 1 Drive 2 Drive 3 9 45 0 48 9 45 0 48 9 45 0 48 9 45 0 48 Sector 520 Sector 520 Sector 520 Sector 520 Drives found with matching code type 4 The following table lists the power up tests in alphabetical not execution order Table 8 4 Disk Array Power Up Tests for the SP620 SCSI SP only Test ID Test Description AEM Append Error Messages Mode AM79C970 Ethernet Controller AM79C970 Tests DIR CEM Clear Error Messages CF Configuration Editor CL1283 Parallel Interface CL1283 Tests DIR CS4231 cs4231 Audio Codec DIR DE Display Errors DEC21040 Ethernet Controller DEC21040 Tests DEM Display Error Messages Power Up Troubleshooting 8 13 Buljooysajqnoly Table 8 4 Disk Array Power Up Tests for the SP620 SCSI SP only cont d
271. e eee 6 61 Change Storage System Memory Configuration on an SP620 SCSI SP 8 Mbyte Minimum 0 0 cece II tenes 6 62 ID Strings for Model 10 and Model 20 0 2 eee eee eee 7 3 Disk Drive Firmware Revisions 0 0 cece eee eee ene eee 7 23 NIKEARRY Commands 0 cece eee eee he 7 47 FRU Map All Disk Array Components at All Bus Locations 7 56 SEM Interfaces ac Re iR ECL Ses 7 68 Available Support Tools 0 2 0 cece ccc eect tenets 7 69 xstm Status Fields for the Information Log 0 ce eee ee eee 7 72 Disk State for the Map in the Information Log 0 0 ee eee 7 73 LUN Bind State for the Map in the Information Log 7 74 Information Log Screen Configuration Section Column Headings 7 75 Firmware Update Tool Menu Items and Descriptions 7 79 Expert Tool Menus and Descriptions 0 0 cece eee eee eee 7 84 Available Support Tools 2202 22 4 a bn ete ae hee ned Dice ba ea ae 7 99 xstm Status Fields for the Information Tool Log 1 102 Disk State for the Map in the Information Tool Log 7 103 LUN Bind State for the Map in the Information Tool Log 7 104 Information Tool Log Screen Headings 0 ce ee eee ee eee 7 105 Firmware Update Tool Menu Items and Descriptions 7 108 Expert Tool Menus an
272. e eee eo eek EE v ER EA 1 41 Performance olefeepua acdsee AIS Shek Pale LE Aa Ded og ae 1 41 Two SPs Load Balancing and Device Names 00 1 42 Completing the Configuration Worksheet 00 0202 eee eee 1 42 Installation Overview of Installation sleeeeeeeeeee e 2 2 Check the SoftWare e csere 2 cece cece eee eee eee n eee eens 2 3 HP UX 9 04 Operating System 1 0 0 2 ee eee ee ee eee eee 2 3 HP UX 9 05 and 9 07 Operating Systems 0 eee eee eee eee 2 3 HP UX 10 x and later Operating System 2 0 0c eee eee eee 2 3 MPEAX Operating System 00 0 cece eee eee een ene eens 2 3 Fast Wide SCSI 2 Host Adapter Firmware 00 02 e eee ee eee 2 4 HARRAY Diagnostic and HARAYMON Monitor 2 5 Check Power Requirements eeseseeeeeee e eee eens 2 5 Unpack the Disk Array 0 0 0 cece cence III 2 6 Unpack a Deskside Disk Array 0 0 cc eee cece 2 6 Check the Anti Tip Legs 0 20 e i ccc eee ene eee 2 7 Engage the Brake i onscetuleetRI e benei 9 NE pS ee A and redes 2 7 Check the Display Orientation Switch 0 0 c cece eee 2 8 Unpack a Rackmount Disk Array 2 0 cee eee eee eee 2 8 Install a Rackmount Disk Array in a Cabinet 0 00202 eee 2 10 U9 U09 Install Additional SCSI 2 Host Adapter Cards on the Host 2 19 Determine Disk Array SC
273. e fan pack When installing the fan cable in the chassis connect the fan end first so the red stripe faces to the outside of the disk array and the plug faces downward The fan cable must be in the cable retainers and must lie flat or the SP cannot be installed on that side Midplane connector Midplane Red stripe Figure 9 23 Removing the Fan Cable 9 40 Model 10 Deskside Component Removal SCSI 2 Cables Figure 9 24 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds Access the rear of the disk array Use a small screwdriver to remove two screws and lift the external SCSI connector and cable away from the IPCB bulkhead Remove the two retaining screws securing the IPCB to the chassis Figure 9 21 Pull both ejectors until they unlock and the IPCB starts out of the chassis Pull the IPCB out of the chassis making sure not to touch any electronic components or circuits Figure 9 21 CAUTION Follow ESD procedures when removing the IPCB 7 Remove the SCSI 2 cables a Pull them from their connectors Model 10 Deskside Component Removal 9 41 jueuie2e deH pue jeAoway 95 Retaining screws Figure 9 24 Removing the SCSI 2 Cables 9 42 Model 10 Deskside Component Removal Internal AC Harness Figure 9 25 and Figure 9 26 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds
274. e host with one SCSI 2 adapter connected by one external Fast Wide differential SCSI 2 bus to two SPs The two SP configuration provides better performance than the basic configuration For better SP performance with this configuration some physical disk units LUNs can be bound on one SP and the rest can be bound on the other SP The SP that binds a physical disk unit is the default owner of that unit The route through the SP that owns a physical disk unit is the primary route and the route through the other SP is the secondary route A secondary route is available on HP UX 10 x and later Host computer Disk array SCSI2 SCSI 2 bus System adapter i SPA console RS 232 SPB RS 232 System disk p To MUX port or terminal Figure 4 2 Model 10 20 Two SP Configuration Two SP Configuration 4 7 008 seues 0006 dH uoHeinByu05 Table 4 3 Two SP Configuration Error Recovery Failing component Disk module Continue after failure Yes 9 04 Yes 10 x What happens and how to recover With RAID of any level other than 0 or an individual disk applications continue running System Administrator or service provider replaces module Provides the same features as the basic configuration but with better performance and higher availability With RAID of any level other than 0 or an individual disk applications can continue to run after the
275. e microcode panic categories grouped by the value of their four high order hexadecimal digits Table 8 16 Microcode Panic Categories Value Category 0x0000nnnn Kernel services 0x0010nnnn Disk management upper level 0x0022nnnn Host read write request management 0x0024nnnn Host non read write request management 0x0025nnnn Unsolicited Event Log management 0x0026nnnn Grid Manager interface 0x0028nnnn RAID services management 0x0030nnnn Memory management 0x0040nnnn Disk management lower level 0x0050nnnn SCSI driver internal buses 0x0070nnnn Disk rebuild process Unsolicited Event Log 8 75 Buljooysajqnoly Table 8 16 Microcode Panic Categories cont d Value Category 0x0080nnnn Configuration manager 0x00AO0nnnn SCSI driver host bus 0x00COnnnn Disk verify process 0x00DOnnnn Disk zero bind process 0x00EOnnnn Board to board communications Ox00F2nnnn Non volatile RAM management Table 8 17 shows some specific panic codes This is not an inclusive listing Many panic codes fall into the same category The codes listed in this table provide more than general category information The inclusion of a panic code in this table does not imply that it is any more or less serious than the panic codes not appearing in this table Table 8 17 Microcode Panic Codes Not a Complete List Value Description 0x00000001 Memory exhausted 0x000000
276. e the event occurred with the most recent messages at the end of the log To display the last line type the letter L In normal user mode a shortened list of the most important events is displayed The comprehensive list of events is displayed in FE mode Unsolicited Event Log 8 53 punoouse qnouj Any event code whose first three characters are 0x9 or Oxa is an error code The error codes and corresponding messages are explained in the following SP codes tables Controlling the Display of Event Records To navigate the Unsolicited Event Log type the following single letter commands Key Action Norn Scrolls to the next screen of entries Porp Scrolls to the previous screen of entries Forf Displays the first screen of entries Lorl Displays the last screen of entries Aora Defines a filter to select event types by FRU or error code For help type A then A filter stays in effect until changed Corc Clears all entries from the Unsolicited Event Log Caution The host computer relies on entries to report subsystem errors clearing the Log may remove recent entries before they reach the host Dord Dump the contents of the Unsolicited Event Log to the screen The user is asked for starting and ending event numbers These must be valid numbers the end value must be greater than or equal to the start value or an error message is displayed The default starting and ending values are the fir
277. e true Data availability is very important Speed of write access is important Applications need logging or recordkeeping This is because fewer modules are required than in a RAID 0 group non redundant array and high availability and fast write access are provided RAID 1 is useful for storing daily updates to a database that resides on a RAID 5 group The updates to the database can then be copied during off peak hours RAID 1 0 Group Mirrored RAID 0 A RAID 1 0 group is recommended for applications in which the following are true Data availability is critically important Planning Disk Unit Configurations 1 39 uonduioseg jonpold e Overall performance is very important RAID 0 Group Non Redundant Individual Access Array A RAID 0 group is recommended for applications in which the following are true e Overall performance is very important e High availability is not important e The best overall performance is important A RAID 0 group is less available than an individual unit e The application uses short term data that requires quick access such as mail A RAID 0 group like a RAID 5 group requires a minimum of three disk modules Individual Disk Unit An individual disk unit is recommended for applications in which the following are true e Speed of write access is somewhat important e High availability is not important Print spooling user file exchange areas and availability are not important
278. ece cece eee ee een eens 2 26 Two Host Configurations 0 0 cece cc hh 2 28 SE SOSLID Modelt 107 unte Go a a AL et ees 2 30 SP SCSI ID Model 20 2 2 2 0 cece cece eet he 2 31 SP Console Port Locations 0 00 cee cece eee ene eens 2 33 Connecting the Power Cord and Enabling Power Typical 2 42 Model 10 20 Basic Configuration 00 0 cece eee eee eee nee 3 6 Model 10 20 Two SP Configuration 0 0 cece eee eee nee 3 8 Model 10 20 Multi Bus Configuration 0 0 0 0 ee eee eee eee 3 10 Model 10 20 Basic Configuration 0 0c cece eee ee 4 5 Model 10 20 Two SP Configuration 0 0 cece eee eee eens 4 7 Model 10 20 Two Adapter Two SP Configuration 00 0200 ee ee 4 9 Multi Bus Configuration 0 0 cece cece ee eee teenies 4 11 Model 10 20 Multi Initiator Configuration 0 cee eee ee eee 4 13 Model 10 20 Multi Bus Multi Initiator Configuration lesse 4 15 Model 10 20 Basic Configuration 00 cee cece ee ee 5 5 Model 10 20 Two SP Configuration 0 0 cee eee eee eens 5 7 Model 10 20 Two Adapter Two SP Configuration 02 0000 ee 5 9 Multi Bus Configuration 0 0 eee eee eee eee ene eens 5 11 Cable J2489A Pinouts 0 0 ec eee eben een e eens 6 6 6 2 6 3 6 5 6 6 6 7 6 8 6 9 6 10 6 11 6 12 6 13 6 14 6 15 6 16 6 17 6 18 6 19 6 20 6
279. ed it indicates that the retry did not work because the disk faulted The microcode sniffs the disk modules every five minutes or after an internal bus reset to determine if they are still functioning properly If the microcode discovers a malfunctioning disk module the failure is logged as an 0xA07 fatal error The microcode begins logging these errors at the 0x08 bit Usually the SP that doesn t own the malfunctioning disk module will report an error with less detailed status It doesn t know why the error occurred and only logs a generic error The owning SP will report an error on the same disk module with more detailed status For example SP B does a sniff of the disk array During the sniff SP B discovers that the disk module in slot AO is powered down Disk module A0 is owned by SP A The Unsolicited Event Log for SP A contains the error 0xA07 with an extended status of 0x09 CRU Enabled Dead SP B will report the 0xA07 with an extended status of 0x80 CRU Powered Down Table 8 15 Extended Status Codes Extended status code Description 0x00 No error information available 8 70 Unsolicited Event Log Table 8 15 Extended Status Codes cont d Extended status code 0x01 Description SCSI protocol error An error was detected and retried during activity on the SCSI bus between the SP and the disk itself 0x02 Device detected parity error Drive returned se
280. ee ee eee eee eee eee ee ee ee ee XO 0 vPl1 8 v OL 9340 TO GG6 dS eademuirj peoTUMOP O4 3nOQV x KKK KKK KK KK KK KK KK KK KK KK KK KK KK KK KK KKK KKK KKK KKK KKK KKK KKK KKK KKH KY AT ngsseoons pea3ooqeuH TYIN aqnutw opip c eqnuru goig ctr spuooes og tcn spuooes Op t 668 spuooes og t000666n spuooes oz 000008 spuooes oT t6 300qea O3 HMIN 30j Tem ssesetg OC OE OK OK XE OE OX OE KKK KKK XE OE OO KKK KK KKK KKK KKK KKH KKK KKK KK KKK KKK KKK KKH HH XX iATTNASSAOONS pepeo uMODp 9Z MULITA HK KKK KKK KK KK KK KK KK KK KK KK KK KKK KKK KK KK KKK KKK KKK KKK KKK KKK KKK KH HY OC OE OE OE QC OE OE OE eee ee eee eee eee eee ee eee eee eee ee eee eee eee eee eee ee ee X9 0 G1 8 v OT 9340 TO GG6 dS STeMUATZ PeOoTUMOP O4 3nOQV x KKK KKK KK KK KK KK KK KK KK KK KKK KKK KK KK KK KKK KKK KKK KKK KKK KKK KKK KH OX KKK KKK KK KK KK KK KK KK KK KK KK KK KK KK KK KK KKK KKK KKK KKK KKK KEKE KEKE KEKE iGHOVNVd dd AVN ASIA HHL HO SSHOOUd SIHI Ld WHHLNI ION OG x i ONINHVM x Ce a COOKIE OX OX OO OO OIX OM X amp b u A eaepdn eu quar enurauoo o4 usrA no og KKK KKK KK KK KK KK KK KK KK KK KK KK KK KK KK KK KK KKK EE EE EE KKK KKK KEKE KEKE KY T0 S96 dS YITM XINHOHd SZeTTOATIUOD 029dS TTe eaepdn o3 anoqv x KEK KKK KKK HK KKH KKK HK KKK KKK KKH KK KKK KKK KKK K HK KK HK KKK KKK KEKE X 7 66 KKKKKKKKKKKKKKK KKK KKK KKK KKK KKK KK KKK KKK KK KKK KKK KKK ck ck ck KKK sk sk sk kx KKK KKKKKKKKKKKKKKK KKK KKK KKK KKK
281. eee cece he 9 71 Removing the Inner Metal Cover 00 e eee cece eee een eens 9 73 Removing the RIP PCA 0 2 ce eee ec m nes 9 74 Removing the Midplane from the Chassis llle 9 74 Removing the RIP PCA 1 0 2 cenn de eee hh e 9 75 Removing the LEDs LED Cable 0 0 ccc eee RII 9 77 Removing the DC Power Distribution Cable 02 0202 9 79 xxiii sainbi4 9 47 9 48 9 49 9 50 9 51 9 52 9 53 9 54 9 55 9 56 9 57 9 58 9 59 9 60 9 61 9 62 9 63 xxiv Removing the Fan Cable 00 eee he 9 81 Removing the Address and SCSI 2 Cable 0 0 00 cece eee eee nee 9 83 Removing the AC Harness Bracket 0 0 0 c eee eee eee neces 9 85 Removing the AC Harness Plugs 2 0 0 0c ee eee eee eese 9 86 On Off Switch and Power Plug Assembly 0 2 2 eee eee eee eens 9 87 Removing the Model 20 Array from the Rack 00 02 ce eee ee eee 9 90 Removing the Top Cover sees 9 92 Removingthe RIP PCA nleieduk LE xDD LR ERE RE ee ee s 9 93 Removing the Midplane from the Chassis lel 9 93 Removing the RIP PCA 22 03 53 8 RR eR A ee Be 9 95 Removing the LEDs LED Cable esses 9 97 Removing the DC Power Distribution Cable 00 00 0020 9 99 Removing the Fan Cable eseseeeeeeeeeeee eh 9 100 Removing the Address and SCSI 2 Cable 0 ccc cece eee eens 9 102 Removing the AC H
282. eee eee 3 2 Disk Array Configurations 00 0 c cece cece eee ene ences 3 5 Basic Configuration for Model 10 and Model 20 Disk Arrays 3 6 Two SP Configuration onion eid eee yack oooh cc oa gk PAS Sa eet ie eed 3 8 Multi Bus Configuration HP UX 10 x and Later 000 002 eee 3 10 4 Configuration HP 9000 Series 800 Disk Array Configurations 00 cece e 4 2 Basic Configuration for Model 10 and Model 20 Disk Arrays 4 5 Two SP Configuration 00 cece ce Rh 4 7 Two Adapter Two SP Configuration 2 0 0 eee eee eee 4 9 Multi Bus Configuration 2 0 cece cee eee eee ne eeee 4 11 Multi Initiator Configuration HP UX 10 x and Later 4 13 Multi Bus Initiator Configuration HP UX 10 x and Later 4 15 5 Configuration HP3000 Series 900 MPEAX Considerations 00 0 cece eee cette tence eee 5 2 Disk Array Configurations 0 00 cece eee e 5 3 Basic Configuration for Model 10 and Model 20 Disk Arrays 5 5 Two SP Configuration 00 cece ehh 5 7 Two Adapter Two SP Configuration 0 0 ee eee eee ee eens 5 9 Multi Bus Configuration eleseeeeeeeeee eee nen ene eens 5 11 Configuring Disk Array SCSI LUNs for MPEAX 0 00 00 0020 5 13 6 Grid Manager About Grid Manager sese hh eee e nee 6 2 Configuring MUX Ports on HP 9000 Sy
283. een LED is on the left and the white LED in on the right 10 Remove the individual LED bezels The bezels are keyed with tabs that fit into grooves in the LED bodies See arrow 2 in Figure 9 33 11 Thread the LED cable back through the sheet metal opening See arrow 3 in Figure 9 33 12 Remove the LED LED cable by lifting up and out through the top of the bracket See arrow 4 in Figure 9 33 LED Bracket LED Cable Connector Midplane Connector Figure 9 33 Removing the LEDs LED Cable 9 58 Model 10 Rackmount Component Removal Fan Cable Figure 9 34 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 Remove the following a b g Fan pack Figure 9 6 SPs or fillers Figure 9 9 Power supplies or fillers Figure 9 11 Front panel Figure 9 18 BBU located in the lower front by pressing in on the two tabs at the sides of the unit Figure 9 13 All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array IPCB Figure 9 32 4 Remove the four retaining screws that hold the chassis in the rack Figure 9 29 WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array 5 Remove the empty chassis from the rack Figure 9 29 6 Remove 11 screws and remove the top pa
284. ejd y pue jeAoway Address and Internal SCSI 2 Cables Figure 9 48 Three cables are connected to the distribution panel two SCSI 2 cables and one address cable 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Remove the four retaining screws and lift the distribution panel away from the array chassis 3 Disconnect the address cable from the address connector on the back of the distribution panel Note When connecting the address cable to the address connector the red trace on the cable is positioned at the bottom of the connector When connecting the address cable to the midplane the red trace on the cable is positioned at the bottom of the connector This connector is not keyed The red trace wire should always face the down side on a rackmount model or face the right side of a deskside model Remove two screws from the top of the midplane and tilt it Disconnect the address cable from the midplane Disconnect the SCSI 2 cables from the midplane Disconnect the two SCSI 2 cable connectors from the distribution panel Poo Se ge cm Remove the cables from the chassis CAUTION Be careful when connecting the SCSI 2 cable to the midplane The connectors are 68 pin connectors and are fragile 9 82 Model 20 Deskside Component Removal Removal and Replacement SCSI cables Distribution Midplane Red trace Address Screws cable
285. emoving the RIP PCA Model 20 Rackmount Component Removal 9 95 jueuie2e deH pue jeAoway LEDs LED Cable Figure 9 57 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 Remove the following a Fan pack Figure 9 6 b SPs Figure 9 9 c Power supplies Figure 9 11 d BBU located under the power supply units Figure 9 13 e All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array f Front panel Figure 9 40 4 Remove the four retaining screws that hold the chassis to the rack 5 Remove the empty chassis from the rack Figure 9 52 6 Remove the top metal panel 11 screws Figure 9 53 7 Lift the LED cable out of the cable retainer The LEDs are soldered to the front of the cable 8 Unplug the LED cable from the midplane 9 Push the LED cable forward out of the bracket See arrow 1 in Figure 9 57 Note The green LED is on the top and the white LED is on the bottom on a deskside model On a rackmount model the green LED is on the left and the white LED in on the right 10 Remove the individual LED bezels The bezels are keyed with tabs that fit into grooves 9 96 in the LED bodies See arrow 2 in Figure 9 57 Model 20 Rackmount Component Removal 11 Thread the LED cable back through the sheet metal opening See arrow 3 in Figure 9 57 12 Remove the LEDs LED ca
286. en 2563443 Average Disk Service Time 14 ms Stripe Boundary Crossing 360 SP failures can occur if after a defective SP is replaced the other SP shows a BOUND BUT NOT ASSIGNED state for one of the LUNs it owns Recovery procedures follow First determine which LUNs are owned by each SP Go to the Grid Manager Main menu and follow the menu path 5 Change Parameters 1 Change Logical Unit Parameters 1 Change Default SP Ownership of Unit At the Logical Unit To Change hex prompt enter each LUN number that has been configured in the array As each LUN number is entered the SP that is the default owner displays Use this information to determine which SP lost ownership of a LUN For example type a 1 at the Logical Unit To Change hex prompt and Grid Manager indicates that SP A is the default owner of LUN 1 Go to SP A s Presentation screen and highlight a disk in LUN 1 and press RETURN The disk state reads BOUND BUT NOT ASSIGNED when it should read ENABLED This means that SP A has lost ownership of the LUN s When the SP that lost default ownership is identified unmount all filesets associated with that SP if possible If the filesets on the SP in question cannot be unmounted unmount all 8 88 Troubleshooting with Grid Manager filesets on the other SP and disable array cache Cycle the SP s enable disable switch or if the SP is a SP620 SCSI SP that does not have an enable disable switch remove
287. en nees 8 95 Bind a LUN Sere n cieie ee eet ne ederent eg ga 8 97 Binding Display Screen 0 0 eee ccc e nee 8 98 Parts Breakdown for the Model 10 Array 00 0 eee eee eee eee 9 2 Parts Breakdown for the Model 20 Array 00 0 cece e eee eee eee 9 5 Removing a Disk Module 00 eee ec ce een teen eee 9 9 Installing a Disk Module 0 cece cee ene eens 9 10 Model 10 20 Fan Packs 1 0 0 eee eect Ie 9 12 Removing a Pan Pack owen nce tsa eee ENA Oo etna E eed 9 12 Installing a Fan Pack 20 eee E 9 13 Setting the Enable Disable Switch llle 9 16 Removing an SP iles ce ELLA ERI LV RR Ads 9 17 Replacing an B sc c Sepe der we pK E RR p Yen 9 18 Removing or Installing a Power Supply Unit 0 0 20 eee 9 20 Location of BBUs in the Model 10 and Model 20 0 9 22 Removing or Installing a BBU 6 0 eee eee eee 9 23 SIMM Locations for the SP 0 2 eee cece ee 9 26 SIMM Locations for the SP620 SCSISP 2 cee eee eee eee ee 9 27 SIMM R mov l eserrz1efsbet e pee E xt Read hes ethan EI EE LES 9 27 SIMM Installation on an SP cece neee 9 28 9 18 9 19 9 20 9 21 9 22 9 23 9 24 9 25 9 26 9 27 9 28 9 29 9 30 9 31 9 32 9 33 9 34 9 35 9 36 9 37 9 38 9 39 9 40 9 41 9 42 9 43 9 44 9 45 9 46 Removing the Outer Covers 2 0 0 cece eee eect e 9 30 Removing th
288. eparately it is necessary to unbind the mirror losing all data on it rebind the module in the configuration desired and reformat the physical disk unit With a disk array RAID 1 hardware mirroring has the following advantages e Automatic operation no user issued commands required for initiation e Physical duplication of images e Auser defined rebuild period during which the SP recreates the second image after a failure For the most highly available mirror of any type the mirror on each disk module should be configured on a different internal SCSI 2 bus Q User data Figure 1 13 RAID 1 Mirrored Pair Hardware Mirrored Pair 1 26 Planning Disk Unit Configurations RAID 1 0 Group Mirrored RAID 0 Group In a RAID 1 0 group disk modules are bound as multiple RAID 1 mirrored pairs For example a six disk RAID 1 0 group consists of three mirrored pairs The number of disk modules in a RAID 1 0 group must be even and can consist of e Model 10 4 6 8 10 disk modules e Model 20 4 6 8 10 12 14 16 disk modules The disk modules make up mirrored images with each image consisting of two to five disk modules on the model 10 or two to eight disk modules on the model 20 The mirrored image consists of the same number of disk modules In this configuration the hardware automatically mirrors the modules in images A RAID 1 0 group uses block striping It combines the speed advantage of RAID 0 with the redundancy a
289. epartments and Agencies are as set forth in FAR 52 227 19 c 1 2 Copyright c 199_ Hewlett Packard Company All Rights Reserved 2 Customer further agrees that Software is delivered and Licensed as Commercial Computer Software as defined in DFARS 252 227 7013 or restricted computer software as defined in FAR 52 227 19 c 1 2 if used respectively in the performance of a Department of Defense or non Department of Defense U S Government contract Any term of this Agreement which is held to be invalid will be deleted but the remainder of the Agreement will not be affected Neither party may assign any rights or obligations without prior written consent of the other party Reference k D m Customer who exports Products assumes responsibility for complying with applicable laws and regulations and for obtaining required export and import authorization Customer will not export or re export the Software any copy or adaptation or any technical data in violation of any applicable laws or regulations Any disputes arising in connection with this agreement will be governed by the laws of California These HP Software License Terms supersede any previous communications representations or agreements between the parties whether oral or written regarding transactions hereunder Customer s additional terms and conditions will not apply Reference R 23 o2ue49JoH Regulatory Statements FCC Statement For U S
290. er SP after the download completes HARRAY Section 17 External Exerciser Type 3 download individual disk microcode to the Enter download type prompt 7 23 sonsouDeig pue seni ISOH Example Download type 3 HARRAY gt down PDEV 24 14 0 is an C1300WDR5 DOWNLOAD UTILITY 4 39 PM Types of Download 1 Download Controller Firmware Only 2 Download Controller Firmware AND Update Controller Prom Code 3 Download Disk Firmware Enter Download Type 1 3 or 99 to exit 1 3 ERROR Peer SP B has been detected as enabled Disk firmware downloads require peer SP be disabled Please disable peer SP now Press cr when disabled enter 99 to abort lt cr gt gt Enter Firmware Filename hit CR to abort download firmware disk c21php0 Download operation started Disk Download Successful Download operation SUCCESSFUL REMINDER Re enable peer SP B 7 24 HARRAY Section 17 External Exerciser EXIT Non exclusive non destructive The EXIT command exits HARRAY Example HARRAY gt exit End of section 17 INTERACTIVE EXERCISER End of pass number 1 HARRAY Exiting HARRAY has terminated HARRAY Section 17 External Exerciser 7 25 sonsoubeiq pue seninf ISOH HELP Non exclusive non destructive The HELP command provides information on the various HARRAY commands Example Several versions of the HELP command follow The HELP command without any options HARRAY gt HELP T
291. er of that unit The route through the SP that owns a physical disk unit is the primary route the route through the other SP is the secondary route A secondary route is available on HP UX 10 x and later Table 3 3 lists configuration error recovery procedures Host Computer Disk Array SCSI 2 SP A System Adapter RS 232 Console a v RS232 System Disk To MUX port or terminal Figure 3 2 Model 10 20 Two SP Configuration 3 8 Two SP Configuration Table 3 3 Two SP Configuration Error Recovery Failing component Disk module Continue after failure Yes What happens and how to recover Provides the same features as the basic configuration but with better performance and higher availability With RAID of any level other than 0 or an individual disk applications continue running System Administrator or service provider replaces module SP Yes With HP UX 10 x if one SP fails ownership of the failed SP s LUNs can transfer automatically to the remaining operational SP If a primary and an alternate path have been configured in LVM LVM can switch automatically to the alternate path LVM applies to HP UX 10 x and later These operations are transparent to the user Fan failure within fan pack Yes Applications continue running System Administrator or service provider replaces module Power supply Yes If a redundant power supply mo
292. erations are transparent to the user If the SCSI 2 adapter or external SCSI 2 bus fails no I O operations are possible between the host and the array 4 2 Disk Array Configurations Table 4 1 Array Configurations cont d Two adapter Two SP One host Two SCSI 2 adapters Two SCSI 2 buses Two SPs Provides highest availability and best array performance for single host configurations With RAID of any level other than 0 or an individual disk applications can continue to run after failure of a disk module HP UX 9 04 If one SP fails the System Administrator must reassign ownership of the failed SP s LUNs to the operational SP by using Grid Manager Also the path to the operational SP must be used and can be changed by way of LVM Once these changes have been made access to all LUNs on the array is possible If the SCSI 2 adapter fails the System Administrator can access the array through the operational SCSI 2 adapter s path LUN ownership and path changes need to be reassigned by using Grid Manager and LVM respectively HP UX 10 x If one SP fails ownership of the failed SP s LUNs can transfer automatically to the remaining operational SP If a primary and an alternate path have been configured in LVM LVM switches automatically to the alternate path If the SCSI 2 adapter fails transfer of LUN ownership and the LVM switch to the alternate path take place automatically These operations are transparent to the user
293. erial Number FRU Capacity 4110034 0x3eb6d2 Blocks Controller Private Space 36864 0x9000 Blocks Bind Signature 346914287 0x14ad7def Bind Time SCSI Bus 4 4 T Bind Time SCSI ID 0 7 56 NIKEARRY HELP The HELP command is processed initially at the ODE level and the syntax used should be as outlined in the ODE Manual The HELP command gives a condensed or expanded help on the NIKEARRY commands If HELP is typed without any parameters it gives out the list of commands supported by NIKEARRY Detailed HELP is available if HELP command is followed by the desired command Example NIKEARRY gt help utilinfo HELP for UTILINFO Utility Information SYNTAX UTILINFO INQUIRY The INQUIRY command issues a SCSI INQUIRY command to get the inquiry data block from the Logical Unit LUN under test INQUIRY takes a LUN index as a parameter If no parameter is provided the value of the TESTUNIT environment variable is used as the LUN index If TESTUNIT is not set the user is prompted for a LUN index LUN index is obtained from the list of LUNs displayed by the DISPMAP command Example NIKEARRY gt inquiry 0 Peripheral Qualifier LUN supported but not connected Peripheral Device Type Direct access device Medium is NOT removable Device type Modifier 0 0x0 ISO Version 0 ECMA Version 0 ANSI Version 2 Device complies to present SCSI 2 Device does NOT support TERMINATE I O PROCESS message INQUIRY da
294. es 26 7 cm 10 5 inches 76 2 cm 30 0 inches model 10 Rackmount 35 6 cm 14 0 inches 48 3 cm 19 0 inches 76 2 cm 30 inches model 20 Deskside 62 9 cm 24 8 inches 35 6 cm 14 0 inches 76 2 cm 30 inches model 20 Weight WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array R 8 Reference Model Minimum Weight Maximum Weight Model 10 Chassis with 3 disks 1 SP 1 power Chassis with 10 disks 2 SPs 2 power supply 51 6 kg 114 0 Ibs supplies and 1 battery backup unit 70 2 kg 156 0 Ibs Model 20 Chassis with 5 disks 1 SP 1 power Chassis with 20 disks 2 SPs 3 power supply 56 3 kg 125 0 Ibs supplies and 1 battery backup unit 83 9 kg 186 4 Ibs To determine the weight of a disk array add the weight of the following components to the weight of a minimum configured disk array Each power supply VSC 2 4 kg 5 3 Ibs Each 4 Gbyte disk module 1 6 kg 8 5 Ibs Each 4 Gbyte LP disk module 0 46 kg 1 0 Ibs Each 9 Gbyte HH disk module 0 82 kg 1 87 lbs Each 9 Gbyte LP disk module 1 0 kg 2 25 Ibs Each 18 Gbyte HH disk module 1 4 kg 8 05 lbs Each storage processor SP 1 2 kg 2 6 lbs Battery backup unit BBU 5 4 kg 12 Ibs Fan pack model 10 3 8 kg 8 5 Ibs Fan pack model 20 5 6 kg 12 5 Ibs AC Power Input Specifications Voltage 100 120 127 and 220 240 VAC 50 to 60 Hz s
295. es as the basic configuration but with better performance and higher availability With RAID of any supported level applications can continue to run after the failure of a single physical disk module If one SP fails the System Administrator must reassign ownership of the failed SP s LUNs to the operational SP by using Grid Manager Also the path to the operational SP must be used and can be changed with SYSGEN or ioconfig Once these changes have been made access to all LUNs on the array within the SCSI LUN limit is possible If the SCSI 2 adapter fails no I O operations are possible between the host and the array Disk Array Configurations 5 3 006 Seles 000 dH uoneunbyuo Table 5 1 Array Configurations cont d Two adapter Two SP One host Two SCSI 2 adapters Two SCSI 2 buses Two SPs Provides highest availability and best array performance for single host configurations With RAID of any supported level applications can continue to run after failure of a disk module If one SP fails the System Administrator must reassign ownership of the failed SP s LUNs to the operational SP by using Grid Manager Also the path to the operational SP must be used and can be changed with SYSGEN or ioconfig Once these changes have been made access to all LUNs on the array within the SCSI LUN limits is possible If the SCSI 2 adapter fails the System Administrator can access the array through the operational SCSI 2 adapte
296. es in redundant RAID groups such as RAID 5 or RAID 1 0 should be replaced only one module at a time The RAID group should complete its disk to disk copy onto the replacement disk module before another disk module is replaced 1 Observe antistatic precautions CAUTION It is recommended that the ESD wrist strap be grounded to the disk array chassis or suitable ground and strapped to your wrist when removing or handling a disk module and that the disk module be placed on the ESD conductive sheet when it is removed from its carton Use ESD kit P N 5182 4119 2 Remove a disk module Figure 9 3 Each disk module should already be labeled with the compartment number from which it was removed a Grasp the disk module by its handle and pull it partially out of the disk array Wait about two minutes to allow the disk to stop spinning before proceeding b Support the disk module with a free hand and pull it the rest of the way out of the disk array Set the disk module down gently Rough handling may damage the disk module 9 8 Model 10 20 Disk Module Removal Replacement 3 Install a disk module Figure 9 4 a Locate the slot where the disk module is to be installed b Ifinstalling an additional disk module into the disk array remove the disk filler module from the slot c Write the slot number the disk is being installed into on the label affixed to the disk module for example A1 d Install the disk module gently b
297. ess to these pages allows the user for example to determine the configuration change the configuration and view the error log on a disk array Detailed Functional Specifications NIKEARRY gives the user low level control of the disk array Among the supported functions are BIND UNBIND and access to several informational mode pages Access to these pages allows the user to do such things as reset a disk array to it s original factory configuration RAID5 A list of all the available Logical Units LUNs are displayed when the utility is first run and the same list can be displayed again using the DISPMAP command All the operational commands or low level commands use a LUN index displayed in the DISPMAP list The disk array can be referenced either by specifying a LUN index with each command or by setting the TESTUNIT environment variable For example set the TESTUNIT environment variable to LUN index 5 and run several commands on that array or type each command with the LUN index embedded in the parameter list The environment variable is for convenience since any information that is needed by a command is prompted for if it is not already set 7 46 NIKEARRY NIKEARRY Operational Commands NIKEARRY commands are listed in Table 7 3 Examples are shown on the following pages Table 7 3 NIKEAR RY Commands Command NIKEARRAY oper Description ational commands BIND Performs a SCSI MODE SELECT call to bind a LUN
298. etry Count 0 Start Unit Retry Count 0 Figure 8 18 First Report of the Display Disk Failure Information Use the following single key commands to navigate the reports Key Action gt Moves to the next page of information about this failure lt Moves to the previous page of information about this failure Norn Moves to the next disk failure Porp Moves to the previous disk failure Rorr Refreshes the screen contents to reflect the current state of the disk failure information Viewing Maintenance Information 8 51 Bunoouse qnoJj Non Vol RAM Error Log To view the Non Vol RAM error log enter FE mode Then from the Grid Manager Main menu follow the menu path 4 View Maintenance Information 6 Display Non Vol RAM Unsolicited Error Log This screen displays the last eight errors seen by the SP Unlike the other Unsolicited Event Log entries this log is stored on the SP in its Non Vol RAM If this log contains eight entries and another event occurs the oldest entry will be lost to make room for the new entry amp xterm Mei X SP A Display Non Yol RAH Error Log FE SP SCSI ID 15 S N 9621954 Total Controller Elapsed Time 199 17 04 13 6 Time CRU Error Hessage 04 23 98 14 22 01 AD 0x9001 Hard SCSI Bus Error 04 23 98 14 22 01 AD 0x9001 Hard SCSI Bus Error 04 23 38 14 22 01 AO OxA007 CRU Powered Down 04 23 98 14 20 17 f 0x9001 Hard SCSI Bus Error 04 23 98 14 20 17 AD Oxf007
299. example Table 8 14 Rebuild Abort Events and Messages CRU Event Message AO 0x606 Unit Shutdown for Trespass BO 0x641 CRU Rebuild Aborted BO 0x606 Unit Shutdown for Trespass CO 0x606 Unit Shutdown for Trespass DO 0x606 Unit Shutdown for Trespass EO 0x606 Unit Shutdown for Trespass AO 0x602 CRU Enabled BO 0x603 CRU Rebuild Started CO 0x602 CRU Enabled DO 0x602 CRU Enabled EO 0x602 CRU Enabled In this sequence of events LUN 0 has lost disk module B0 and a rebuild is in progress An auto trespass occurs on LUN 0 and a 0x641 error occurs When LUN 0 is re enabled a 0x603 CRU Rebuild Started event occurs and the rebuild resumes at the end of the last fully rebuilt 10 Kbyte block The rebuild process reconstructs data in 10 Kbyte blocks A Unsolicited Event Log 8 65 Buljooysajqnoly partially rebuilt block less than 10 Kbytes is lost when a 0x641 error occurs The rebuild normally resumes within 30 seconds following an auto trespass Recovered Error Log Reporting Unit Shutdown Unsolicited Event Log errors can be produced by certain configurations One configuration that was known to produce these errors contained a piggyback SCSI jumper cable between SP A and SP B The configuration consisted of a host with two Fast Wide differential adapters and a 10 slot array with two SPs Initially the host adapters were set at different addresses and the SPs were jumpered together Fi
300. example of using the Online Diagnostic Environment ODE is shown below ISL ODE ODE run mapper I O Configuration Type HW SW Revisions Path Component Name ID Mod Mod Hdwr Firm 8 HP PB Fast Wide SCSI 4H 4H 3BH 0 0 Resetting SCSI Bus 8 1 1 HPC1300WDR5 RAID 5 disk array HP02 8 1 3 HPC1300WDR1 RAID 1 disk array HP02 8 1 4 HPC1300WDR1 RAID 1 disk array HP02 On path 8 1 1 is a RAID 5 group of five physical disk modules but SYSGEN and the rest of MPE X perceive this as one MPE iX Ldev and one SCSI LUN 5 2 MPEAX Considerations Disk Array Configurations Four hardware configurations are available Table 5 1 lists the hardware components making up each configuration and summarizes the features of each A section for each configuration explains the error recovery features including examples of disk configurations for a model 10 and a model 20 array The section for each configuration is modular and self contained To provide this modularity some information is repeated Table 5 1 Array Configurations Configuration type Basic One host One SCSI 2 adapter One SCSI 2 bus One SP Features With RAID of any supported level applications can continue to run after the failure of a single physical disk module within a logical disk unit LUN Applications cannot continue to run if a SP or SCSI 2 adapter fails Two SP One host One SCSI 2 adapter One SCSI 2 bus Two SPs Provides the same featur
301. f disk arrays can be installed in the rack if a second PDU is installed Install a Rackmount Disk Array in a Cabinet 2 11 uonejelsu A rackmount disk array can be mounted in a Hewlett Packard cabinet using the hardware mounting kit A3231 60030 which consists of Arails 2 keeper tabs 14 Torx 25 screws 12 clip nuts 1 To reduce the weight of the disk array remove the disk modules See Model 10 20 Disk Module Removal Replacement in chapter 9 for information on removing disk modules Make sure each disk module s label which contains the ID number of the slot the disk module is installed into AO for example is present The disk modules must be replaced in their original slots For the slot ID numbers refer to the disk array chassis illustrations in chapter 1 CAUTION Handle disks carefully as they can be damaged easily 2 Remove the disk array s front panel as shown in Figure 2 2 The procedure is the same for both the model 10 and the model 20 2 12 Install a Rackmount Disk Array in a Cabinet To remove front panel pull panel to one side then the other To install front panel press panel onto ballstuds Ballstud Figure 2 2 Removing or Installing the Front Panel 3 Power down the cabinet into which the disk array is to be installed Refer to the cabinet documentation for information 4 Remove any cabinet trim and open the cabinet doors Refer to the
302. for cache SIMMs part numbers A3234A or A3235A must be made according to Table 6 7 Non cache SPs have two 1 Mbyte SIMMs installed which must be removed and replaced with the SIMM combinations from the table in order to use caching Table 6 7 SIMM Sizes for Model 10 and Model 20 Model 10 Model 20 8 Mbyte 4 x 4 Mbyte 8 Mbyte 4 x 4 Mbyte 4 A3235A SIMMs 16 Mbyte 8 x 4 Mbyte 16 Mbyte 8 x 4 Mbyte 8 A3235A SIMMs Setting up Disk Array Caching for Microcode Revision 8 xx 9 xx 6 59 Jobeue N pus Table 6 7 SIMM Sizes for Model 10 and Model 20 cont d 4 or 8 A3235A SIMMs 32 Mbyte 4 x 16 Mbyte 4 A3234A SIMMs 64 Mbyte 8 x 16 Mbyte 8 A3234A SIMMs To set up disk array caching use Grid Manager to specify the cache page size and cache size then change the cache state from disable to enable Specify the cache page size and cache size only if disk array caching is disabled Before specifying these parameters for the first time they have the following default values Cache Page Size 2 Kbytes Cache Size 0 Specifying Memory Configuration for Microcode Revision 8 xx 9 xx The values that can be specified for the storage system memory configuration are Table 6 8 Storage System Memory Configuration Values SP SP620 SCSI SP SPA SP B SPA SPB Total Physical Memory 64 Mbyte 64 Mbyte 64 Mbyte 64 Mbyte System Buffer Size 2 Mbyte 2 Mbyte 4 Mbyte 4 Mbyte Write Cache Size 41 Mbyte 41 Mbyte 40 Mbyt
303. for first host s SCSI 2 adapter Cabling a multi initiator configuration only one array Connect to connector for second host s SCSI 2 adapter Cable not needed for HP UX 10 x Connect to connector for first host s SCSI 2 adapter ANS o A Connect to connector 8 E Connect to connector for first host s second A 5 h for second host s second SCSI 2 adapter SCSI 2 adapter wy CA Connect to connector 7 Connect to connector for first host s first uz for second host s first SCSI 2 adapter SCSI 2 adapter 2 V Connector IN SCSI Bus Cable SCSI Terminator Plug ds Connector OUT Figure 2 9 Two Host Configurations 2 28 Connect the Disk Array s to the Host Computer Set SCSI IDs for SPs in the Disk Array s Each SP connected to the same SCSI 2 bus requires a unique SCSI ID number The back of the model 10 disk array has a 10 switch SCSI ID switch pack Figure 2 10 Eight of the switches are used to set the SCSI ID Four switches set the SCSI ID for SP A and the other four switches set the SCSI ID for SP B Each switch set has four switches IDO through ID3 as shown in Figure 2 10 The uppermost switch labeled Tower is used to set the orientation rackmount or deskside of the Grid Manager Presentation screen The back of the model 20 disk array has two SCSI ID switch packs Figure 2 11 One switch pack sets the SCSI ID for SP A and the other sets
304. g at the SYSDIAG DUI prompt fwscsipb pdev x section 13 where x is the hardware path from the ioscan output Firmware download takes about 40 seconds Check the Software 4 Verify the firmware download At the SYSDIAG DUI prompt type fwscsipb pdev x section 3 where x is the hardware path shown by ioscan The code revision listed in the output must be greater than 0x3621 Contact your WTEC representative to determine the exact adapter firmware revision number HARRAY Diagnostic and HARAYMON Monitor The HARRAY disk array diagnostic and the HARAYMON disk array monitor need to be updated on HP UX 9 x systems See chapter 7 for information about HARRAY and HARAYMON for HP UX 9 x and 10 x versions of the operating system Check Power Requirements See the Reference chapter for AC and DC power requirements and make sure the installation site meets those requirements Check Power Requirements 2 5 uonejjeisuj Unpack the Disk Array Before installing the disk array unpack it from its shipping crate Follow the instructions in the READ ME FIRST card on the shipping carton If the card has been misplaced the instructions are repeated below Unpack a Deskside Disk Array 1 Move the disk array on its shipping pallet to the location where it is to be installed 2 Cut the plastic bands around the shipping carton and pallet Open the shipping carton by cutting the tape over the seams WARNING The shipping carton is held on the
305. g cache synchronization During PROM updates very intermittent Pulling a hot spare during an equalize process Microcode incorrectly pointing to an internal table structure incorrectly caused a panic with the extended status of 0x000000011 Microcode revision 9 x5 fixes problems that resulted in microcode panics 0x00000008 with an extended status of 0x00000003 All known causes of this panic have been fixed If this panic occurs obtain a memory dump and Unsolicited Event Logs in FE mode from both SPs and send them to your HP service representative 0x0000000A Memory exhausted An attempt to allocate a memory resource that is required to always be available failed Rebooting the disk array has been reported to clear the problem Unsolicited Event Log information should be obtained and your HP service representative notified 0x0000000C Stack overflow The kernel detected a stack overflow condition 0x0000000E Memory exhausted Unsolicited Event Log 8 77 punoouse qnouj Table 8 17 Microcode Panic Codes Not a Complete List cont d Value 0x00000010 Description Dual SPs fault simultaneously It was possible for two SPs configured for write cache array to fault at approximately the same time with SIMM parity error panics This would happen if the first SP s backend SCSI processor detected a SIMM parity error reading data from SP memory panic 0x0050300B while it was moving dat
306. g the fan pack It also has an enable disable switch to power it down before inserting or removing it This switch is used to reduce power transients on the power distribution bus that provides power to the SP The SP620 SCSI SP does not have an enable disable switch The SP620 SCSI SP is enabled and disabled using Grid Manager See Enabling Disabling the SP620 SCSI Storage Processor in chapter 6 for information SP microcode is always loaded from disk when the array is booted SP PROM code is updated directly on the SP board An obvious indicator that an SP has failed is the red FAULT LED is on FAULT LED descriptions are in Table 8 2 SP Failures on HP UX 9 0x If a SP fails on HP UX 9 04 access to the LUNs owned by the failed SP is lost The SP must be replaced to regain access to the LUNs Changing Paths and Ownership if an SP Fails HP UX 9 04 Only one path from a host to an SP can be configured at a time In the case of an SP failure the SP can be hot replaced with a new SP and operation can continue without any software modification Following are the steps to accomplish this Troubleshooting the Disk Array 8 27 punoouse qnouj Ifa 1 2 3 4 Ifa gee ON m 8 28 new SP is available Disable the failed SP Remove the failed SP With SP power disabled install the replacement SP Enable the SP power on the replacement SP new SP is not available Stop all I O to the disk array Disable SP power on both SPs
307. gue for unbinding a LUN 4 View Unsolicited Event Log Displays the Unsolicited Event Logs starting at entry number 1 Must use FE mode to see all entries USER mode only shows 0x9xx and OxAxx errors 6 22 Grid Manager Menus FE FE FE FE Allows selection of a filter for either error code or FRU 5 Change Parameters 1 Change Physical Unit Parameters 1 Change Default SP Ownership of Unit Displays dialogue for changing SP ownership 2 Change Unit Rebuild Time Displays dialogue for changing rebuild time 3 Change Unit Auto Assign Displays dialogue for changing auto assign 4 Change Unit Caching Parameters Displays dialogue for changing unit caching 2 Change Storage System Parameters 1 Change Storage System Caching Parameters Change caching parameters N Y Write Cache State Enable Disable Disable Cache Page Size in KB 2 4 8 16 2 Read Cache State Enable Disable Disable Peer Read Cache State Enable Disable Disable BBU Testing Enable Disable Enable BBU Test Day of the Week 1 7 1 BBU Test Hour 0 23 0 BBU Test Minutes 0 59 0 2 Change System Memory Configuration Change Memory Configuration N Y SP A System Buffer Size in MB 2 MB min 2 SP B System Buffer Size in MB 2 MB min 2 Write Cache Size in MB 0 SP A Read Cache Size in MB 0 SP B Read Cache Size in MB 0 SP A Optimized RAID 3 Memory Size in MB 0 SP B Optimized RAID 3 Memory Size in MB 0 Gr
308. gure 8 22 Note that in microcode revision 7 23 Unit Shutdown messages were numbered 0x906 In microcode revisions 8 xx this type of message is a 0x606 Unit Shutdown for auto trespass Neither message is actually an error Adapter A Adare SP A LUN 0 Adapter B SP B Addr 7 LUN Figure 8 22 SCSI Jumper Cable Example The array was on an HP UX 10 0 system and had two LUNs configured Ioscan showed eight I instance numbers for the two LUNs because each adapter sees LUN 0 and LUN 1 through both SP A and SP B and auto trespass generates twice as many instance numbers two instance numbers for each of the two LUNs for each of the two SPs When the piggyback SCSI jumper cable was removed the 0x906 errors did not occur Also when the piggyback cable was removed ioscan showed four I instance numbers which is the correct number when two LUNs are configured in the array on HP UX 10 x When using two host adaptors in a single host be sure to terminate both SP A and SP B Asecond case logging 0x906 errors had a similar configuration but each SP was terminated The problem had to do with how paths were configured in LVM 8 66 Unsolicited Event Log When configuring LVM on HP UX 10 x set up a primary and an alternate path to take advantage of the high availability features of the disk array The 0x906 errors can appear when the primary and alternate paths in LVM are ch
309. gure 9 42 and Figure 9 43 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Remove the outer covers Figure 9 40 3 Remove the following a b Fan pack Figure 9 6 Power supplies Figure 9 11 SPs Figure 9 9 BBU located next to the power supply units Figure 9 13 All disk modules Figure 9 3 Be sure to label each disk module with the compartment number it was removed from 4 Remove the inner metal cover 11 screws Figure 9 41 5 Remove the midplane a 9 72 Remove the two screws that secure the midplane to the chassis one screw at each end of the top of the midplane Disconnect the LED cable Remove the RIP PCA retaining screw Figure 9 44 Pull the midplane out approximately 1 2 inch Tilt the midplane approximately 2 inches and remove the RIP PCA Disconnect all cables from midplane Be careful not to damage the midplane connectors Model 20 Deskside Component Removal Note The presence or absence of a jumper on plug P1 changes the orientation of the Grid Manager Presentation Utility screen If a jumper is on P1 Grid Manager displays a deskside system If there is no jumper Grid Manager displays a rackmount system When replacing the midplane take the jumper off the old midplane and place it in the same orientation on the new midplane When installing a midplane make sure it is seated in the chassis before tighteni
310. h disk array has a matching 25 pin RS 232 port available at a DTC e One RS 232 I O cable 17355D 40234A 13242M or 13242N to connect an RS 232 port on the disk array to its matching DTC RS 232 port To install the RS 232 hardware 1 Install the DTCs and the MDP or DDP RS 232 or both distribution panels in the HP 3000 Series 900 system 2 Connect LAN cables to the DTCs and the LANIC 802 3 interface in the HP 3000 Series 900 system 3 Connect one RS 232 cable between each RS 232 port on the disk array and a corresponding RS 232 port on the DTC Verify Array Connectivity Power on or reboot the HP UX system after all SCSI 2 cards and cables RS 232 serial port or serial MUX cards and cables have been attached and target addresses for the disk array controllers have been set up Connect Asynchronous Console Cables 2 39 uonejjeisuj Note If configuring the disk array via a hard terminal it is also possible to complete any configuration modifications before attaching the disk array to the host computer If planning to use one of the terminal emulation models the disk array will come up in a factory default configuration of RAID 5 After changing any array configuration unbinding or rebinding LUNs it is necessary to reboot the HP UX system to force appropriate driver and special device files to be set up Run the ioscan f n Cdisk command to verify connectivity Examine the outputs of the command The outpu
311. h configuration explains the error recovery features and includes examples of a disk configuration for a model 10 and a model 20 array Table 3 1 Array Configurations Configuration type Basic One host One SCSI 2 adapter One SCSI 2 bus One SP Features With RAID of any level other than 0 or an individual disk applications can continue to run after the failure of a single disk module within a physical disk unit LUN Applications cannot continue to run if an SP SCSI 2 adapter or external SCSI 2 bus fails Two SP One host One SCSI 2 adapter One SCSI 2 bus Two SPs Provides the same features as the basic configuration but with better performance and higher availability With RAID of any level other than 0 or an individual disk applications can continue to run if a single disk module fails With HP UX 10 x if one SP fails ownership of the failed SP s LUNs can automatically transfer to the remaining operational SP If properly configured these operations are transparent to the user Auto trespass is not available on HP UX 9 05 9 07 If the SCSI 2 adapter or external SCSI 2 bus fails no I O operations are possible between the host and the array Multi bus HP UX 10 x and later Two hosts Two SCSI 2 adapters one per host Two SCSI 2 buses one per host Two SPs Resembles two basic configurations side by side and provides modest high availability with the external SCSI 2 buses not shared betwee
312. he chassis six on one side seven on the other b Disconnect the LED cable from the midplane c Lift the inner chassis out of the disk array 8 Remove the midplane by removing 13 screws then remove the midplane PCA from the inner chassis Note The midplane is part of the inner chassis To install the inner chassis in the system lift it over the lip at the rear of the system Top panel Figure 9 30 Removing the Top Panel Model 10 Rackmount Component Removal 9 53 jueuie2e deH pue jeAoway Midplane ue Inner chassis Six retaining screws Bottom Seven retaining screws Figure 9 31 Removing the Inner Chassis 9 54 Model 10 Rackmount Component Removal Interconnect Printed Circuit Board IPCB Figure 9 32 1 SU NS Power off the disk array and disconnect all cables The fans continue to run for a few seconds Access the rear of the disk array Remove the two screws that secure the IPCB to the chassis Figure 9 32 Pull both ejectors until they unlock and the IPCB starts out of the chassis Pull the IPCB out of the chassis making sure not to touch any electronic components or circuits CAUTION Follow ESD procedures when removing the IPCB 6 For each SCSI 2 cable using a small screwdriver a Remove two screws and lift the SCSI connector and cable away from the IPCB bulkhead b Unplug the SCSI connector attaching the cable to the IPCB Model 10 Rackmount Componen
313. he disk array was powered up The number has the format nK where n is a numeric value and K means 1 000 The number of I O requests processed by an individual disk module can be displayed by highlighting the slot for the module on the grid and pressing RETURN Presentation Utility Screen 6 35 Jobeue N pus Cache State Indicates the state of disk array caching DISABLED or ENABLED Write cache cannot be enabled unless the following hardware is present e Afully charged BBU e Two SPs each with at least 8 Mbytes of memory e Disk modules in slots A0 A1 A2 A3 and A4 on a model 10 e Disk modules in slots AO BO CO DO and EO on a model 20 Disk Module Status In the grid in the middle of the screen the blocks of boxes two for the model 10 four for the model 20 identify disk compartments in the chassis Associated with each compartment is an ID that consists of a letter indicating the internal SCSI 2 bus for the compartment and a number indicating the compartment s position on the bus A disk module in the chassis is identified by its compartment s ID A model 10 disk array has two buses A and B each with positions 0 1 2 3 and 4 In a full model 10 chassis the module IDs are AO Al A2 A3 A4 BO B1 B2 B3 and B4 A model 20 disk array has five buses A B C D and E each with positions 0 1 2 and 3 In a full model 20 chassis the module IDs are AO Al A2 A3 BO B1 B2 B3 CO C1 C2 C3 DO D1 D2 D3 E0
314. he disk modules for the primary p image on one internal SCSI 2 bus and the disk modules for the secondary s image on the other internal SCSI 2 bus For example the RAID 1 0 group shown in Figure 1 15 has six modules that make up the primary and secondary image pairs of the group In this example the group can continue after failure of AO or BO A1 or B1 and A2 or B2 because p1 and s1 are in slots AO and BO p2 and s2 are in slots Al and B1 and p3 and s3 are in slots A2 and B2 When binding modules into a RAID 1 0 group it is necessary to select them in the following order pl s1 p2 s2 p3 s3 and so on AO pi A1 p2 A2 p3 Rackmount model BO s1 f Deskside model B1 s2 B2 s3 Primary mirror image p1 p2 p3 Secondary mirror image s1 s2 s3 Mirrored pairs p1 and s1 p2 and s2 and p3 and s3 Figure 1 15 Mirrored Modules in a RAID 1 0 Group in a Model 10 Disk Array Planning Disk Unit Configurations 1 29 uonduioseg jonpold RAID 1 0 for a Model 20 Array When planning a RAID 1 0 group for highest availability and best performance it is necessary to configure each disk module in the group on a different internal SCSI 2 bus For example the RAID 1 0 group shown in Figure 1 16 has six modules that make up the primary p and secondary s image pairs of the group In this example the group can continue after failure of AO or DO BO or EO and CO
315. he following commands are available CAPACITY CLEAR LOGS CONT REP DESCRIBE DIAG DOWNLOAD EXIT HELP CMD ALL INQUIRY LUN REPORT MECH READ BLOCK REQUEST SENSE REVISION RESET RO MT RUL SHOW MECHS SUSPEND AUTO TRESPASS WIR MT Bracketed parameters are optional The HELP command with the ALL option HARRAY gt help all COMMAND DESCRIPTION The following commands are available CAPACITY Displays drive capacity and parameters CLEAR LOGS Allows the user to clear ALL unsolicited sense logs CONT REP Displays configuration information about the controller DESCRIBE Display miscellaneous configuration info about disk DIAG Allows the user to restart controller firmware DOWNLOAD Allows the user to download controller firmware EXIT Exits HARRAY and returns to the DUI prompt HELP CMD ALL INQUIRY LUN REP MECH READ BLOCK REQUEST SENSE Provides this list of commands as well as more detailed descriptions and syntax of each command Displays drive type firmware revision and capacity Displays statistics and configuration information about each FRU configured in the LUN under test Allows user to enter issue non destructive commands at the FRU mech level Reads and displays a block of data from the disk Issue a SCSI Request Sense to the device HARRAY Section 17 External Exerciser RESET Issue a Bus Device Reset REVISION Display firmware revisions RO MT Invoke Mechanism internal READ ONLY Media Test
316. he new options package can be recognized 7 Use IOMAP or DISKINFO to verify that the disk array identifies correctly Checking Auto Trespass and Option Package There is no place to look in Grid Manager to determine which option package has been set on the array for microcode revision 7 23 Grid Manager includes a field that displays the system type for microcode revision 8 20 For microcode revision 7 23 auto trespass one of the parameters in the Storage Systems Option Package is on for option package 2 and off for option package A If the option package is set correctly the following example output should appear after running the ioscan command OPTION PACKAGE A FOR HP UX 9 04 Two SPs in one array with jumper cable installed Class LU H W Path Driver H W Status S W Status disk 0 8 5 0 scsi3 target disc3 ok 0x20650202 ok disk 1 8 6 0 scsi3 target disc3 ok 0x650202 ok disk 2 8 6 1 scsi3 target disc3 ok 0x650202 ok OPTION PACKAGE 2 FOR HP UX 10 0 Two SPs in one array with jumper cable installed Class I H W Path Driver S W State H W Type Description disk 0 16 5 0 disc3 CLAIMED DEVICE DGC C2300WDR5 disk 1 16 5 1 disc3 CLAIMED DEVICE DGC C2300WDR5 disk 2 16 6 0 disc3 CLAIMED DEVICE DGC C2300WDR5 disk 3 16 6 1 disc3 CLAIMED DEVICE DGC C2300WDR5 Grid Manager includes a field that displays the auto trespass setting for microcode revision 8 20 HARRAY can also be used to determine if auto trespass is on or off Iosc
317. he same as the PDC s and MAPPER s Path Product ID Product Identification This is what MAPPER displays as the product string Bus Bus is the type of bus the device is attached to FWSCSI Fast Wide SCSI 2 FLEX Fiber Link and so forth Size A rough calculation of the size of the LUN MB Megabytes Mbytes GB Gigabytes Gbytes Rev Firmware revision of the disk drive Example NIKEARRY gt dispmap Indx Path Product ID Bus Size Rev 0 10 16 12 0 0 HPC2300WD RAID 0 disk array SCSI 1 8 GB HPO1 1 10 16 12 0 1 HPC2300WD RAID 1 disk array SCSI 1 9 GB HPO1 Note 7 52 Due to different calculation methods used the device size shown is only an approximation NIKEARRY DOWNLOAD DOWNLOAD provides the ability to download controller microcode or disk firmware to a specific device The device must be specified by the index number from the NIKEARRY scan and not the hardware path Example NIKEARRY gt download KKKKKKKKKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KK KKK WARNING NIKEARRY must have exclusive access to all NIKEs you wish to update K If you are in a multihost environment such as Switchover amp ServiceGuard make sure all other hosts are powered down before continuing ae KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KK KEK KKK KKK KKK KKK KKK KKK KKK ko ko ko KK Press lt Return gt to continue Please wait while I cre
318. hen updated to revision 8 56 The microcode must be updated to revision 8 56 before the old SP s are removed and new SP620 SCSI SP s are installed Failure to do so will prevent the SP620 SCSI SPs from booting To upgrade the SP microcode from revision 8 20 to 8 56 1 Disable caching and set the cache size to zero From the Grid Manager Main menu follow this menu path 5 Change Parameters gt 2 Change Storage System Parameters 1 Change Storage System Cache Parameters a Type Y to the Change Storage System Cache Parameters prompt b Disable caching by pressing RETURN in response to the prompt Cache State Enable Disable Disable If the default appears as Enable type D to disable the cache c The cache page size prompt appears next as Upgrading from an SP to a SP620 SCSI SP 8 35 punoouse qnouj Cache page size in KB 2 4 8 16 2 Press RETURN to accept the default page size d Set the cache size to zero by typing 0 zero in response to the following prompt Cache size in Mbyte 8 CAUTION Failure to disable caching and set the cache size to zero could cause data loss Verify that the revision of microcode running on the disk array is 8 20 If microcode revision 7 23 is running use the HARRAY DOWNLOAD procedure to upgrade to microcode revision 8 20 before continuing with these procedures Contact your WTEC representative for information about where to obtain the microcode See the H
319. i DL EU LV C E RA EAR ES 6 34 Microcode Revision 0 cece eee cent e eee e ne nens 6 34 Statistics Logging ssas 1 cee teen Ih 6 35 Peer SPs nma eue aea lade eee ete ieee eee dee SA 6 35 Requests Completed isa resero nni EE eee eee EE i 6 35 Cache State ie qui eme R ee e ete ea ee ex eee Sie eae 6 36 Disk ModuleStatus lesse hh 6 36 PROM Cache and Other Components Status 0 000 cece eee 6 37 PROM Revision 25 eI era EA per ERR ee ee 6 37 FAN X utis pIPUSH preteen filled teed Sis 6 39 Power Supply Unit VSC XX 0 eee eee eee eens 6 39 BBU 4e dtd eb puit E eeaeet espe ea 6 39 ADEPEm BDPn UDP1T uw Gah ce eee Ad a eee ae 6 40 Moving the Grid Cursor 0 0 cece eee eee een eee nes 6 41 Grid Manager Procedures 2 cece ee eee eee eee eee 6 41 Enabling Disabling the SP620 SCSI Storage Processor 000 005 6 42 Creating Logical Disk Units LUNs 00 0 cece eee eee 6 43 Preparing to Create Logical Disk Units 0 cc eee ee eee 6 43 Binding Disk Modules as Logical Disk Units 0 04 6 43 Logical Unit Number LUN 0 cece eee eee eee eee nee 6 47 Maximum Rebuild Time 2 0 2 0 cece eee ee eee enn 6 48 Sectors per Stripe Element 0 0c eee cece eee eese 6 48 Enable Caches LUN Cache 0 cece eee eee ee ene 6 48 Unbinding a Logical Disk Unit LUN eese 6 51 Setting up Di
320. i Di IBM DGHS18 CLRR18 17 89b E1 E1 IBM DGHS18 CLRR18 17 89b Weoovovuvuvu v eee ey a Ue a Ul a i a i a Figure 7 29 Information Tool Log Screen Configuration Section 7 104 Support Tools Manager STM The information displayed includes the following Table 7 17 Information Tool Log Screen Headings Column heading Description If information is not obtainable for example LUN not supported or disk faulted the Lun LUN to which current disks are bound displays if no LUN C D Current Default ownership of the LUN displays if none Fru BS Disk FRU FRU bind slot if they exist Cch LUN read R on off write W on off cache ATr LUN auto trespass state on off hot spare Type Vendor LUN type or disk vendor as appropriate Product ID LUN or disk product identification as appropriate AprxCap LUN or disk approximate user capacity as appropriate where AprxCap GB user bytes 1 000 000 000 BSz LUN block size B R Build Rebuild percentage unobtainable information is reported as a The disk section presents detailed information about the disk and vendor Support Tools Manager STM 7 105 sonsoubeig pue seni ISOH Information Tool Log for C2300 Array on path 874 9 BAA BAA BAA RAR RAR AAA BAA Ps Ps PS PS Ps PS PS Ps PS Ps Ps PS Ps Ps PS PS Ps PS Ps Ps PS Ps Ps PS PS Ps PS Ps Ps PPS
321. iagnostic Trouble Tree HARRAY Section 17 External Exerciser NIKEARRY Support Tools Manager STM SAM System Administration Management Tool LVM Logical Volume Manager EMS Hardware Event Monitoring 7 1 sonsouDeig pue seni ISOH Overview The diagnostics utilities and monitors available for use with the array are e JOMAP offline e SYSMAP online e LOGTOOL e HARAYMON e HARRAY online e NIKEARRY offline e STM online e SAM e LVM e EMS 7 2 Overview IOMAP The offline IOMAP utility identifies any connected peripheral for HP UX 9 0x For HP UX 9 04 at the ISL prompt type IOMAP and press RETURN For HP UX 10 x at the ISL prompt type ODE and then MAPPER Search the display for the following ID strings Table 7 1 ID Strings for Model 10 and Model 20 ID string for a model 10 ID string for a model 20 Description DGCC1300WD DGCC2300WD Individual disk DGCC1300WDR5 DGCC2300WDR5 RAID 5 group DGCC1300WDR3 DGCC2300WDR3 RAID 3 Not Supported DGCC1300WDR1 DGCC2300WDR1 RAID 1 mirrored pair DGCC1300WDRO DGCC2300WDRO RAID 0 group DGCC1300WDR10 DGCC2300WDR10 RAID 1 0 group DGCC1300WDUNB DGCC2300WDUNB Unbound On HP UX 10 x and MPE iX operating systems run ODE Offline Diagnostic Environment and type MAPPER for peripheral identification The following sample output appears from the GENERAL IOMAP under SYSDIAG 8x7 CPU example This is an exa
322. iated negotiation DISABLED ENABLED ENABLED Substitute busy for qfull DISABLED ENABLED DISABLED Mode page 8 DISABLED ENABLED DISABLED Recovered error log reporting DISABLED ENABLED ENABLED Auto Trespass DISABLED ENABLED ENABLED 6 View Cache Statistics Displays dialogue for displaying cache statistics FE 7 Back End Exerciser Displays dialogue and instructions for running the back end exerciser Grid Manager Menus 6 21 Jobeue N pus FE Must be in FE mode to access these screens or functionality To enter FE mode type CONnTROL P followed by in upper case FE To return to user mode type CONTROL P followed by in upper case USER Grid Manager Menus for Microcode Revision 8 20 Menu number from the Grid Manager Main Menu 1 Presentation Utility Displays the Presentation screen 2 Bind Physical Units 1 RAID 5 Group Individual Access Array Displays dialogue for making a RAID 5 LUN 2 RAID 3 Group Parallel Access Array Displays dialogue for making a RAID 3 LUN 3 RAID 1 Mirrored Pair Displays dialogue for making a RAID 1 LUN 4 RAID 0 Group Non redundant Individual Access Array Displays dialogue for making a RAID 0 LUN 5 RAID 1 0 Group Mirrored RAID 0 Group Displays dialogue for making a RAID 1 0 LUN 6 Individual Disk Unit Displays dialogue for making an independent mode disk 7 Hot Spare Displays dialogue for making a hot spare 3 Unbind Physical Units Displays dialo
323. ice UC C PS RS XJ X ES A X2 AD A ike ike ike ike ike ike ike ike ike ike ike ike K K K K K Index number used for referencing the device Firmware Revision of the device Bound HP02 Bound HP02 Bound HP02 Bound HP02 Bound HP02 Bound HP02 Bound HP02 Bound HP02 Bound HP02 Bound HP02 Bound HP02 Bound HP02 Rev is the firmware revision of the device storage firmware revision Note Due to different calculation methods used the size of the device shown is only an approximation NIKEARRY 7 49 sonsouDeig pue seni ISOH BIND The BIND command configures disk FRUs into usable Logical Units LUNs FRUs selected to be bound cannot be part of an existing LUN This command prompts the user for all the necessary information Defaults should be used in all cases except LUNs on binding Note When a range is requested use the following definition of range to enter the response Range lt number gt lt number gt lt number gt lt number gt For example to enter four numbers as a range type 0 1 2 3 or 0 3 or 0 2 3 Example NIKEARRY gt gt bind Enter test array cr 0 3 0 Enter Bind Type 1 Disk Array Individual Disk Disk Mirror Striped Disk Array No Parity eR Hot Spare Raid 1 0 Enter selection 1 1 7 q 1 Enter Rebuild Time in hours 4 0 255 q Enter LUN number 0 0 29 q Enter Auto Assign Perm
324. icrocode or firmware file to be downloaded to the disk array must be available on the host system The procedure to update the microcode or firmware with the Firmware Update Tool starts with the following two steps 1 Select the microcode or firmware file to be used for the update with the Select File option on the File menu 2 Identify the type of file selected either SP microcode or disk module firmware with the Update menu SP Microcode Updates If the SCSI disk array uses two SPs both should be enabled before firmware is updated so that they are updated simultaneously 7 110 Support Tools Manager STM The download update PROM and restart controllers property from the Controller Firmware option on the Update menu is recommended for general use The other options provide a greater degree of control needed only in special circumstances Disk Module Firmware Updates The peer SP must be disabled when downloading disk module firmware in a dual SP SCSI disk array configuration If a peer SP is enabled the Firmware Update Tool requests that it be disabled and the process does not continue until the request is satisfied Grid Manager must be used to disable an SP STM does not support this functionality CAUTION Never reenable a peer SP while a disk module firmware update operation is in progress Doing so can interrupt the operation causing permanent disk module and or data loss Exclusive access is required to download d
325. id Manager Menus 6 23 Jobeue N pus 3 Change Main Menu Password Displays dialogue for changing password 4 Change System Date and Time Displays dialogue for changing date and time FE 5 Select Storage System Options Package Displays dialogue for changing system type FE 6 Change Storage System Fast RAID 3 Parameters Displays dialogue for changing FAST RAID 3 cache vault area 3 Change SP Parameters 1 Enable Disable Statistics Logging Displays dialogue for controlling statistics logging 2 Enable Disable Periodic Error Reporting Displays dialogue for controlling periodic error reporting 3 Change Console Parameters Displays several modifiable console parameters default is 9600 baud 8 bit no parity 4 Update SP PROMs and Reboot SP Displays dialogue for updating SP PROMs 5 Change Host Interface Options Addressing model LUN TARGET LUN Target initiated negotiation DISABLED ENABLED ENABLED Substitute busy for qfull DISABLED ENABLED DISABLED Mode page 8 DISABLED ENABLED DISABLED Recovered error log reporting DISABLED ENABLED ENABLED Allow non Mirrored Cache DISABLED ENABLED DISABLED Auto Trespass DISABLED ENABLED ENABLED 6 View Cache Statistics Displays dialogue for displaying cache statistics FE 7 Back End Exerciser Displays dialogue and instructions for running the back end exerciser 6 24 Grid Manager Menus FE Must be in FE mode to access these screens or functionalit
326. ide s the multiplexed SCSI 2 bus traffic between two internal synchronous single ended SE SCSI 2 buses identified by the letters A and B The SP s in a model 20 disk array divide s the multiplexed SCSI 2 bus traffic between five internal synchronous single ended SCSI 2 buses identified by the letters A B C D and E Each internal synchronous SE SCSI 2 bus supports multiple disk modules SCSI 2 adapter controller The name for the printed circuit board within the computer chassis that allows the host to access the external Fast Wide differential SCSI 2 bus SCSI ID A number that identifies a device on an external SCSI 2 bus An SP s SCSI ID is selected using dip switches located on the back of the disk array The default SCSI ID for SP A is 0 the default SCSI ID for SP B is 1 secondary disk array path or secondary See primary disk array path SP See storage control processor SP SP memory modules Memory modules SIMMs that provide the local storage cache for an SP An SP must have at least 8 Mbytes of memory in the form of two 4 Mbyte SIMMs to support the storage system cache Glossary G 5 Aiesso 5 storage control processor SP A printed circuit board with memory modules that control the disk modules in the storage system chassis The SP runs Grid Manager which is used to bind and unbind disk units set up disk array caching observe array status and view the SP event log The SP in a model 10 disk arr
327. if disk modules are moved to different slots or different arrays so that they are out of order after being BOUND If this is done in most cases the data on the disk modules will be destroyed The typical failure occurs as follows 1 The disk modules are bound into LUNs either at or after initial install 2 Disk modules are moved from the one array to another 3 The disk modules are moved usually with power off but sometimes with power on 4 The disk modules that were in a NON HA mode not RAID 1 RAID 1 0 or RAID 5 now show up as a blank with the LUN number showing in the box 5 Later another disk module is added to the array and the disk module seems to pick up the old LUN number but cannot be accessed by either SP A check of the Unsolicited Event Logs in FE mode shows Bind Signature Error for the given LUN To clear this up 1 Enter FE mode by pressing CONTROL P followed by SHIFT FE 2 In FE mode follow the menu path from the Grid Manager Main menu 5 Change Parameters gt 2 Change Storage System Parameters 1 Change Storage System Caching Parameters 8 68 Unsolicited Event Log 3 Inresponse to CHANGE CACHE PARAMETERS type YES then select DISABLE and press RETURN repeatedly past the bottom of the screen BBU TEST portion 4 Wait one or two minutes for the CACHE to complete the Dump Disable process 5 Select Change Storage System Caching Parameters again 6 In response to Change Cache Pa
328. ify cmd i B5535 None Lasr bitk to seerify misst he Pest block oo verify Figure 7 23 Tests Verify LUNs Select Blocks Screen Support Tools Manager STM 7 95 sonsoubeiq pue seni ISOH Utilities gt System Configuration System Cache ITE Pim ES Barer iy Ear shoe caras sl lin Head Cache Eraba Ce Di salad ova Fuera Alc aai ume bran T8 Pridni i Dassin Wie Caches Enabiud Se Di ead ied Sots L ngrai so Enabhiid d Hes iow Watermark value QU 105 High trem art value Uc 11 v Cacka Pape Bice 2K JK Be ME EJ Figure 7 24 Utilities System Configuration System Cache Screen 7 96 Support Tools Manager STM Utilities System Configuration System Memory Becht y Thee pinsi eee ceni allon values SP A Tirai Piegsiial Meme DHA nes H SP B Tow Physical Bbsenary Riis B d SP A Syrien Better alee Mbyte 4 d Xp HO Eyewear Euer sire Mipit d aP A Dpiiemizamd FUSE 3 rear ee y sce EAE e I x AP B prbrdonad FID X miary size Ayres fi Wiha Cache eue Missa I SP A lead Cache Slane EEvoas 1 SP B Reed Cecha Stew MBs 1 Pes SF aures hulter doe and 5E MADI mamay acne may any Ba glared i sil ayas caching 9x digabiad Changing vestem bufer sioe pr FU rra sie will cower the EF being aririciwssur in resiart arbo iino ie slo nire oly bie alle EE ite bacs ks disahiiendi SP A acid E Pede Cacus abe m
329. igh Availability Features 2 0 0 0 eese 1 18 RAID Level 5 1 or 1 0 Data Redundancy 0 cece eee ee eee 1 18 Disk Array Caching 00 cece eee Ih 1 19 Configurations with Two SPs or Two Hosts or Both 04 1 20 Auto Trespass i e ERIER ee oe Leet ae sai I Ed 1 22 Planning Disk Unit Configurations 0 0 0 ee eee eee ee eee 1 23 RAID Concepts ee iat eed eh ed teers ee tle ele 1 23 Disk Siping os 4 0 24e he cate See CE HERS Pee REDE Ee ye EUR 1 24 Mirroring ss de ed are eng ele eE ln abn vet il eS ERE 1 24 RAID 0 Group Non Redundant Array 00 cece eee ee eee 1 25 RAID 1 Mirrored Pair 12 2 0 a cece cee ee eee eens 1 26 RAID 1 0 Group Mirrored RAID 0 Group 2 2 000 1 27 RAID 5 Group Individual Access Array 0 000 e ee ee eee 1 30 Individual Disk Unit 2 0 cee cee ee ee eens 1 32 HOt Spares reee veh oe US y dikes MES he mS ee ee Lee dk 1 32 Data Availability and Disk Space Usage 00 0202 eee eee 1 35 Guidelines for Logical Disk Unit Configurations 1 38 RAID 5 Group Individual Access Array 0 202s eee eee eee 1 39 RAID 1 Mirrored Pair coL Ghee ee a ess 1 39 RAID 1 0 Group Mirrored RAID 0 0 0 cee eee eee eee 1 39 RAID 0 Group Non Redundant Individual Access Array 1 40 Individual Disk Unit eseseeeeeeeee eA 1 40 HotSpare 2 214 a S
330. ight disk support for the SP only 2 Gbyte disk support for the SP only for the SP only 4 Gbyte LP low profile and 9 Gbyte HH disk support for the SP only 4 Gbyte HH disk support for the SP620 SCSI SP only for the SP620 SCSI SP only 4 Gbyte LP and 9 Gbyte HH disk support for the SP620 SCSI SP only 4 Gbyte LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP only 4 Gbyte LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP620 SCSI SP only The following table shows the minimum supported microcode and PROM revisions required for each disk module supported on the High Availability SCSI disk array Table 6 2 Minimum Microcode and PROM Revisions for Supported Disk Modules Model 10 Model 20 iE appz Disk Disk module Capacity Microcode PROM Microcode PROM Microcode PROM revision revision revision revision revision _ revision A3240A 2 GB 7 16 7 97 7 16 7 97 8 51 1 49 A3396A 4 2 GB 7 16 7 97 7 16 7 97 8 51 1 49 A3541A 8 8 GB 8 27 8 02 8 27 8 02 8 58 1 57 A3666A 4 2 GB 8 27 8 02 8 27 8 02 8 58 1 57 A5132A 8 8 GB Not supported 9 25 00 8 02 9 55 01 1 73 6 38 Presentation Utility Screen Table 6 2 Minimum Microcode and PROM Revisions for Supported Disk Modules cont d Model 20 with SP620 SP Model 10 Model 20 Disk Disk module Capacity Microcode PROM Microcode PROM Microcode PROM revision revision revision revision revision revision
331. iled description of LVM see the following manuals e How HP UX Works Concepts for the System Administrator e System Administrator Tasks e Solving HP UX Problems The following are basic instructions to assist in configuring the disk array with LVM Creating Physical Volumes Execute the pvcreate command for each physical volume on the system For example a model 10 array is configured as follows LUN 0 RAID 1 2 disks device file c5t1d0 volume group id vg01 size 1988 Mbytes LUN 3 RAID 5 5 disks device file c5t1d8 volume group id vg02 size 7952 Mbytes LUN 5 RAID 5 3 disks device file c5t1d5 volume group id vg03 size 3976 Mbytes Note Device files vary in form from HP UX 9 04 to HP UX 10 x Auto trespass must be set at installation when manually creating physical volumes and defining the primary and alternate data paths to them SAM does this automatically In the following example the physical volume is equal to the LUN therefore pvcreate is run once for each LUN Run the ioscan f n command if necessary to determine the raw files c5d0S2 is an example of an HP UX 9 04 device file c5t1d0 is an example of an HP UX 10 x device file LVM Logical Volume Manager 7 129 sonsoubeig pue seni ISOH For example HP UX 9 04 HP UX 10 x pvcreate f dev rdsk c5d0S2 pvcreate f dev rdsk c5t1d0 pvcreate f dev rdsk c6d0S2 pvcreate f dev rdsk c5t1d3 pvcreate f dev rdsk c7d0S2 pvcreate f
332. ing a dual adapter dual SP configuration i Oo 8 E 2 9 Connect to host s second SCSI 2 adapter v EF AN cS e x A E A 3 Connect to host s first SCSI 2 adapter lel V Connector IN SCSI Bus Cable SCSI Terminator Plug A Connector OUT Figure 2 8 Single Host Configurations 2 26 Connect the Disk Array s to the Host Computer Connect the SCSI 2 Cables in Two Host Configurations Each two host configuration requires two SPs If an SP needs to be installed refer to Model 10 20 SP Removal Replacement in chapter 9 for instructions Identify which of the following configurations is to be set up e Multi bus e Multi initiator not supported on MPE iX e Multi bus multi initiator not supported on MPE X For detailed information on these configurations see chapters 3 4 and 5 Note The total length of the SCSI 2 bus must not exceed 25 m 81 ft including all external and internal SCSI 2 bus cables for the host s and the disk array s Securely connect the SCSI 2 cable s and terminator s to the disk array as shown in Figure 2 9 Connections should be finger tight Connect the Disk Array s to the Host Computer 2 27 uonejjeisuj SCSI Bus Cable Terminator Plug Thumb Thumb Screws Screws C SCSHB Connect to connector for second host s SCSI 2 adapter SCSA N SCSHA 3 Connect to connector
333. ingle phase power factor corrected 100 120 127 220 240 VAC power systems Range 90 132 and 180 264 VAC 47 to 63 Hz Reference R 9 o2ue49JoH Current Model draw VA Ww BTU hr K cal hr Model 10 4A 393 391 1 333 336 Model 20 8A 693 694 2 367 597 Marked electrical rating under maximum load Power Consumption Apparent power True power DC Power Specifications Fan pack SP Disk module Environmental Specifications 473 VA maximum for model 10 768 VA maximum for model 20 472 W power factor corrected maximum for model 10 766 W power factor corrected maximum for model 20 Operating Environment 24 VDC power from power supplies 24 VDC on model 10 power from power supplies by on board voltage regulator 5 VDC developed 48 VDC on model 20 power from power supplies 5 VDC developed by on board voltage regulator 5 VDC and 12 VDC power from the disk module voltage regulator which receives 24 VDC from the power supplies For continuous trouble free operation the high availability disk array should not be operated at its maximum environmental limits for extended periods of time The recommended operating range provides a less stressful operating environment Operating within the recommended operating range ensures maximum reliability for the disk array R 10 Reference The following specifications were type tested under controlled conditions Hewlett
334. ins the property of HP aa KKKKK KKKKK ERFAN Version A 01 15 KEREK KKKKKKKK KKKKKKKK KKKKK KKK e KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KK KKK KKK KKK KKK KKK KKK kk KKK KK Welcome Today is MON APR 29 1996 at 4 04 PM HIGH AVAILABILITY Disk Array Diagnostic initializing 5 HIGH AVAILABILITY Disk Array Diagnostic verifying path to device HIGH AVAILABILITY Disk Array Diagnostic identifying selected disk Initialization SUCCESSFUL Section 17 INTERACTIVE EXERCISER PDEV 24 14 0 is an C1300WDR5 HARRAY Section 17 External Exerciser 7 13 sonsouDeig pue seni ISOH AUTO TRESPASS HP UX 10 x and later Non exclusive non destructive The AUTO TRESPASS command is used to turn the auto trespass function on and off Examples The first example shows the AUTO TRESPASS command with option 1 selected HARRAY gt AUTO TRESPASS Enable Disable Auto Trespass current state 2 ON 1 Turn Auto Trespass OFF 2 Turn Auto Trespass ON 99 Cancel command Enter Selection 1 2 99 99 gt 1 AUTO TRESPASS COMMAND SUCCESSFUL The second example shows the AUTO TRESPASS command with option 2 selected HARRAY gt AUTO TRESPASS Enable Disable Auto Trespass current state 1 OFF 1 Turn Auto Trespass OFF 2 Turn Auto Trespass ON 99 Cancel command Enter Selection 1 2 99 99 gt 2 AUTO TRESPASS COMMAND SUCCESSFUL The next example shows the AUTO TRESPASS command with option 99 selected HARRAY gt AUTO TRESPASS Enable Disab
335. ion A configuration that has one host with one SCSI 2 adapter for the disk array and a disk array with two SPs host In the context of disk arrays a processor that runs an operating system and uses a disk array for data storage and retrieval hot repair See replace under power hot pluggable image disk image See mirroring individual disk unit A disk module bound as an individual unit independent of any other disk modules in the cabinet An individual unit has no inherent high availability feature but since the G 2 Glossary operating system supports software mirroring it can be made highly available by software mirroring it with another individual unit preferably on a different internal bus within the array logical unit LUN A logical unit is one or more disk modules bound into a single entity accessible by logical unit number LUN Logical unit is a SCSI term This manual generally uses the term physical disk unit which has the same meaning as logical unit memory module See SP memory modules mirroring Maintenance of a second and optionally via software a third copy of a logical volume image that provides continuous access if an image becomes inaccessible The system and user applications continue running on the good image without interruption There are two kinds of mirroring hardware mirroring in which the disk array maintains synchronization of the disk images and software mirroring in which the ope
336. is already configured or the Fast Wide SCSI 2 port of a model 9x9 Core I O card is being used HP PB Fast Wide SCSI 2 adapter example PATH TD PMGR io ap 10 16 4 id hp28696a io lp 10 16 4 10 16 4 HP28696A FWSCSI DAM LMGR LD TY EV PE PMGRPRI MAXIOS DA 6 0 ID number A2372 60003 FWSCSI is used when configuring a the Fast Wide SCSI 2 port on a model 9x9 Core Multi Function I O card Configuring Disk Array SCSI LUNs for MPEAX 5 13 006 Seles 000 dH uoneunbyuo Core I O example io gt lp 10 4 24 PATH 10 4 24 LDEV ID A2372 60003 FWSCSI TYPE DA PMGR FWSCSI DAM PMGRPRI 6 LMGR MAXIOS 0 Configure the path of the SP using the SCSI Target ID of the SP 1 in this example and a SYSGEN ID of PSEUDO io ap 10 16 4 1 id PSEUDO io lp 10 16 4 1 PATH 10 16 4 1 LDEV ID PSEUDO TYPE DA PMGR TRANSPARENT MGR PMGRPRI 6 LMGR MAXIOS 0 Configure an Ldev for each LUN owned by the SP in this example LUNS 0 1 and Z using a SYSGEN ID that corresponds to the disk array model number displayed by ODE Mapper 10 Slot Array HPC1300WD 20 Slot Array HPC2300WD Note ODE Mapper suffixes the model IDs with a two character RAID mode identifier of R1 or R5 Do not include the suffix in the SYSGEN ID as it causes the add device command to fail MPE iX determines the RAID mode of the configured LUN by directly querying the disk array Configure disk array
337. isk array has two SIMM pairs of different sizes then use the crossover address see Table 8 9 to identify which SIMM pair has failed If the address of the parity error is below the crossover address then the larger of the SIMM pairs failed Otherwise the smaller of the SIMM pairs failed Note 3 Ifthe disk array has two SIMM pairs of the same size then use the crossover address to identify which SIMM pair has failed If the parity error is below the crossover address the failure is in the first pair of SIMMs slots XU1 and XU2 If the address of the parity error is above the crossover address the failure is in the second pair of SIMMs slots This is true only if the larger SIMM pair is in the first slots XU1 and XU2 XU3 and XU4 Table 8 9 SIMM Configuration SIMM size MB SIMM pair SIMM pair Total memory Crossover address XU1 XU2 XU3 XUA 4 4 8 4 4 4 4 16 0x00800000 16 16 32 0x02000000 16 16 4 4 40 0x02000000 16 16 16 16 64 0x02000000 Troubleshooting the Disk Array 8 33 punoouse qnouj Example 1 In a 40 Mbyte disk array that has one pair of 4 Mbyte SIMMs and one pair of 16 Mbyte SIMMs the failing address is 0x02345678 This is above the crossover address of 0x02000000 which means the 4 Mbyte SIMM pair has failed Example 2 In a 64 Mbyte disk array that has four 16 Mbyte SIMMs the failing address is 0x01234567 This address is below the crossover address of 0
338. isk firmware All file systems should be deactivated while updating disk firmware Expert Tool Expert Tool provides useful SCSI disk array support capabilities including e Information retrieval e Functionality testing e Failure analysis e Data integrity checking e SCSI disk array management Expert Tool offers much of the functionality available through Grid Manager Access the Expert Tool window by selecting Tools Expert Tool Run This window provides six menu items in addition to Lab Use Only and Help Figure 7 34 The Lab Use Only item is intended for use only by Hewlett Packard factory engineers and is not documented here Support Tools Manager STM 7 111 sonsoubeig pue seni ISOH This window also provides a status of the activities performed on the selected device hpterm Iof x Expert Tool for device type C2300 Array on path 874 9 File Logs Info I 0 Tests Utilities Lab Use Only Help Read Disks command CANCELED per user request Read of DISK AQ beginning blocks 0 thru 125 125 cmd lt max gt Read of DISK AO completed blocks 0 thru 125 0 errors detected Read Disks command SUCCESSFUL 1 disks processed 0 errors detected Read of LUN 0 beginning blocks 0 thru 127 127 cmd max Read of LUN 0 completed blocks 0 thru 127 0 errors detected Read Luns command SUCCESSFUL 1 luns processed 0 errors detected Test Unit Ready command SUCCESSFUL Verify
339. ission Nu FP WD l 0 Auto assignment is disabled 1 Auto assignment is enabled Enter Permission 0 0 1 q Enter Stripe Count 0 0 4294967295 q Enter Element Size 128 0 4294967295 q Enter Default Ownership 0 Peer controller will own this unit 1 This controller will own this unit Enter Selection 1 0 1 q Enter Auto Trespass Enable 0 Auto Trespass Disabled 1 Auto Trespass Enable 7 50 NIKEARRY Enter selection 1 0 1 q Enter Write Cache Enable 0 Do not allow Write Caching 1 Enable Write Caching on this unit Enter selection 0 0 1 q Enter Read Cache Enable 0 Do not allow Read Caching 1 Enable Read Caching on this unit Enter selection 0 0 1 q Enter 1 to 16 FRUs between 0 and 29 to be bound 0 4 0 29 q 0 2 Bind command will create a LUN and may take up to 4 hours to complete Would You Like To Continue y n y Enter MODESENSE 0x2A to verify the results of this command The FRU numbers of the FRUs to be bound need to be specified Refer to the following FRU map for FRU numbers NIKEARRY 7 51 sonsoubeiq pue seninf ISOH DISPMAP The DISPMAP command displays a list of all the LUNs found on the system or if given a lt LUN index gt only displays that LUN The fields displayed are Indx Device Index number Used by other NIKEARRY commands to reference that specific device Path Logical Path to each device The Path constructed is t
340. it trespass The microcode has been coded to eliminate internal resource deadlocks between SNiiFFER and normal I O operations Resource starvation caused by heavy I O operations during rebuilds All known occurrences of this panic have been fixed in microcode revisions 9 xx and later If this panic occurs obtain a memory dump and Unsolicited Event Logs from both SPs and send them to your HP service representative for failure analysis 0x00240010 SCSI protocol problems The microcode panics 0x00A000B0 0x00240010 and 0x00220050 occurred in SCSI environments that have severe SCSI protocol problems due to improperly working initiators or noisy conditions Fixed in microcode revisions 8 29 8 61 and 9 xx 0x0024001A Front end SCSI reset Caused by a frontend SCSI bus reset during a SCSI write buffer mode 1 command Fixed in microcode revision 8 20 A SCSI bus reset on the front end while a Mode Select is in progress might intermittently cause the microcode to panic Fixed in microcode revision 8 20 0x0026000E PROM erase failed An attempt to erase a FlashROM failed 0x0026000F PROM program failed An attempt to program a FlashROM failed 0x00260022 GridMgr failure If a change configuration type request was issued during extremely heavy I O GridMgr failed to read the request from the pipe Fixed in microcode revision 8 20 0x00280002 Bad unit type detected Executing a SCSI Verify or Writ
341. ite cache be enabled LUN cache may show enabled but will not be operational unless storage system cache also is enabled Two SPs SIMMs and a BBU must be present to enable storage system cache Storage system and LUN cache use the SIMMs installed on the SP Enabled LUN caches randomly share the storage system cache pool No specific cache addresses are assigned on a LUN by LUN basis The storage system cache state can be observed on the Grid Manager Presentation screen Note To change the disk s bind options answer N to the Confirm bind options Y N question then reselect bind options To reselect different disks to bind answer Y to the Confirm bind options Y N question Answering Y does not start the binding process 6 48 Creating Logical Disk Units LUNs 10 11 Bind these disk modules Y N N Answering Y starts the binding process for the disks that have been selected Answering N allows reselection of the disk s to be bound Binding takes approximately 45 to 60 minutes for microcode revision 7 23 and 30 to 45 minutes for microcode revision 8 xx A RAID 5 LUN consisting of five 4 2 Gbyte disk modules takes about 70 minutes to bind A RAID 5 LUN of five 8 8 Gbyte disk modules binds in about 140 minutes for microcode revision 9 xx If Y was typed Grid Manager changes the UNB to BIN for a RAID group or individual unit or to BIN H for a hot spare and starts creating the LUN specified As soon as Gri
342. ith better performance and higher availability With RAID of any level other than 0 or an individual disk applications can continue to run after the failure of a single disk module SP No 9 04 Yes 10 x HP UX 9 04 If an SP fails the System Administrator must use Grid Manager to reassign ownership of the failed SP s LUNs to the operational SP Also the path to the operational SP must be used and can be changed by way of LVM Once these changes have been made access to all LUNs on the array is possible HP UX 10 x If one SP fails ownership of the failed SP s LUNs can transfer automatically to the remaining operational SP If a primary and an alternate path have been configured in LVM LVM automatically switches to the alternate path These operations are transparent to the user Fan pack Yes 9 04 Yes 10 x Applications continue running System Administrator or service provider replaces module Power supply 9 04 10 x Yes 9 04 Yes 10 x If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply The System Administrator restarts the operating system if necessary SCSI 2 adapter cable No 9 04 Yes 10 x HP UX 9 04 If the adapter cable fails the System Administrator must access the array through the operational SCSI 2 adapter s path LUN ownership and path changes need to be made u
343. itiators or noisy conditions Fixed in microcode revisions 8 29 8 61 and 9 xx 0x00220054 I O panic Caused by issuing a SCSI Write Verify command several times under certain conditions Fixed in microcode revision 8 20 0x00220055 I O panic Executing a SCSI Verify or Write Same commands to a Fast RAID 3 LUN would result in 0x00280002 or 0x00220055 panics Fixed in microcode revision 9 xx Unsolicited Event Log 8 79 Buljooysajqnoly Table 8 17 Microcode Panic Codes Not a Complete List cont d Value 0x00220056 Description I O panic The microcode would occasionally panic during a rebuild if another disk faulted during a write operation The following was a cause of this panic A large write requests larger than 0x200 sectors is received by the microcode and broken into multiple parts Due to lack of resources a busy system the parts are placed on a waiting queue Some other request completes but does not free the specific resources the waiting requests need The microcode de queues the first event on the waiting queue in an attempt to restart it But since the resources it needs are still not available it cannot be started The microcode then places the event back on the waiting queue Atsome point in the future the microcode attempts to start the next event on the waiting queue which is now the second part of the request This causes the 0x00220056 panic because
344. ittently panic with a 0x00220011 Fixed in microcode revision 8 20 0x00220013 Alternating LUN assignment with cache and auto trespass enabled If the host continuously alternates assignment of a LUN between peer SPs the microcode would panic with 0x00220013 This would only happen if cache and auto trespass were enabled Fixed in microcode revision 8 20 0x00220023 Memory or resource panic Normally occurred during large file I Os Fixed in microcode revision 9 xx 0x00220026 Memory or resource panic Occurred while LUNs were continuously trespassed under a heavy I O load Fixed in microcode revision 8 20 8 78 Unsolicited Event Log Table 8 17 Microcode Panic Codes Not a Complete List cont d Value 0x00220030 Description Unexpected lock The execution of a disk read write operation detected an invalid conflict with another operation 0x00220032 Cache page processing Executing large block reads gt 64Kb could result in this panic when a SCSI front end process attempts to process a cache page Fixed in microcode revision 7 23 0x00220040 Inter process communication panic Occurred when taking an error while trying to respond to a request from a host Fixed in microcode revisions 8 29 8 61 and 9 xx 0x00220050 I O panic The microcode panics 0x00A000BO0 0x00240010 and 0x00220050 occurred in SCSI environments that had severe SCSI protocol problems due to malfunctioning in
345. ive Failed A drive was shut down 0x691 Checksum Error on Device read A checksum error was detected on a read from an individual disk There was no data transfer involved 0x692 Incoherent Stripe Data and parity were not consistent in a RAID group 0x693 Uncorrectable Stripe Inconsistent write or time stamps could not be corrected in RAID group 0x694 Parity Invalidated Parity has been invalidated in a RAID group 0x695 Uncorrectable Sector An uncorrectable sector was detected on a RAID 1 mirrored pair 0x696 Mirror Sector Invalidated A sector on a RAID 1 mirrored pair was invalidated on a rebuild Table 8 11 Thresholded Codes Event code 0x801 Description Recommended threshold Soft SCSI error This code indicates Notify the user after 25 errors day that an abnormal SCSI bus or disk FRU drive event was detected A retry of the operation cleared the condition Unsolicited Event Log Buljooysajqnoly Table 8 11 Thresholded Codes cont d owns this disk encountered a condition which required the firmware to invalidate a sector on the unit The firmware did this to ensure that incorrect data is not returned to the host in the future Table 8 12 Error Codes Event code Description Recommended threshold 0x802 Illegal SCSI Bus interrupt An Notify the user after 10 errors month inconsistent interrupt situation has FRU been detected on the controller 0x803 Recommend Disk Replacemen
346. k ck ck ck ck ck KKK KKK KKK KK ck ck ck KKK KKK KKK KKK KKK ck ck ck KKK ck ck ck KKK ck ck ck ck oko KKK KK ko ko ko ko ko ko KKK File name Intended Product Size SP 829 NIKE SP Controller Sauna Firmware 1310720 SE 925 00 NIKE SP Controller Sauna Firmware 1310720 SP 861 20 NIKE SP620 Controller Phoenix Firmware 1572864 SP 955 01 NIKE SP620 Controller Phoenix Firmware 1572864 NIKE2GHH03 NIKE Disk FRU Firmware 269824 NIKE2GLP08 NIKE Disk FRU Firmware 269824 NIKE4GHH03 NIKE Disk FRU Firmware 269824 NIKE4GLP53 NIKE Disk FRU Firmware 335872 NIKE9GHH53 NIKE Disk FRU Firmware 335972 Legend File name name of the firmware file Intended Product firmware file s intended product Size exact byte size of the firmware image Enter the firmware file name will display files Default will quit SP 955 01 7 64 NIKEARRY Host Utilities and Diagnostics S9 7 AIV AMIN 0 GI 8 01 O3UO I0 GG6 dS ezemuxrg peo uMop o4 3noqv x KKK KKK KK KK KK KK KK KK KK KKK KKK KK KK KK KK KKK KKK KKK KKK KKK KEKE KEKE KEKE iIngsseoons pe3ooqeu AMIN spuooes og ttc nn spuooes Op t6 spuooes og 006668 spuooes oz crt ttt tte spuooes oT 0066s 300qe43 O3 HMIN 30j Tem ssesetg OC OE OK OK XE OE ee eee eee XE OE OO eee ee eee eee eee eee eee ee ee ee eee ee X iATTNASSAOONS pepeo uMODp SAeMUAT Y HK KK KK KKK KKK KK KK KK KK KK KK KK KK KK KK KK KKK KKK KKK KKK KKK KKK KKK KK HH HY Tete eee eee eee ee eee eee eee ee eee eee e
347. k is bound to a RAID 0 LUN R 1 disk is bound to a RAID 1 LUN R 3 disk is bound to a RAID 3 LUN R 5 disk is bound to a RAID 5 LUN R 10 disk is bound to a RAID 1 0 LUN Hot disk is bound to a hot spare LUN AO B2 etc LUN ownership and LUN number The configuration section displays detailed LUN and disk information 7 74 Support Tools Manager STM CONFIGURATION amp Information Tool Log for C2300 Array on path 874 8 PR RTr Tupe Vendor Product ID AprxCap On C238BHDRS IBM D6HSB3Y IBM DGHSB8SY CLARGS IBM DGHSB8SY CLARGS IBM DGHS SY CLAROS IBM DGHS SY CLARGS RAID 5 C238B8HDRS SERGRTE ST15158N SERGRTE ST15158N SERGRTE ST15158N RRID 1 C2388HDR 1 IBM DCHSB84F CLARG4 IBM DCHSB84F CLARG4 Figure 7 5 Information Log Screen Configuration Section The information displayed includes the following Table 7 10 Information Log Screen Configuration Section Column Headings Column heading Description Lun LUN to which current disks are bound displays if no LUN C D Current Default ownership of the LUN displays if none Fru BS Disk FRU FRU bind slot if they exist Cch LUN read R on off write W on off cache ATr LUN auto trespass state on off hot spare Support Tools Manager STM 7 75 sonsoubeiq pue seninf ISOH Table 7 10 Information Log Screen Configuration Section Column Headings cont d
348. kup unit BBU N A A3231 60003 A3231 69103 Storage processor SP without SIMMs N A A3231 60004 A3231 69104 Power supply VSC N A A3231 60006 A3231 69006 Model 10 fan pack Reference R 15 o2uoJ9JoH Table R 1 Replaceable Parts Model 10 cont d N A 5062 9398 A3232 69022 various vendor numbers including MT12D136G 7 MH1M36ADJ 7 HB56D136BR HB56D136BV 7B MH1M36BDJ 7 THM361020ASG 70 MCM36100ASG70 KMM5361000AG KMM5361000BG 7 KMM5361000CG 7 MH1M36BYJ 7 KMM5361003CG 7 KMM5361003BG 7 TM124MBK36B 70 KMM5361203AWG 7 SM5361000WG 7DG 4 Mbyte SIMM N A 5062 9399 A3232 69020 numerous vendor numbers including HB56D436BR 7A MH4M36AXJ 7 MH4M36BXJ 7 HB56D436B 7RB SCM421186SG70 SM5364000G 7DG KMM5364100AG 7 SM536044002P3G7 KMM5364103AKG 7 16 Mbyte SIMM N A 005 040801 N A Model 10 AC line filter PCB N A 123 002825 N A Model 10 20 caster N A 5063 5397 N A Model 10 20 disk slot filler panel N A 005 041690 N A Model 10 rackmount front bezel N A R 16 005 041691 Reference N A Model 10 deskside front bezel Replaceable Parts Model 20 Reference A L EP 4d N Figure R 2 Parts Breakdown for the Model 20 Disk Array R 17 Reference Table R 2 Replaceable Parts Model 20 Figure index New HP E
349. lability SCSI Disk Array rackmount field racked e HP A3550AZ Model 20 High Availability SCSI Disk Array factory racked Models Discontinued June 1 1998 e HP A3539A Model 10 High Availability SCSI Disk Array deskside e HP A3540A Model 10 High Availability SCSI Disk Array rackmount field racked e HP A3540AZ Model 10 High Availability SCSI Disk Array factory racked Options Available Storage Capacity Options Note 9 Gbyte and 18 Gbyte disk modules are not supported on MPE iX or the model 10 disk array The options for the 4 Gbyte disk modules are not supported on the model 10 disk array e 341 59 Gbyte LP disk modules for model 20 e 351 518 Gbyte HH disk modules for model 20 e 361 54Gbyte LP disk modules for model 20 Order up to four of the above options to fill the array High Availability Options e Option 204 2 SPs with BBU and 64 Mbytes of cache per SP model 20 only Reference R 5 o2ue49JoH e Option 205 Single SP with 64 Mbytes of cache model 20 only e Option 206 Single SP with 8 Mbytes of cache model 20 only e Option 421 Add on power supply required for full power redundancy Patch Media Options e Option AA4 HP UX 9 04 patch QIC format e Option AAF HP UX 9 04 patch CD format e Option AAH HP UX 9 04 patch DDS format SCSI 2 Cable Options e Option 801 9m68 pin HD male to 68 pin HD male cable e Option 802 2 5m 68 pin HD male to 68 pin HD male cable e Option 803 5 0 m68 pin HD male to 68 pin HD
350. later software mirroring the RAID 0 group provides high availability Note RAID 0 is not supported on models A3549 or A3550 because it does not provide data redundancy For microcode revision 8 56 and later this RAID level can only be configured in FE mode with the Grid Manager utility See chapter 6 Grid Manager for more information about FE mode Stipe Blocks User data Q 384 511 768 895 1152 1279 1536 1663 First Module O Stripe Size lt 128 255 512 639 896 1023 1280 1407 1664 1791 Second Module 256 383 640 767 1024 1151 1408 1535 1792 1919 Third Module O Figure 1 12 Distribution of User Data in a RAID 0 Group Planning Disk Unit Configurations 1 25 uonduioseg jonpold RAID 1 Mirrored Pair This arrangement consists of two disk modules bound as a mirrored pair The disks in a RAID 1 pair are mirrored by the disk array hardware The disk array writes the same data to both disk modules in the mirrored pair The mirror cannot be split into its individual units A software mirror composed of two individual disk units can be split In a RAID 1 configuration the hardware mirrors the two disk modules automatically RAID 1 hardware mirroring within the disk array is not the same as hardware mirroring for other kinds of disks Functionally the difference is that mirroring cannot be stopped manually on a RAID 1 mirrored pair To use one of the disk modules in such a mirror s
351. le The fan pack is customer replaceable Individual fans within the fan pack however cannot be changed The fan pack can be replaced while the disk array is operating although there is a limited amount of time about two minutes during which the disk array can continue to operate without cooling before the array shuts itself down The fan pack on both the deskside and rackmount models pivots for easy access to the power supply units and SPs The fans run on DC voltage 18 VDC minimum to 26 VDC maximum supplied by the power supplies When the disk array is operating normally the fans run at 18 VDC If power fails and the BBU begins operating to supply power the fans speed up because the BBU supplies the maximum DC voltage 26 VDC to the fan pack s Figure 1 9 illustrates the fan packs About Disk Array Components 1 13 uonduioseg jonpold Sue NS N Ys y Figure 1 9 Fan Packs for the Model 10 and Model 20 Disk Arrays Power Supply Units The power subsystem consists of either one or two power supply units voltage semi regulated converters VSCs in the model 10 or two or three power supply units in the model 20 disk array The power supplies are auto ranging Depending on configuration the units can be N 1 redundant one more component than is needed for normal operation The number of power supplies in the subsystem depends on the number of disk modules e Five to 10 disk
352. le The System Administrator can transfer control to the other SP and adapter by changing host s and array configuration If the SCSI 2 adapter or SCSI 2 cable fails can regain access to the LUNs associated with the failed SP host adapter or SCSI 2 cable only if changes are made manually in Grid Manager and LVM Multi Bus Initiator Configuration HP UX 10 x and Later CONFIGURATION HP3000 SERIES 900 MPE iX Considerations Disk Array Configurations Basic Configuration for Model 10 and Model 20 Disk Arrays Two SP Configuration Two Adapter Two SP Configuration Multi Bus Configuration Configuring Disk Array SCSI LUNs for MPE iX 5 1 006 Seuos 000 dH uoneanByuo2 MPE iX Considerations For the MPE iX user who is new to disk arrays here are some tips e With conventional SCSI disks there is a one to one mapping of individual physical disk drive modules to logical device Ldev in MPE iX terminology or logical unit LUN in SCSI terminology e A maximum of eight LUNs are recognized by MPE X 5 5 e With a RAID disk array the Grid Manager discussed in chapter 6 can be used to associate logically or bind individual physical disk modules into a LUN e The HP 3000 Series 900 views the LUN as a single disk or Ldev For example the System Software Installation and Maintenance Manual part number 30216 90223 describes how to determine what drives are on the system e An
353. le 9 2 Model 20 Disk Array CE Replaceable Parts see Figure 9 2 cont d N A 5062 9398 A3232 69022 various vendor numbers including MT12D136G 7 MH1M36ADJ 7 HB56D136BR HB56D136BV 7B MH1M36BDJ 7 THM361020ASG 70 MCM36100ASG70 KMM5361000AG KMM5361000BG 7 KMM5361000CG 7 MH1M36BYJ 7 KMM5361003CG 7 KMM5361003BG 7 TM124MBK36B 70 KMM5361203AWG 7 SM5361000W 7DG 4 Mbyte SIMM N A 5062 9399 A3232 69020 various vendor numbers including HB56D436BR 7A MH4M36AXJ 7 MH4M36BXJ 7 HB56D436B 7RB SCM421186SG70 SM5364000G 7DG KMM5364100AG 7 SM536044002P3G7 KMM5364103AKG 7 16 Mbyte SIMM N A 008 040535 N A DC power distribution cable N A 008 040685 N A Address RS 232 PCA N A 111 001497 N A Display orientation jumper N A 123 002825 N A Model 10 20 caster N A 5063 5397 N A Model 10 20 disk slot filler panel N A 005 041695 N A Model 20 rackmount front bezel N A 005 041696 N A Model 20 deskside front bezel Illustrated Parts Breakdown for the Model 20 9 7 1ueuieoe deu pue jeAoway Model 10 20 Disk Module Removal Replacement Install a replacement or add additional disks by performing the following steps CAUTION Disk modules having different storage capacities or different model numbers should not be bound into the same RAID group Replace a disk module with one having the same model number Disk modul
354. le Auto Trespass current state 2 ON 1 Turn Auto Trespass OFF 2 Turn Auto Trespass ON 99 Cancel command Enter Selection 1 2 99 99 gt 99 AUTO TRESPASS UTILITY EXITED BY USER 4 19 PM 7 14 HARRAY Section 17 External Exerciser CAPACITY Non exclusive non destructive The CAPACITY command is used to determine the capacity of a LUN Example HARRAY gt capacity PDEV 24 14 0 is an C1300WDR5 CAPACITY UTILITY 4 04 PM Max Block Address 12219263 Current Block Size 512 bytes LUN Capacity 6256263168 bytes 5966 MB CAPACITY UTILITY COMPLETED 4 04 PM HARRAY Section 17 External Exerciser 7 15 sonsoubeiq pue seninf ISOH CLEAR LOGS Non exclusive non destructive The CLEAR LOGS command is used to clear logs Example HARRAY gt clear logs WARNING This will Clear ALL Unsolicited Logs Continue Y N N y Clear Logs Command SUCCESSFUL All logs have been cleared CLEAR LOGS UTILITY COMPLETED 4 05 PM 7 16 HARRAY Section 17 External Exerciser CONTROLLER REP Non exclusive non destructive The CONTROLLER REP command is used to access information about the controller the current software level and some other controller parameters and configurations Example HARRAY gt cont rep PDEV 24 14 0 is an C1300WDR5 CONTROLLER REPORT UTILITY 4 06 PM Cabinet ID 3 C1100 or C1300 RACK Controller Model Number 7341 Firmware Major Revision 8 Firmware Minor Revision
355. ler Product Revision Level INQUIRY UTILITY COMPLETED supports Asynchronous Event Notification as a target a processor type initiator does not support the Terminate I O Process message does not support relative addressing does not support 32 bit wide SCSI does support 16 bit wide SCSI supports synchronous data transfers does not support Linked CDB s supports tagged command queuing responds to the RESET condition with a Hard of the SCSI Interface DGC C1300WDR5 HP02 4 08 PM HARRAY Section 17 External Exerciser LUN REPORT Non exclusive non destructive Displays status information for the LUN under test Example HARRAY gt LUN REPORT LUN REPORT UTILITY 4 09 PM Current SP Accessed A SCSI Back End Address 14 Peer SP is Present This LUN is configured as a RAID 5 Disk Array SCSI Bus Number Backend 0 SCSI Device ID Backend 0 FRU Status 7 FRU Enabled Number of Hard Read Errors 0 Number of Hard Write Errors 0 Number of Soft Read Errors 0 Number of Soft Write Errors 0 Hit CR to continue or E to Exit FRU SLOT NUMBER AO cont Average Disk Request Service Time ms 5 Average Address Difference sectors 2B Number of Reads 3E86 Number of Writes 65A9 Number of Read Retries 0 Number of Write Retries 0 Number of Sectors on FRU 3EB6D2 Number of Sectors of User Space 3E2680 Number of Sectors of Private Space 9000 Maximum Requests in Queue
356. lights all modules in the logical disk unit when one of the modules is selected If an attempt is made to unbind a logical disk unit owned by the other active SP the following message appears Dual board unbind error 6 Press the RETURN key to select the highlighted logical disk unit for unbinding Grid Manager warns that unbinding destroys all data on the logical disk unit and then asks for confirmation to unbind this logical disk unit Unbinding a Logical Disk Unit LUN 6 51 Jobeue N pus 7 To unbind the logical disk unit type Y and press RETURN otherwise cancel the unbind operation by typing N 8 Grid Manager unbinds the LUN For each module the status changes from ENA or RDY to UNB and the unit number to the unbound module symbol e xterm JP A Unbind Logical Units SP SCSI ID 15 S N 9621954 Unbinding Mill DESTROY All Data on this Unit Unbind this unit Y N N ff Figure 6 12 Unbinding a Logical Disk Unit To unbind another LUN return to step four To bind newly unbound modules into another logical disk type caret twice to return to the Main menu and then refer to Creating Logical Disk Units LUNs on p 43 of this chapter 6 52 Unbinding a Logical Disk Unit LUN Setting up Disk Array Caching for Microcode Revision 7 23 The disk array supports caching if it has the following modules Two SPs each with at least 8 Mbytes of memory A fully charged BBU Disk modules in slots A0 A1 A2
357. ling executing or displaying Software on a computer processor or controller or making a copy of Software for archival or backup purposes only c Products include hardware Software options documentation accessories supplies spare parts and upgrades on HP s Price List on the date HP receives Customer s order LICENSES a Inreturn for the associated fee HP grants Customer a non exclusive license to Use one copy of the Software listed in the Customer s order in conformance with the applicable Software Licenses specified in the Price List HP s standard Software Licenses are described below Other types of Software Licenses may be made available for some Software If no license is specified for Software then in return for the applicable fee HP grants Customer a license to Use one copy of the Software on one computer processor or controller at any one time SYSTEM Customer may use the Software at any one time 1 onany one computer or processor Basic System License 2 on any one ofa class of computers or processors Class License or 3 onasingle registered computer or processor Nodelocked License Some System Licenses may be limited to a maximum number of users as described in the Price List CONCURRENT USER An identified number of user may Use the Software concurrently at any time TEMPORARY Customer may Use the Software only an identified period of time Reference RUNTIME Customer may Use only the e
358. ll fault If this happens the BBU can be removed and reinserted to start another charging cycle Do not exceed two charging cycles If the batteries do not recharge in two charging cycles replace the BBU Battery life is two to five years BBUS are customer replaceable and hot replaceable A failed BBU can be replaced while the disk array is operating without disruption however the cache is disabled if the BBU fails or is removed Hot swap of a working BBU on a model 20 running SP microcode revision 7 23 is not supported Figure 1 11 Battery Backup Unit 1 16 About Disk Array Components SCSI 2 Adapter The SCSI 2 adapter is not a disk array component It is a printed circuit board that resides in an I O slot in the host computer It transfers data between host memory and the external Fast Wide differential SCSI 2 bus which connects to the disk array One or more SCSI 2 host adapters can be installed in the host computer although the S700 supports only one SCSI 2 adapter connected to a disk array An external SCSI 2 bus cable connects a SCSI 2 adapter to the SP s in the disk array cabinet Only service personnel authorized to open the computer and replace parts should install or replace SCSI 2 adapters in the host computer About Disk Array Components 1 17 uonduioseg jonpold Disk Array High Availability Features RAID Level 5 1 or 1 0 Data Redundancy RAID Redundant Array of Inexpensive Disks technology suppor
359. lly completed Controller restarting please wait approx 2 minutes 20 Percent Complete 40 Percent Complete 60 Percent Complete 80 Percent Complete 100 Percent Complete Diagnostic Status OK no errors INITIATE DIAGNOSTIC UTILITY COMPLETED 5 52 PM HARRAY Section 17 External Exerciser 7 19 sonsoubeiq pue seninf ISOH DOWNLOAD Exclusive non destructive Use the DOWNLOAD command to download new SP microcode and disk module firmware e The host system to which the disk array is attached must be quiescent If the array is connected in a multiple host configuration all the host systems must be quiescent e Disable all cache on all SPs Use the Grid Manager Change Parameters menu to set the cache size to zero before starting the download e Allthe filesets on the array must be unmounted e The HARRAY diagnostic can be run from single user mode or all volume groups on the array must be deactivated To deactivate volume groups use the HP UX Ivchange a n and vgchange a n commands Updating Storage Processor Microcode for Non Root Disk Arrays To update the SP microcode for example from revision 7 23 to revision 8 20 1 Loadthe new microcode file to a known directory from the update media DDS format update media P N B3108 13471 or get FTP information from WTEC 2 Change the run level to single user on all hosts attached to the disk array whose firmware is being updated or shut down all applications processes an
360. logical disk unit is the default owner Host computer Disk array SCSI 2 bus SCSI 2 SPA adapter System RS 232 console EL SCSI 2 SCSI2 bus A A adapter SPB RS 232 System disk To DTC port or terminal Figure 5 3 Model 10 20 Two Adapter Two SP Configuration Two Adapter Two SP Configuration 5 9 006 Seles 000 dH UOHeunByuoD Table 5 4 Two Adapter Two SP Configuration Error Recovery Failing component Disk module Continue after failure Yes What happens and how to recover With RAID of any level supported by MPE ix applications continue running System Administrator or service provider replaces module Provides the same features as the basic configuration but with better performance and higher availability With RAID of any supported level applications can continue to run after the failure of a single disk module SP No If an SP fails the System Administrator must reassign ownership of the failed SP s LUNs to the operational SP by using Grid Manager Also the path to the operational SP must be used and can be changed using SYSGEN or ioconfig Once these changes have been made access to all LUNs on the array within the SCSI LUN limits is possible LUN ownership need to be reassigned by using Grid Manager Fan pack Yes Applications continue running System Administrator or service p
361. m to be bound back to the right slots The two key factors are that cache must be changed to zero and the SPs must be rebooted twice during this process It is recommended that FE mode be used as it provides all the available capabilities Unsolicited Event Log 8 69 punoouse qnouj Decoding Extended Status Codes Extended status codes are reported in the Unsolicited Event Log display in Grid Manager Extended status codes represent abnormal activity either on the buses or the disks that are attached to the internal SCSI controller channels of the SP When an error occurs the extended status codes and CRU event messages are listed in the disk array s Unsolicited Event Log Extended status codes are classified as either hard or soft If the status corresponds to a number in the extended status code list it is classified by the SP microcode as a soft status A hard status is represented by adding 0x40 to the extended status code For example 0x01 is a soft status and 0x41 is a hard status A hard status of 0x41 most likely indicates that a solid fault has occurred in the array An extended status code of 0x6A means a disk module has faulted The 0x6A code is a hard status of extended status code 0x2A The 0x2A status is a SCSI protocol error which states that an error was detected during activity on the SCSI 2 bus between the SP and the disk and that a retry was unsuccessful When the 0x6A extended status code is logg
362. microcode revision 8 56 and later these RAID levels can only be configured in FE mode with the Grid Manager utility See chapter 6 Grid Manager for more information about FE mode CAUTION Disk unit configurations must be planned carefully Once the disk modules are bound they cannot be unbound without losing the data on them To change a physical disk unit it must be unbound rebound and again made accessible to the operating system To decide when to use a RAID 5 group RAID 1 mirrored pair RAID 1 0 group RAID 0 group individual disk unit or hot spare the following factors must be weighed e Importance of data availability e Importance of performance 1 38 Planning Disk Unit Configurations Amount of data stored Cost of disk space RAID 5 Group Individual Access Array A RAID 5 group is recommended for applications in which the following are true Data availability is very important Large volumes of data will be stored Multi tasking applications use I O transfers of different sizes A database uses a normal or less than normal percentage of write operations where writes are 33 or less of all I O operations Write caching can significantly enhance the write performance of a RAID 5 group Multi tasking applications require a large history database with a high read rate Transaction processing is needed RAID 1 Mirrored Pair A RAID 1 mirrored pair is recommended for applications in which the following ar
363. minal Figure 4 4 Multi Bus Configuration Multi Bus Configuration 4 11 008 seues 0006 dH uoneanbiuo2 Table 4 5 Multi Bus Configuration Error Recovery Failing Continue after component failure What happens and how to recover Disk module Yes 9 04 Provides the same features as the basic configuration but Yes 10 x with better performance and higher availability With RAID of any level other than 0 or an individual disk applications continue running System Administrator or service provider replaces module SP No 9 04 If one SP fails the System Administrator can access the LUNs Yes 10 x owned by the failed SP by accessing the remaining operational SP with the other host If a host adapter fails the System Administrator can access the LUNs in the array with the other host adapter The System Administrator must change the path and the SP ownership of the LUNs for the failed SP or failed host adapter by using LVM and Grid Manager respectively HP UX 10 x There is no automatic transfer of LUN ownership or path for this configuration type Fan pack Yes 9 04 Applications continue running System Administrator or Yes 10 x service provider replaces module Power supply Yes 9 04 If a redundant power supply module is present applications Yes 10 x continue running otherwise disk array fails Service provider replaces power supply The System Administrator restarts the operating system if ne
364. mmary Displays the current configuration of all bound LUNs in the disk array 5 Display Disk Failure Information Displays information on the most recent disk failures Can be viewed in FE mode only 6 Display Non Vol RAM Unsolicited Error Log Displays the last eight errors seen by the SP Unlike the other Unsolicited Event Log entries this log is stored on the SP in the Non Vol RAM Can be viewed in FE mode only Display Unsolicited Event Log To view the Unsolicited Event Log from the Grid Manager Main menu follow the menu path 4 View Maintenance Information 1 Display Unsolicited Event Log The Unsolicited Event Log describes a variety of disk array events in chronological order oldest to newest Periodically the SP writes this log to disk to retain it when disk array power is off The Unsolicited Event Log holds over 2 000 entries 8 44 Viewing Maintenance Information The Unsolicited Event Log and the different types of events are described in detail in Unsolicited Event Log on p 53 of this chapter For listings and descriptions of event codes see Table 8 10 through Table 8 13 xterm Ei x SP A Display Unsolicited Event Log SP SCSI ID 15 S N 9621954 Event Date CRU Event Hessage gt Extended Status 1 04 24 97 14 08 51 IB Oxa08 Database Sunc Error 0x03 2 04 24 97 14 09 08 AS Oxa08 Database Sync Error 0x03 3 04 24 97 14 11 23 IB Oxa08 Database Sync Error 0x03 4 04 24 97 14 11 23 A3 Oxa08 Dat
365. module A3 fails Immediately the SP begins rebuilding the RAID 5 group using the hot spare When it finishes the RAID 5 group consists of disk modules A1 A2 B2 A4 and B1 No hot spare is now available When the failed module in A3 is replaced the SP starts copying the structure on B2 to A3 When it finishes the RAID 5 group once again consists of modules A1 through A4 and B1 and the hot spare in B2 becomes available for use if any other module fails A similar sequence would occur if module AO in the mirrored pair failed In summary Planning Disk Unit Configurations 1 33 uonduioseg jonpold e The RAID 5 group is Al A2 A3 A4 and B1 the RAID 1 mirrored pair is AO and BO the hot spare is B2 e Disk module A3 fails e The RAID 5 group is Al A2 B2 A4 and B1 no hot spare is available e Failing disk module A3 is replaced with working module e The RAID 5 group is Al A2 A3 A4 and B1 the hot spare is B2 Hot all is Spare Deskside Model Rackmount Model Figure 1 18 Example of a Hot Spare in a Model 10 Disk Array Hot Spare for a Model 20 In this example the modules in slots
366. mount Mode Figure 1 19 Example of a Hot Spare in a Model 20 Disk Array Data Availability and Disk Space Usage A RAID 5 group provides high data availability but requires more disk modules than a mirrored pair In a RAID 5 group of five disk modules 80 of the disk space is available for user data and 20 is for parity information A RAID 5 group is usually more suitable than a RAID 1 mirrored pair for applications in which high data availability good performance Planning Disk Unit Configurations 1 35 uonduioseg jonpold and efficient disk space use have relatively equal importance Write caching significantly enhances the write performance of a RAID 5 group A RAID 1 mirrored pair or RAID 1 0 group provides very high data availability It also is expensive because only 50 of the total disk capacity is available for user data The remaining 50 is the mirror image of the user data A RAID 0 group non redundant individual access array provides all its disk space for user files but does not provide any high availability features For high availability it is necessary to use a RAID 1 0 group or a software mirror with a RAID 0 group A RAID 0 group can be mirrored in HP UX 10 x and later A RAID 1 0 group or a software mirrored RAID 0 group provides the best combination of performance and availability at the highest cost per Mbyte of disk space A RAID 1 0 group offers higher reliability than a software mirrored RAID 0 group
367. mple of a model 20 configured into four physical disk units LUNs The physical disk units are 0 through 3 The SCSI address of the SP is 0 Each physical disk unit is configured as a RAID 5 group Note that IOMAP does not indicate how many disks are in each group Path Component Name 8 HP PB Fast Wide SCSI Approx 10 seconds Resetting SCSI Bus Unknown SCSI device type C2300WDR5 Unknown SCSI device type C2300WDR5 Unknown SCSI device type C2300WDR5 Unknown SCSI device type C2300WDR5 OO Oe x WNrR Oo IOMAP 7 3 sonsouDeig pue seni ISOH Where the digits in the Path 8 0 2 for example are leftmost digit host adapter address 8 in the example middle digit SP SCSI address 0 in the example rightmost digit physical disk unit number LUN 2 in the example 7 4 IOMAP High Availability Disk Array Monitor HARAYMON The high availability disk array monitor HARAYMON monitors FRUs within the disk array and notifies the system of all FRU failures For HARAYMON to report errors correctly the disk activity must be taking place and in addition either be configured in a high availability or redundant configuration RAID 5 or RAID 1 0 for example or if configured in a non high availability configuration RAID 0 for example have more than one active LUN When HARAYMON detects a FRU failure it displays a message on the system console and sends an e mail message to the root user on the host system The e mail function i
368. n Info menu option details this information The Information Tool menu options are shown in Figure 7 27 7 100 Support Tools Manager STM e hpterm File System Device Tools Options Help salt rose hp com 10 4 4 Arra TATATATA TATATA TATATATA TATATATA TATATATA TATATA TATATATA TATATATA TATATA TATATA TATATA TATATATA TATATATA TATATATA TATATATA TATATATA ATATEN TATATATA TATATATA fusr sbin stm ui bin stm Product Bus Adapter 580 SCSI Interface 145 C2300 Array amp DGCC2300HDR5 C1300 Array DGCC1S00WDRS C2300 Array DGCC2300HDR5 Fibre Channel Interface FC Storage lt FCPProtocol Bus Adapter 4580 SCSI Interface 15 gt SCSI Disk lt SEAGATEST34371 SCSI Disk lt SEAGATEST34371 SCSI Disk lt SEAGATEST34371 SCSI Disk lt SEAGATEST34371 Bus Adapter 45005 NIO Terminal Multiplexor NIO Fast Wide SCSI Interf Figure 7 27 mstm Device Map and Menu The following screens depict how the Information Log is organized The log is separated Last Last Op Active Tool Status into three sections map configuration and disks The map section presents an overview of the SCSI disk array s current status This includes fan power supply and BBU status cache usage and microcode revision The SCSI disk array s physical and logical layout comprises cabinet orientation the location of disks and LUN binding and ownership Figure 7 28 shows the map section of the Information Log Use the Up and Down arro
369. n chapter 6 If a model 10 disk array has the minimum number of disk modules configured reconfiguration is optional not required If more than the minimum number of disk modules are configured in a model 10 or a model 10 is configured the disk array must be reconfigured To reconfigure all LUNs must be unbound To unbind LUNs go to the Grid Manager Main menu by typing the caret character and pressing RETURN Select option 3 Unbind Physical Units Then follow the procedures in Unbinding a Physical Disk Unit LUN in chapter 6 Be sure to unbind all LUNs that were already configured in the disk array when it was received After unbinding all LUNs return to the Grid Manager Main menu by typing and pressing RETURN Select option 2 Bind Physical Units Follow the procedures in Creating Physical Disk Units LUNs in chapter 6 to reconfigure the disk array When selecting a RAID mode it is very strongly recommended that only the data protecting modes which are RAID 1 RAID 1 0 and RAID 5 be used Other RAID modes do not protect data which can result in permanent data loss in the event of a disk module failure When reconfiguring the disk array the following items must be taken into consideration e Operating system patches see your WTEC representative for the latest patch information Installing and Reconfiguring on S700 Systems 3 3 00Z Seues 0006 dH uoneunDuuo25 e The maximum number of individual LUNs that can be configured
370. n explaining circumstances surrounding the failure SP s microcode and PROM levels Number of SPs in disk array LUN configuration and LUN RAID modes Presentation screen output preferable Cache parameter settings Firmware level on disk mechanisms Inventory screen output preferable For output Disk Array Installation Troubleshooting Checklist Revision 1 3 U a zd type the letter i or the character at the Presentation screen prompt Host operating system level and installed patch list Type of host system used Type of F W host adaptor used and firmware level if applicable System map Total number of disk arrays on site ___ Number of disk arrays connected to each host adaptor Diagram of SCSI cable connections from host to disk array s FE mode Unsolicited Event Logs for both SPs at least five days prior to the failure Host system I O logs around time of failure List of disk array parts already changed Disk Array Installation Troubleshooting Checklist Revision 1 3 2 47 uonejjeisuj 2 48 Disk Array Installation Troubleshooting Checklist Revision 1 3 CONFIGURATION HP 9000 SERIES 700 Disk Array Support Requirements for HP UX Release 9 05 9 07 Installing and Reconfiguring on S700 Systems Disk Array Configurations Basic Configuration for Model 10 and Model 20 Disk Arrays Two SP Configuration Multi Bus Configuration HP UX 10 x and Later 3 1 004 Sales 0006 dH uoneun
371. n on the new midplane CAUTION When re installing the midplane in the chassis make sure the midplane is seated before tightening screws Make sure the DC power distribution cable is not pinched beneath the midplane Top metal jRetaining screws 11 cover E Li Figure 9 53 Removing the Top Cover 9 92 Model 20 Rackmount Component Removal Mounting bracket Retaining screw Retaining screws A P1 jumper location Midplane Figure 9 55 Removing the Midplane from the Chassis Model 20 Rackmount Component Removal 9 93 u w vejd y pue jeAoway RIP PCA Figure 9 56 1 Noe a Fw WN 9 94 Power off the disk array and disconnect all cables The fans continue to run for a few seconds Access the rear of the disk array Remove the SPs Figure 9 9 Remove the four retaining screws that hold the chassis to the rack Figure 9 52 Remove the empty chassis from the rack Figure 9 52 Remove the top metal panel 11 screws Figure 9 53 Remove the RIP PCA a Remove the two screws that secure the midplane to the chassis One screw at each end of the top of the midplane b Remove the RIP PCA retaining screw c Pull the midplane out approximately 1 2 inch d Tilt the midplane approximately 2 inches e Remove the RIP PCA from its connector Model 20 Rackmount Component Removal Mounting bracket N Retaining screw Figure 9 56 R
372. n rack 1 10 110 4 7 Yes 2 10 110 8 14 Yes 3 10 110 12 21 Yes 4 10 110 16 28 No E amp M 5 10 110 20 35 No E amp M 1 10 220 2 7 Yes 2 10 220 4 14 Yes 3 10 220 6 21 Yes 4 10 220 8 28 No M 5 10 220 10 35 No M 1 20 110 8 9 Yes 2 20 110 16 18 Yes 3 20 110 24 27 No E amp M 4 20 110 32 36 No E amp M 1 20 220 4 9 Yes 2 20 220 8 18 Yes 3 20 220 12 27 No M 4 20 220 16 36 No M No E amp M Not electrically or mechanically possible to install this number of disk arrays in the rack No M Not mechanically possible to install this number of disk arrays in the rack 2 10 Install a Rackmount Disk Array in a Cabinet Table 2 2 1 6 m Cabinet Number Current Supported of arrays Model VAC amps EIA units in rack 1 10 110 4 7 Yes 2 10 110 8 14 Yes 3 10 110 12 21 Yes 4 10 110 16 28 Yes 5 10 110 20 35 No E amp M 1 10 220 2 7 Yes 2 10 220 4 14 Yes 3 10 220 6 21 Yes 4 10 220 8 28 Yes 5 10 220 10 35 No M 1 20 110 8 9 Yes 2 20 110 16 18 Yes 3 20 110 24 27 No E 4 20 110 32 36 No E amp M 1 20 220 4 9 Yes 2 20 220 8 18 Yes 3 20 220 12 27 Yes 4 20 220 16 36 No M No E amp M Not electrically or mechanically possible to install this number of disk No M No E arrays in the rack Not mechanically possible to install this number of disk arrays in the rack Electrical limitation This number o
373. n two hosts With RAID levels other than 0 or an individual disk applications can continue to run if a disk module fails If a host adapter fails the adapter on the other host can only access its own LUNs The System Administrator must manually change the path and SP ownership of the LUNs for the failed SP or failed host adapter by using Grid Manager No automatic transfer of LUN ownership or path takes place for this configuration type Disk Array Configurations 3 5 004 seues 0006 dH uoneunDuuo25 Basic Configuration for Model 10 and Model 20 Disk Arrays The basic configuration has one host with one SCSI 2 adapter connected by an external Fast Wide differential SCSI 2 bus to the SP in the array as shown in Figure 3 1 The system can survive the failure of a disk module within a redundant RAID group or mirror but cannot continue after failure of the SCSI 2 adapter or SCSI 2 cable see Table 3 2 for more information Host computer Disk array adapter RS 232 System console To MUX port or Pide terminal Figure 3 1 Model 10 20 Basic Configuration 3 6 Basic Configuration for Model 10 and Model 20 Disk Arrays Table 3 2 Basic Configuration Error Recovery Failing Continue after component failure What happens and how to recover Disk module Yes With RAID of any level other than 0 or an individual disk applications continue running System Admi
374. nel of the chassis Figure 9 30 Model 10 Rackmount Component Removal 9 59 jueuie2e daH pue jeAoway Remove the inner chassis Figure 9 31 a Remove the retaining screws from each side of the chassis six on one side seven on the other b Disconnect the LED cable from the midplane c Lift the inner chassis up and out of the system i Rotate upward the rear of the inner chassis about 3 inches ii Lift up and out Unplug the fan cable Lift the cable out of the cable retainers Lift the cable out of the chassis Note The cable is keyed by a red stripe located at the bottom of the cable at the 9 60 midplane connector and as the cable is routed through the chassis The red stripe is positioned to the outside of the chassis at the fan pack When installing the fan cable in the chassis connect the fan end first so the red stripe faces to the outside of the system and the plug faces downward The fan cable must be in the cable retainers and lie flat or the SP cannot be installed on that side Model 10 Rackmount Component Removal Midplane connector Midplane Red stripe Figure 9 34 Removing the Fan Cable Model 10 Rackmount Component Removal 9 61 jueuie2e deH pue jeAoway SCSI 2 Cables Figure 9 35 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds Access the rear of the disk array Use a small screwdriver to remove tw
375. nent status fan pack power supply units BBU etc e Date and time SCSI disk array e Disk module information by slot location vendor product identification bind slot e Firmware revisions disk module e Microcode revisions SP Support Tools Manager STM 7 69 sonsouDeig pue seni ISOH e LUN binding RAID type LUN number current default ownership associated disk modules e System serial number SCSI disk array e Hardware Path The Information Tool provides access to the following three logs and detailed descriptions of the contents of the Information Log Selecting any of the logs displays the information in anew window e Information Log e Activity Log e Failure Log To use the Information Log first build it with Tools Information Run The log displays after it is built If it doesn t display it with Tools Information gt Information Log The Information Log contains map configuration and disks information The Tools Information Info menu option details this information The Information Tool menu options are shown in Figure 7 3 7 70 Support Tools Manager STM Canc hel Iis brar calle hg cos Ihe corruni hori has been pri bo sali rires Pp com AP Ody Deane bvrizlad fer tale ugankar Figure 7 3 Information Tool Menu Options The following screens depict how the Information Log is organized The log is separated into three sections map configuration and disks The map secti
376. nformation Engineering MS5668 Hewlett Packard Company Enterprise Storage Business Unit 8000 Foothills Boulevard Roseville CA 95747 5668 NO POSTAGE NECESSARY IF MAILED IN THE UNITED STATES Tape Fold Here Please Do Not Staple Tape
377. ng checklist disk array 2 45 disk array functionality and configuration 2 46 physical damage sitepower site environment 2 44 required information for escalated sites 2 46 SCSI and RS 232 cables and SCSI terminators 2 45 internal AC harness removal 9 43 9 63 9 84 9 103 IOMAP 7 2 ioscan shows 2 LUN Os if 2 SPs on HP UX 9 0x 6 67 IPCB removal 9 35 J jumper cable 2 21 1 4 Index L LEDs 1 9 8 7 LEDs LED cable removal 9 37 9 57 9 76 9 96 load balancing 1 42 locations of SCSI 2 bus connectors 2 23 logical disk configurations 1 38 logical disk units binding 6 43 logical volumes creating 7 131 using 7 131 LOGTOOL 7 2 LUN Logical Unit Number binding 6 43 in STM 7 85 7 114 capacity 7 84 7 86 7 113 7 115 creating 6 43 inquiry 7 84 7 86 7 113 7 115 reading 7 84 7 88 7 89 7 90 7 113 7 117 7 119 tips for MPE AX 5 2 unbinding 6 51 in STM 7 85 7 114 verify 7 84 7 94 7 95 7 118 7 121 7 123 LUN Configuration Summary 8 49 LUN REPORT HARRAY command 7 29 LVM 1 22 7 129 M maintenance information viewing 8 44 MECH HARRAY command 7 31 memory configuration in STM 7 85 7 97 7 114 7 125 microcode rev 8 xx 6 60 microcode R 2 1 11 1 33 6 34 6 37 6 45 7 20 7 21 7 39 8 10 8 16 8 19 8 25 8 27 8 30 8 35 panics 8 75 power up sequence 8 13 revision on Presentation screen 6 34 updates 7 82 7 110 midplane removal 9 31 9 72 9 91 mirror hardware 1 26 software 1 26 mirrored pair 1 1
378. ng the address cable to the address PCA the red trace on the cable is positioned at the bottom of the connector When connecting the address cable to the midplane the red trace on the cable is positioned at the bottom of the connector This connector is not keyed The red trace wire must always face the down side on a rackmount model or face the right side of a deskside model 4 Remove two screws from the midplane and tilt it Figure 9 55 5 Disconnect the address cable from the midplane 6 Disconnect the two SCSI 2 cable connectors from the distribution panel 7 Disconnect the SCSI 2 cables from the midplane 8 Remove the cables from the chassis Note Be careful when connecting the SCSI 2 cable to the midplane The connectors are 68 pin connectors and are fragile Model 20 Rackmount Component Removal 9 101 jueuie2e deH pue jeAowaYy pc gt NS S LXX S D Red trace wire Address cable Figure 9 60 Removing the Address and SCSI 2 Cable 9 102 Model 20 Rackmount Component Removal S ec FH SCSI cables Distribution panel Four retaining screws Internal AC Harness Figure 9 61 and Figure 9 62 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds Access the rear of the disk array Remove the following a Fan pack Figure 9 6 b SP Figure 9 9 c Power supplies Figure 9 11 d BBU located under the power s
379. ng the screws Check that the DC power distribution cable is not pinched beneath the midplane Inner metal cover EHNA MTT 1r Retaining screws 11 Figure 9 41 Removing the Inner Metal Cover Model 20 Deskside Component Removal 9 73 jueuie2e deH pue jeAoway Retaining screw gt Ne Midplane connector Tilt the midplane A Mounting bracket P1 jumper location Figure 9 43 Removing the Midplane from the Chassis 9 74 Model 20 Deskside Component Removal RIP PCA Figure 9 44 1 gu mw qe Power off the disk array and disconnect all cables The fans continue to run for a few seconds Remove the outer covers Figure 9 40 Remove SPs Figure 9 9 Remove the inner metal cover 11 screws Figure 9 41 Remove the RIP PCA from the midplane Figure 9 44 a Remove the RIP PCA retaining screw b Remove the RIP PCA from its connector Retaining screw x L Mounting bracket Midplane connector Figure 9 44 Removing the RIP PCA Model 20 Deskside Component Removal 9 75 u w vejd y pue jeAoway LEDs LED Cable Figure 9 45 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Remove the outer covers Figure 9 40 3 Remove the inner metal cover 11 screws Fig
380. nge Unit Auto Assign Displays dialogue for changing auto assign 4 Change Unit Caching Parameters Displays dialogue for changing unit caching 2 Change Storage System Parameters 1 Change Storage System Caching Parameters Change caching parameters N Y Write Cache State Enable Disable Disable Cache Page Size in KB 2 4 8 16 2 Read Cache State Enable Disable Disable Peer Read Cache State Enable Disable Disable BBU Testing Enable Disable Enable BBU Test Day of the Week 1 7 1 BBU Test Hour 0 23 0 BBU Test Minutes 0 59 0 2 Change System Memory Configuration Change Memory Configuration N Y SP A System Buffer Size in MB 2 MB min 2 SP B System Buffer Size in MB 2 MB min 2 Write Cache Size in MB 0 FE SP A Read Cache Size in MB 0 FE SP B Read Cache Size in MB 0 FE SP A Optimized RAID 3 Memory Size in MB 0 FE SP B Optimized RAID 3 Memory Size in MB 0 6 26 Grid Manager Menus FE FE FE 3 Change Main Menu Password Displays dialogue for changing password 4 Change System Date and Time Displays dialogue for changing date and time 5 Select Storage System Options Package Displays dialogue for changing system type 6 Change Storage System Fast RAID 3 Parameters Displays dialogue for changing FAST RAID 3 cache vault area 3 Change SP Parameters 1 Enable Disable Statistics Logging Displays dialogue for controlling statistics logging 2 Enabl
381. ngineers CEs For more information see chapter 9 Removal and Replacement 8 2 About Field Replaceable Units FRUs Model 10 Deskside Cabinet Power supply units Ei Disk drive slots o o o o M Fan pack a Front view Storage processor modules ai A Rear view Figure 8 1 Model 10 FRU Locations Model 10 Deskside Cabinet 8 3 punoouse qnouj Model 20 Deskside Cabinet Power Storage Supply Fan Disk Drive Processor Units Pack Slots Modules Front View Rear View Midplane Side view of the array with cover removed RIP PCA Figure 8 2 Model 20 Deskside Disk Array 8 4 Model 20 Deskside Cabinet Model 10 Rackmount Chassis Disk drive slots N Front view
382. ning the HP UX 9 04 operating system check for the latest version by typing what hp ux at the system prompt Look for date information similar to the following 9245XA HP UX A 09 04 0 Wed Oct 26 09 44 06 PDT 1994 If the date is equal to Wed Oct 26 09 44 06 PDT 1994 or later a correct revision of HP UX 9 04 is installed If the date is earlier the operating system must be updated to the latest version of HP UX 9 04 before installing the disk array HP UX 9 05 and 9 07 Operating Systems The HP UX 9 05 and 9 07 operating systems must be updated before installing the disk array See Disk Array Support Requirements for HP UX Release 9 05 9 07 in chapter 3 for more information HP UX 10 x and later Operating System The disk array is supported on HP UX 10 01 and greater Note The disk array is not supported as a boot device in either multi initiator or shared bus configurations on HP UX 10 x or later MPE iX Operating System The disk array can be used on HP 3000 Series 900 systems running PowerPatch 1 C 55 01 or later for release 5 5 of the MPE X operating system Contact your WTEC representative for the latest patch information Check the Software 2 3 uonejjeisuj GridMgr iX the interface to the disk array s Grid Manager configuration utility is part of the FOS Fundamental Operating System and is restored automatically onto the system during an install or update as program file gridmgr pub sys For informa
383. nistrator or service provider replaces module SP No Disk array fails System Administrator or service provider replaces SP and the System Administrator restarts operating system Fan failure Yes Applications continue running System Administrator or within fan pack service provider replaces module Power supply Yes If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply and System Administrator restarts operating system if necessary SCSI 2 adapter No I O operations to array disk units fail Authorized service provider replaces adapter and System Administrator restarts operating system and applications SCSI 2 cable No I O operations to array disk units fail Authorized service provider replaces cable and System Administrator restarts operating system and applications Basic Configuration for Model 10 and Model 20 Disk Arrays 3 7 002 seues 0006 dH uoneunDuuo25 Two SP Configuration The two SP configuration shown in Figure 3 2 has one host with one SCSI 2 adapter connected by one external Fast Wide differential SCSI 2 bus to two SPs The two SP configuration provides better performance than the basic configuration For better SP performance with this configuration some physical disk units LUNs can be bound on one SP and the other physical disk units on the second SP The SP that binds a physical disk unit is the default own
384. non destructive The SUSPEND command suspends HARRAY and returns to the DUI level where DUI commands can be run To return to HARRAY type RESUME at the DUI prompt Example The following example shows the user SUSPENDing HARRAY executing the HELP RESUME command at the DUI prompt and then returning to HARRAY at the next DUI prompt HARRAY gt suspend DUI gt help resume SYNTAX RESUME program id BACKGROUND ALL BACKGROUND The RESUME command causes the execution of a previously suspended program to continue If the program which is to be resumed is not specified and the user has only one suspended program that one program will be resumed If the program which is to be resumed is not specified and more than one program is suspended a list of currently suspended programs will be displayed The user should then repeat the command this time giving one or more of the program identifiers from the list ABBREVIATIONS AND ALTERNATE TOKENS RES MODIFIERS program id The unique identifying number assigned to a process by the operating system ALL Resume running all suspended programs Continue Y N y BACKGROUND Place the specified programs in the background if they are not there already and then resume them COMMAND EXAMPLES DUI resume DUI res DUI res 24 26 background DUI res all background DUI gt resume HARRAY has been resumed HARRAY Section 17 External Exerciser 7 43 sonsouDeig pue seni I
385. not exist in the array configuration Troubleshooting the Disk Array 8 25 Buljooysajqnoly Table 8 7 Sense Key Additional Sense Code Additional Sense Code Qualifier Values Not a Complete List cont d Sense key Description 05 25 01 This LUN is a hot spare This LUN can not be accessed for data operations by the host 05 26 02 A parameter was set to an invalid value such as a non changeable field being changed on a SCSI Mode Select 05 27 00 The initiator is attempting to download drive firmware without disabling the peer SP 05 3A 00 The initiator is performing a drive microcode download operation There are no disk drives to which the SP can appropriately install the microcode 0x06 Unit Attention 06 29 00 This LUN has just been powered up or is reporting a front end SCSI Bus Reset condition 06 2F 00 A SCSI Clear Queue message has been received by this LUN from another initiator This status indicates that all of the initiator s requests to this LUN have been aborted by the Clear Queue message 06 80 01 The SCSI Reservation for this LUN has been changed by a Force Reserve operation This Sense Key is reported to all initiators other than the one issuing the Force Reserve command 0x0B Aborted Command 0B 00 00 The SP has aborted a command due to a condition where the initiator is encouraged to retry the command at a later time 0B 00 06 The SP ha
386. not been reported to the system log but an SP failure has been reported by the array monitor the RIP is also suspect Location of RIP PCB LEDs Power LED green LED green flashes when storage processor fault is detected Midplane Sde view of array with cover removed gt RIP PCB Figure 8 13 RIP PCB LEDs Power Block Diagram 8 43 punoouse qnouj Viewing Maintenance Information The View Maintenance Information menu provides fault information about disk array components and accesses the SP s Unsolicited Event Log Select 4 View Maintenance Information from the Grid Manager Main menu to display the following submenus 1 Display Unsolicited Event Log Displays events maintained by the SPs such as Hard Errors start ups and shut downs of disk modules fans SPs power supply units and the BBU More than 2 000 events can be stored in the Unsolicited Event Log A filter feature allows selection of events by device or error message 2 Display Faulted CRUs Displays only currently faulted CRUs Once a fault has been repaired the Unsolicited Event Log entry is removed 3 Display Verify Results Displays the status of Verifies and Sniffs which have been running on a unit To change the Unit Verify features from the Grid Manager Main Menu follow the menu path 5 Change Parameters 1 Change Logical Unit Parameters 6 Change Unit Verify Configuration 4 Display LUN Configuration Su
387. nse data indicating data sourced at the SCSI interface device on the SP encountered an internal parity error 0x03 Device busy Drive returned a SCSI Busy status indicating more than one request was sent to the disk before the first one completed This may be seen most frequently in dual initiator configurations 0x04 Improper SCSI status byte An unrecognized status byte was received during a SCSI transfer possibly due to a SCSI protocol violation 0x05 Media error A hard media error was detected at the disk level and the sector re mapped 0x06 Time out error The SP microcode allocates ten seconds for a disk to complete a command This error is a result of a disk not completing a command within the allocated response time 0x07 Selection time out The targeted drive did not respond to SCSI selection A status of 0x47 indicates retries failed 0x08 Not ready The drive returned a Not Ready SCSI sense key 0x09 Hardware error The drive returned Hardware Error sense key Indicates an internal electro mechanical problem on the disk module Ox0A Bad format Indicates the drive format has changed from the expected sector standard for the array 0x0B Wrong device The drive reporting is a type not supported 0x0C User abort This error code should never be seen 0x0D Illegal request The drive returned an Illegal Request sense key This indicates an illegal command was sent to the disk OxOE Device saw bus reset The dri
388. number SP heartbeat and the fault field SP Identifier The SP identifier located at the top left of screen specifies the SP SP A or SP B connected to the asynchronous RS 232 port being used Presentation Utility Screen 6 33 JoDeue N piis SP SCSI ID Under the SP identifier is the SP SCSI ID which lists the SCSI ID number Serial Number Under the SCSI ID is the serial number for the disk array The number displayed matches the serial number affixed to the back of the disk array Heartbeat To the right of the screen title Presentation Utility is an asterisk in parentheses This asterisk indicates the heartbeat status of the SP It should blink every few seconds to indicate the SP is functioning If it does not appear to blink the SP may not be working In this case try typing CONTROL L or CONTROL Q If the asterisk still does not blink the SP may require service Fault Field If the words SUBSYSTEM FAULT display at the top right of the Presentation screen look for the fault condition in the Grid Manager Unsolicited Event Log From the Grid Manager Main menu follow the path 4 View Maintenance Information 1 Display Event Log System Status The rightmost column shows five fields the microcode revision on the SP statistics logging state whether a second SP is present number of I O requests and disk array cache states Microcode Revision This field gives the revision of the SP s microcode Each SP in
389. o LUNs associated with failed SP by making changes manually in Grid Manager and LVM 4 8 Two SP Configuration Two Adapter Two SP Configuration The two adapter two SP configuration shown in Figure 4 3 has one host with two SCSI 2 adapters each connected by an external Fast Wide differential SCSI 2 bus to a different SP in the array For better performance with this configuration some physical disk units LUNs can be bound on one SP and the rest can be bound on the other SP The SP that binds a physical disk unit is the default owner The route through the SP owning a physical disk unit is the primary route the route through the other SP is the secondary route Host computer Disk array SCSI 2 bus SCSI 2 SPA adapter System RS 232 console pe SCSI 2 SCSI 2 bus adapter SPB RS 232 System disk To MUX port or terminal Figure 4 3 Model 10 20 Two Adapter Two SP Configuration Two Adapter Two SP Configuration 4 9 008 seues 0006 dH uoHeinByu05 Table 4 4 Two Adapter Two SP Configuration Error Recovery Failing component Disk module Continue after failure Yes 9 04 Yes 10 x What happens and how to recover With RAID of any level other than 0 or an individual disk applications continue running System Administrator or service provider replaces module Provides the same features as the basic configuration but w
390. o S700 systems Host computer 1 Disk array SCSI 2 adapter Host computer 2 SCSI 2 adapter System console To MUX port or terminal To MUX port or terminal System console Figure 3 3 Model 10 20 Multi Bus Configuration 3 10 Multi Bus Configuration HP UX 10 x and Later Table 3 4 Multi Bus Configuration Error Recovery Failing component Disk module Continue after failure Yes What happens and how to recover Provides the same features as the basic configuration but with better performance and higher availability With RAID of any level other than 0 or an individual disk applications continue running System Administrator or service provider replaces module SP Yes If one SP fails the System Administrator can access the LUNs owned by the failed SP by accessing the remaining operational SP with the other host If a host adapter fails the System Administrator can access the LUNs in the array with the other host adapter The System Administrator must change the path and SP ownership of the LUNs for the failed SP or failed host adapter by using LVM applies to HP UX 10 x and later and Grid Manager respectively There is no automatic transfer of LUN ownership or path for this configuration type Fan failure within fan pack Yes Applications continue running System Administrator or service provider replace
391. o check for basic power up problems Due to the redundant design features of the array it does not have to be powered down while replacing the disk modules SPs power supplies or BBUs NEVER remove replace or rearrange faulted disks with array power off or data can be lost Table 8 5 Power up Troubleshooting Condition Symptom Power up tests fail Solution Inserting a FRU while the power up tests are running can cause the tests to fail Insert all FRUs and cycle power to rerun the power up tests No FRU LEDs on and fans not running Indicates no AC power is being delivered to the array s power supply s Check input AC power If AC power is present remove array power supply s and check for AC power at the internal AC power supply plug A single power supply on a model 10 may have failed Array fails power up test at letter W and displays message Storage system failure contact your HP service representative The SP620 SCSI SP displays a message that a valid microcode image could not be found on any data base drives Microcode for the SP s is loaded from specific disks within the array If the disk modules were unloaded for installation purposes they may have been reinserted in the wrong slots and SP microcode is not available to be loaded A slot number label is affixed to each disk module at the factory Match the slot number label on the disk module to the appropriate slot in the array
392. o screws and lift the external SCSI connector and cable away from the IPCB bulkhead Remove the two retaining screws securing the IPCB to the chassis Pull both ejectors until they unlock and the IPCB starts out of the chassis Pull the IPCB out of the chassis making sure not to touch any electronic components or circuits CAUTION Follow ESD procedures when removing the IPCB 7 Remove the SCSI 2 cables and pull them from their connectors Figure 9 35 Removing the SCSI 2 Cables 9 62 Model 10 Rackmount Component Removal Internal AC Harness Figure 9 36 and Figure 9 37 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 Remove the following a b g Fan pack Figure 9 6 SPs or fillers Figure 9 9 Power supplies or fillers Figure 9 11 Front panel Figure 9 18 BBU located in the lower front by pressing in on the two tabs at the sides of the unit Figure 9 13 All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array IPCB Figure 9 21 4 Remove the four retaining screws Figure 9 29 WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array 5 Remove the empty chassis from the rack Figure 9 29 6 Remove 11 screws and remove the top panel of the chassis Figure
393. ocessor The microcode has been modified to handle the condition Fixed in microcode revision 8 61 The panic is a result of the SCSI chip and the SCSI driver not agreeing on either the slot or ID of the drive The SP should be replaced 0x00502001 Parity error Firmware revision greater than 8 00 A SCSI controller chip detected a parity error during a transfer from data RAM Reference the 0x00000012 code 0x00503003 SCSI processor error Occasionally the microcode would panic with a 0x00501003 or 0x00503003 code This was being caused by a known problem with the Symbios 53810 SCSI processor The microcode has been modified to handle this issue Fixed in microcode revision 8 61 0x0050300B Parity error Firmware revision greater than 8 00 A SCSI controller chip detected a parity error during a transfer from data RAM Reference the 0x00000012 code 8 82 Unsolicited Event Log Table 8 17 Microcode Panic Codes Not a Complete List cont d Value 0x00600000 Description RAID 5 write operation returns incorrect data This panic could occur in earlier microcode for example revision 8 20 when a disk drive in the LUN returned incorrect data to the SP during a RAID 5 write operation The microcode could create a condition where data written back out to the disk causes more 0x1234560n panics when the sector is accessed again The panic condition was observed to reoccur either immediately on power up of
394. oe Node license for systems specified by the license HP EE E Bc alert shall not be liable for any damages resulting from misuse UXOR NUR FREER or unauthorized use of this program This program RAK RR RoE remains the property of HP eA KkKKKKK KkKKKKK FREES Version A 01 00 A ap oe KERRE E kkkkkxk kkkxkxkxkxkxkxkxkxkxkxkxkxkkxkkxkkxkxkxkkxkxkkkkkxkkkkkkkkkkxkkkkxkkkkkxkkkkkkkkkkkkkxkkkkxkkkkkkkkx k Enter password or a lt CR gt to exit Type HELP for command information Please wait while I scan the device busses Sel Indx Path Bus Controller Type Bus Type 0 10 4 8 HP PB Fast Wide SCSI SCSI 1 10 0 Core Fast Wide SCSI SCSI 10 8 Core Fast Wide SCSI SCSI Legend Sel Select flag to indicate if the bus is marked for possible action Bus is marked Indx Index number used for referencing the specific bus 3 device busses found Mapping each bus will take several minutes Please enter a range of the device busses you are interested in Type for HELP Enter Range default is 0 5 O Indx Path Product ID Bus State Rev 7 48 NIKEARRY H rnpbuo20 1oU PWN EF CO HL c 0 4 8 15 0 HPC2300W 0 4 8 15 1 HPC2300W 0 4 8 15 2 HPC2300W 0 4 8 15 3 HPC2300W 0 4 8 14 0 HPC2300W 0 4 8 14 1 HPC2300W 0 4 8 14 2 HPC2300W 0 4 8 14 3 HPC2300W 0 4 8 15 0 HPC2300W 0 4 8 15 1 HPC2300W 0 4 8 15 2 HPC2300W 0 4 8 15 3 HPC2300W Legend Il Indx Rev where Indx is the index number used for referencing the dev
395. of DISK AO beginning blocks 0 thru 32758 32768 cmd max Verify of DISK AO completed blocks thru 32768 0 errors detected Verify Disks command SUCCESSFUL 1 disks processed 0 errors detected Verify of LUN 0 beginning blocks 0 thru 65535 65535 cmd lt max gt Verify of LUN 0 completed blocks 0 thru 65535 0 errors detected Terify Luns command SUCCESSFUL 1 luns processed errors detected Figure 7 34 Expert Tool Top Level Menu The properties for each menu are documented in Table 7 19 The section following the table contains example screens showing the most common Expert Tool features Table 7 19 Expert Tool Menus and Descriptions Menu items Option Property Description File Save As N A Saves the contents of main STM window to a file Print N A Prints the contents of main STM window Exit N A Exits the Expert Tool 7 112 Support Tools Manager STM Table 7 19 Expert Tool Menus and Descriptions cont d Menu items Option Property Description Logs View Unsolicited Log Event Codes All Displays Unsolicited Event Log information Event Codes gt 0x800 for the SP to which the command is issued Note that the log can be saved to a file and Event Codes gt 0x900 further filtering can then be done using Event Codes gt 0xA00 HP UX commands such as grep Clear Unsolicited Log N A Clears all Unsolicited Even
396. of memory 0x602 Drive Enabled The unit this drive belongs to is ready for normal host I O 0x603 Rebuild Started The background rebuild process has begun rebuilding the data on this drive 0x604 Rebuild Completed The background rebuild process has completed rebuilding the data on this drive 0x606 Unit Shutdown for Trespass The unit of which this drive is a member has been shut down by this SP at the request of the peer SP due to a Trespass operation Recovered Error Log Reporting on p 66 of this chapter has more information 0x607 Unit Shutdown for Change Bind The LUN this module belongs to has been shut down by this SP in order to change its operating parameters This shutdown is needed only when the SP is operating in Target Addressing Mode 0x608 Drive Ready Specified drive is powered up and physically ready for use 0x609 Format Started The given FRU is being formatted 0x610 Format Failed Disk module could not be physically formatted and thus cannot be used in the disk array Make sure the disk module is a valid model If the model is valid consult your HP service representative for recovery service 0x611 SP PROM Rev Notes PROM revision extended status is this SP s SCSI ID This message is logged once at the completion of every powerup 0x613 CRU Equalize Started The SP has started a Rebuild Equalize operation 0x614 CRU Equalize Completed The SP has completed a Rebuild Equalize operation 0x615 CRU Equalize Aborted
397. off the disk array and disconnect all cables The fans continue to run for a few seconds Remove the outer covers Figure 9 18 Remove the following a Fan pack Figure 9 6 b Power supplies or fillers Figure 9 11 c SPs or fillers Figure 9 9 d All disk modules or fillers Figure 9 3 Be sure to label each disk module with the compartment number it was removed from e Interconnect PCB Figure 9 21 f BBU if present Figure 9 13 Turn the disk array over onto its top a Remove the wheel base i Remove the retaining screws from the front of the wheel base ii Slide wheel base 1 inch toward back and remove Remove the inner right metal panel 11 screws Figure 9 19 Remove the inner chassis Lay the chassis onto its left side a Remove the six retaining screws from the top b Lay the chassis onto the covered side c Remove the seven retaining screws from the bottom d Lift the inner chassis up and out of the system i Rotate upward the rear of the inner chassis about 3 inches ii Lift up and out Model 10 Deskside Component Removal 9 39 jueuie2e deH pue jeAoway 7 Unplug the fan cable from the midplane 8 Lift the cable out of the cable retainers 9 Remove the cable from the chassis Note The cable is keyed by a red stripe located along the bottom of the cable at the midplane connector and as the cable is routed through the chassis The red stripe is positioned to the outside of the chassis at th
398. ole Host computer 2 Disk array SCSI 2 adapter To MUX port or terminal Figure 5 4 Multi Bus Configuration Multi Bus Configuration 5 11 006 Salas 000 dH UOHesNByu0D Table 5 5 Multi Bus Configuration Error Recovery Failing component Disk module Continue after failure Yes What happens and how to recover Provides the same features as the basic configuration but with better performance and higher availability With RAID of any level supported by MPE iX applications continue running System Administrator or service provider replaces module SP Yes If one SP fails the System Administrator can access the LUNs owned by the failed SP by accessing the remaining operational SP with the other host If a host adapter fails the System Administrator can access the LUNS in the array with the other host adapter The System Administrator must change the path and SP ownership of the LUNs for the failed SP or failed host adapter by using SYSGEN and ioconfig respectively There is no automatic transfer of LUN ownership or path for this configuration type Fan failure within fan pack Yes Applications continue running System Administrator or service provider replaces module Power supply Yes If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power s
399. olumn of the worksheet When determining RAID modes consider whether or not to apply any high availability features at a higher level through LVM including software mirroring HP UX 10 x and later a For high availability built into the hardware configuration RAID 5 RAID 1 or RAID 1 0 are appropriate choices b Be careful to select an appropriate number of disks for each LUN and that the disks selected have taken into account which internal SCSI 2 bus es are used This is important for true fault tolerance as an internal SCSI 2 bus can be a single point of failure c Fillin the diagram of either the model 10 or model 20 at the top and the RAID column of the worksheet d If mirroring in software choose a RAID level that does not have inherent high availability such as RAID 0 or individual disk units because any other mode would waste capacity If selecting RAID 5 RAID 1 0 or RAID 0 it is necessary to select a stripe size This is the size in disk sectors that is placed on a single spindle before data is continued on the next spindle a Setthe stripe size using Grid Manager or SAM when the LUN is bound Once selected the stripe size cannot be changed without unbinding a LUN and losing all data b Always select a value that is a multiple of 16 sectors In general stripe size should be set as small as possible but large enough to prevent forcing I O operations to access more than one stripe A performance consultant may be
400. on a warning incomplete or failure status refer to the tool Activity Log Access this screen from xstm s main menu by selecting Tools Information gt Activity Log Support Tools Manager STM 7 77 sonsoubeiq pue seninf ISOH amp Information Tool Activity Log for C2300 Array on path 874 8 OF x salt rose hp com 15 43 214 34 Information Tool Activity Log for C2300 Array on path 8 4 8 Log creation time Tue Feb 24 12 41 37 1998 Tue Feb 24 12 41 37 1338 Information tool scsi nike starting on path 8 4 8 Tue Feb 24 12 41 38 1998 Attempt to retrieve the identification information from the device was successful Tue Feb 24 12 41 38 1998 Tool completed with exit status SUCCESSFUL B indicating tool completed without errors Figure 7 7 Activity Log Screen If a tool fails refer to the tool Failure Log by selecting Tools Information Failure Log from xstm s main menu for a description of the failure and recommended course of action if any Firmware Update Tool The Firmware Update Tool is an interactive tool that provides information about and allows updating of disk module firmware and SP microcode Access it by selecting the Tools gt Firmware Update Run menu option from xstm s main menu 7 78 Support Tools Manager STM irmware Update Tool for device type C2300 Array on path 874 8 File List Update Help INSTRUCTIONS Use File commands to
401. on presents an overview of the SCSI disk array s current status This includes fan power supply and BBU status cache usage and microcode revision The SCSI disk array s physical and logical layout comprises cabinet orientation the location of disks and LUN binding and ownership Figure 7 4 shows the map section of the Information Log Support Tools Manager STM 7 71 sonsoubeiq pue seni ISOH amp Information Tool Log for C2300 Array on path 874 8 This SP A 8 7305 SP628 Array Date 82 25 38 Peer SP B 9 Array Time 17 83 88 Serial 95 7535 656 Cabnt Type Rack Mnt Firmware Rev Enabled Microcode Rev 9 55 41 Auto Trespass Enabled Prom Rev Levl 1 73 88 Stat Logging Figure 7 4 Information Log Screen Map Section Status fields surround the physical map Table 7 7 xstm Status Fields for the Information Log Status field values Description This SP ID hard address loop ID and SP model of this SP Peer SP ID hard address and loop ID of peer SP Serial Serial number of the SCSI disk array FAULT DETECTED Displayed only if the SP reports a SCSI disk array fault the Information Tool completion status will be set to Information Warning 7 72 Support Tools Manager STM Table 7 7 xstm Status Fields for the Information Log cont d Status field values Description Array Date SCSI disk array s date in mm dd yy format Ar
402. on the disk drive module takes the 24 VDC from the midplane and converts it to 5 VDC used by the disk drive and other logic in the module e 24 volts model 10 and 48 volts model 20 to the SP module are converted to 5 VDC and 12 VDC by way of the SP power regulator Disk Array Power Troubleshooting All FRUs in the array power line disk drive modules power supply units and fan pack are monitored by the SP If a fault is detected the SP informs the host and shuts down the faulty FRU Information concerning the fault is placed in the host error log recorded in the Grid Manager status and shown by an illuminated LED on the faulted FRU s front panel If the power bus on the midplane is disrupted so the DC voltages are not distributed to some or all of the disk drive modules and or the fan pack the host or array may indicate Power Block Diagram 8 41 punoouse qnouj that an FRU has a fault when it does not The following are suggestions to determine whether a midplane has a power problem e The green POWER LED on the front panel is off while the green READY LED on the power supply unit is illuminated This indicates that the power supply unit has developed the DC voltages but they are not detected on the midplane If the disk drive modules and fan pack power up normally the suspected faulty unit is the front panel LEDs and cable otherwise the following could be the result The cable connecting the power supply units to
403. onnect all cables The fans continue to run for a few seconds Remove the outer covers Figure 9 18 Remove the following a Fan pack Figure 9 6 b Power supplies or fillers Figure 9 11 c SPs or fillers Figure 9 9 d IPCB Figure 9 21 e All disk modules or fillers Figure 9 3 f BBU Figure 9 13 Turn the disk array over onto its top a Remove the wheel base i Remove the retaining screws from the front of the wheel base ii Slide wheel base 1 inch toward back and remove Remove the inner right metal panel 11 screws Figure 9 19 Remove the inner chassis Lay the chassis onto its left side a Remove the six retaining screws from the top b Lay the chassis onto the covered side c Remove the seven retaining screws from the bottom d Lift the inner chassis up and out of the system i Rotate upward the rear of the inner chassis about 3 inches ii Lift up and out 9 46 Model 10 Deskside Component Removal Note The on off switch is attached to the inner chassis 7 Remove the four spade lug wires from the back of the switch 8 Use a flat blade screwdriver to press down on the retaining clips 9 Work the switch back and forth to loosen Retaining clips 2 on top 2 on bottom 4 L Four To wiring Figure 9 27 Removing the On Off Switch Model 10 Deskside Component Removal 9 47 jueuie2e deH pue jeAoway AC Line Filter Figure 9 28 1
404. ons are supported only on the HP 9000 Series 800 1 Determine if the disk array is to be connected to a single host or to two host computers 2 Ifthe disk array is to be connected to a single host go to the next section Connect the SCSI 2 Cables to a Single Host 3 Ifthe disk array is to be connected to two hosts go to Connect the SCSI 2 Cables in Two Host Configurations on p 27 Connect the SCSI 2 Cables to a Single Host 1 Identify which configuration is being set up Basic Dual adapter Dual adapter dual SP For detailed information on these configurations see chapters 3 4 and 5 Have the system administrator shut down the host computer After the shutdown completes power down the system Securely connect the SCSI 2 cable s and terminator to the disk array as shown in Figure 2 8 Connections should be finger tight Note The total length of the SCSI 2 bus must not exceed 25 m 81 ft including all external and internal SCSI 2 bus cables for the host and the disk array Connect the Disk Array s to the Host Computer 2 25 uonejjeisuj SCSI Bus Cable Terminator Plug Thumb Screws Cabling a basic configuration scsb 3 jl C SCSHB O gt al l SCSA C gt Connect to host s SCSI 2 connector Cabling a dual SP configuration o A C SCSHB SCSLA gt gt Connect to host s SCSI 2 adapter Cabl
405. op panel of the chassis Figure 9 30 7 Remove the inner chassis Figure 9 31 9 68 Model 10 Rackmount Component Removal 8 Disconnect all four spade lug wires on the AC line filter PCA 9 Remove three screws and lift the AC line filter PCA away from the mounting studs Four To wiring spanedug harness AC line filter Retaining screws 1 hidden Figure 9 39 Removing the AC Filter Assembly Model 10 Rackmount Component Removal 9 69 u w vejd y pue jeAoway Model 20 Deskside Component Removal Outer Covers Figure 9 40 1 Remove the front panel a Lift from the bottom b Pull up and away 2 Remove the top cover a Remove the two front screws b Slide top 1 inch toward rear c Lift up and away 3 Remove the sides a Lift up approximately 1 inch to release the hubs from the keyhole slots b Remove sides Note For these procedures the disk array must be powered off and unplugged Also it is recommended that ESD procedures be followed Use ESD kit P N 5182 4119 The covers should be removed only when the internal parts of the disk array must be accessed The covers are installed only on the deskside models The rackmount models do not have covers 9 70 Model 20 Deskside Component Removal Figure 9 40 Removing the Outer Covers Top cover Model 20 Deskside Component Removal 9 71 jueuie2e deH pue jeAoway Midplane Figure 9 41 Fi
406. opy of a logical disk image The copy provides continuous access if the original image becomes inaccessible The system and user applications continue running on the good image without interruption There are two kinds of mirroring e Hardware mirroring in which the SP in the disk array synchronizes the two disk images e Software mirroring HP UX 10 x in which the host operating system synchronizes the disk images A hardware mirror can be created by binding disk modules as a RAID 1 mirrored pair or a RAID 1 0 group a mirrored RAID 0 group The hardware mirrors the modules automatically Using software mirroring RAID 0 groups or individual units with no inherent data redundancy can be mirrored With software mirroring the operating system mirrors the images Instructions on software mirroring can be found in the LVM User s Guide 1 24 Planning Disk Unit Configurations RAID 0 Group Non Redundant Array A RAID 0 group uses block striping A RAID 0 group can have the following number of disk modules e Model 10 three to 10 modules e Model 20 three to 16 modules Unlike the other RAID levels the hardware does not maintain parity information on any disk module There is no data redundancy nor any high availability features RAID 0 offers enhanced performance through simultaneous I Os to multiple disk modules A RAID 0 group configuration for a physical disk unit offers fast access rather than high availability With HP UX 10 x and
407. or service provider replaces module Provides the same features as the basic configuration but with better performance and higher availability With RAID of any level other than 0 or an individual disk applications can continue to run after the failure of a single disk module SP Yes If one SP fails ownership of the failed SP s LUNs can transfer automatically to the remaining operational SP If a primary and an alternate path have been configured in LVM LVM switches automatically to the alternate path These operations are transparent to the user Fan pack Yes Applications continue running System Administrator or service provider replaces module Power supply Yes If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply The System Administrator restarts the operating system if necessary SCSI 2 adapter cable I O operations fail to all physical disk units configured on SP attached to a failed SCSI 2 adapter cable The System Administrator can transfer control to the other SP by host s and array configuration changes If the SCSI 2 adapter or SCSI 2 cable fails no I O operations are possible between the host and the array Can regain access to LUNs associated with failed SP host adapter or SCSI 2 cable by making changes manually in Grid Manager and LVM 4 14 Multi Initiator Configuration HP UX 10
408. ormation to tell whether the sector holds data or parity information 0x683 Data Sector Reconstructed The hardware reconstructed a data sector that had a checksum or write stamp error 0x684 Parity Sector Reconstructed The hardware reconstructed a parity sector that had a checksum or write stamp error 0x685 Hard Error The hardware detected an error other than a parity or write stamp error 0x686 Command Complete A command completed after a soft error was corrected 0x687 Stripe Reconstructed Inconsistent write time stamp in a RAID group were corrected 0x688 Command Dropped An optional command was dropped 0x689 Sector Reconstructed On a read from a RAID 1 mirrored pair a corrupted sector was reconstructed 8 60 Unsolicited Event Log Table 8 10 Informational Codes contd Event code Description 0x68A Uncorrectable Parity Sector A hard error on a parity sector could not be corrected 0x68B Uncorrectable Data Sector A hard error on a data sector could not be corrected 0x68C Hard Read Checksum Error A hard checksum error was detected on a data transfer from the host 0x68D Soft Read Checksum Error A soft checksum error was detected on a data transfer from the host Ox68E Inconsistent Stripe Inconsistent write or time stamp were detected in a RAID group Ox68F Inconsistent Time Stamps Inconsistent time stamps were detected in a RAID group on a verify 0x690 Dr
409. orting 0 eee eee eee eens 8 66 Hard Media Error versus Hot Spare 2 e eee ee ee eee eee eee 8 67 Bound but Unassigned Problems 0 eee eee eee tenes 8 68 Decoding Extended Status Codes 0 c cece esee 8 70 Microcode Panic exui LT wee ede eee eee 8 75 Troubleshooting with Grid Manager 0 eee eee eee eee eee 8 86 Checking Status with the Presentation Utility 0 8 86 Bound but Not Assigned Status after Replacing an SP 8 87 Troubleshooting with SAM ssseeeeeeee eme 8 90 View Status of an Array meieriet ponu e n a ene neee 8 90 Disk Module Identification 0 0 0 eee eee eee eee 8 93 Selecting a Disk Module in the Array 0 00 e ee ee eee 8 93 UnbimndabUN i denda dette Rete Perle ee p nie de dde 8 93 Changing a LUN s Cache Parameters 0 000 c cece cence ence eens 8 94 Binda LUN scere nequeam veda ipei ur DEAN ee a 8 96 Set Up TTY Connection a an eee he 8 99 Creating a File System on a Series 700 System 8 101 9 Removal and Replacement xiv Illustrated Parts Breakdown for the Model 10 00 0202 eee eee 9 2 Illustrated Parts Breakdown for the Model 20 0 2 0 ce ee eee ee 9 5 Model 10 20 Disk Module Removal Replacement 2 000 9 8 Model 10 20 Fan Pack Removal Replacement eeeesee eee 9 11 Model 10 20 SP Removal
410. ox contains cables manuals and terminators 4 Lift the shipping carton off the disk array 5 Remove all foam strips from around the disk array 6 Lift the disk array off the pallet WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array CAUTION Do not lift the disk array either by the fan pack handle or the disk module handles because doing so can damage those components Model 10 disk array Place your hands under the bottom of the disk array and lift it off the pallet Lower the disk array onto a suitable surface Model 20 disk array Remove the cardboard braces from the ends of the disk array Place your hands under the bottom of the disk array and lift it off the pallet Lower the disk array onto a suitable surface 7 Remove the protective covering from the disk array Unpack the Disk Array 2 9 uonejeisuj Install a Rackmount Disk Array in a Cabinet Before installing a rackmount disk array into a cabinet refer to Table 2 1 and Table 2 2 to determine the number of disk arrays that fit into the 1 1 m and 1 6 m cabinets When an Uninterruptible Power Supply UPS is added to either cabinet the number of disk arrays that can be installed decreases due to the UPS s EIA units Table 2 1 1 1 m Cabinet Number Current Supported ofarrays Model VAC amps EIA units i
411. p it takes that group s unit number When it returns to hot spare status it regains its original unit number Only LUNs numbered 0 through 7 should be bound because SCSI 2 specifies only three bits for the LUN field This limits the number of LUNs eight and the names 0 through 7 available when binding LUNs Creating Logical Disk Units LUNs 6 47 Jobeue N pus Maximum Rebuild Time The maximum rebuild time specifies the maximum number of hours the disk array can take to integrate a hot spare or new disk module inserted to replace a defective one into the group The four hour default value is recommended Selecting lower rebuild times impacts system performance The rebuild time can be changed with Grid Manager later without damaging the information stored on the LUN However changing the rebuild time during a rebuild does not speed up the rebuild and does not take affect until the current rebuild has completed Sectors per Stripe Element The number of sectors per stripe element specifies the stripe size in 512 byte disk sectors on each disk in the RAID group The stripe element size specified becomes an integral part of the LUN It cannot be changed without unbinding the unit which destroys all data Enable Caches LUN Cache Three options are available for LUN caching e Read and write cache can be enabled e Read or write cache can be enabled e Read and write cache can be disabled It is recommended that both read and wr
412. p Disk Array Caching for Microcode Revision 8 xx 9 xx Specifying Memory Configuration for Microcode Revision 8 xx 9 xx Changing Storage System Parameters Grid Manager Passwords Grid Manager Inventory Screens Jobeue N pus About Grid Manager The Grid Manager utility can be used to e Create logical disk units LUNs e Bind and unbind disks e Determine FRU status e Set up disk array caching Use Grid Manager procedures in the following situations The first time a LUN sometimes referred to by Grid Manager as a physical unit or physical unit number configuration is created or when an existing configuration is changed The operating system cannot access a disk module in the disk array until the module has been bound into a logical disk unit e Ifthe disk array is updated to support caching e Unbinding e fadditional memory is added to existing SPs or replacing an SP when the disk array caching is enabled Note The procedures in this chapter assume the disk array configurations have been planned in advance Chapter 1 has information on planning configurations 6 2 About Grid Manager Configuring MUX Ports on HP 9000 Systems 1 Attach the male end of the 17355D cable to the system MUX box Attach the female end of the 17355D cable to the RS 232 port Console SP A or Console SP B on the back of the disk array for a Series 800 CPU Use the 24542H cable for a Series 700 Only two connections
413. part of another logical disk unit Unbound disks and disks available for binding are identified as UNB Before binding disks read the restrictions and recommendations in Table 6 4 Table 6 4 Restrictions and Recommendations for Binding Disks into LUNs Group to bind Any group Restrictions and recommendations Only unbound UNB disk modules can be bound If an attempt is made to bind a module that is binding BIN Grid Manager displays an error message All disk modules in a group must be of the same type and same capacity to use the modules storage space fully The capacity of a disk module can be determined in Grid Manager by selecting that module in the Presentation screen and pressing RETURN RAID 5 group A minimum of three and a maximum of 16 disk modules must be bound However for more efficient use of disk space it is recommended that five disk modules be bound Note MPE iX has a 16 Gbyte limit for volumes and a 4 Gbyte limit for Ldevs RAID 1 mirrored pair An even number of disk modules from a minimum of two to a maximum of 16 must be bound The disk modules can be selected in any order and each disk module in a pair should be on a separate bus The maximum number of disks is 10 for model 10 and 16 for model 20 RAID 0 group A minimum of three and a maximum of 16 disk modules must be bound Modules can be selected in any order The maximum is 10 for model 10 and 16 for model 20 RAID 0 is no
414. play lvdisplay lvdisplay lvdisplay L 1988 dev vg01 L 2000 dev vg02 L 2000 dev vg02 L 2000 dev vg02 L 1952 dev vg02 L 2000 dev vg03 L 1976 dev vg03 dev vg01 1voll dev vg02 1voll dev vg02 1vo12 dev vg02 1vo13 dev vg02 1vol14 dev vg03 1voll dev vg03 1vo12 Using Logical Volumes For raw I O use the following dev vg01 rivoll dev vg01 rivoll dev vg01 rivo12 dev vg01 rlvo13 dev vg01 rivol4 dev vg01 rivoll dev vg01 rivo12 dev vg01 1voll dev vg01 1voll dev vg01 1vo12 dev vg01 1vo13 dev vg01 1vol14 dev vg01 1voll dev vg01 1vo12 For file system use the following LVM Logical Volume Manager 7 131 sonsoubeig pue seni ISOH EMS Hardware Event Monitoring Hewlett Packard s Event Monitoring System EMS for HP UX is available for use with the disk array An EMS hardware monitor watches the disk array and reports any unusual occurrence called an event Hardware event monitoring is an important tool for implementing high availability Using hardware event monitors you can virtually eliminate undetected hardware failures that interrupt system operation or cause data loss The EMS Hardware Monitors User s Guide and other information about EMS Hardware Monitors is available at http www docs hp com hpux systems Setting Up the Hardware Event Monitor Separate monitors watch over the disks and the disk array You need to install and configure the Disk Monitor disk em
415. ponding status information or to mark the disk for binding Unbind a LUN Use this action to unbind a LUN on the model 10 or model 20 arrays Unbinding a LUN frees the constituent disk modules After a LUN is unbound the disk modules can be re bound into new LUNs Unlike the bind operation unbinding a LUN takes only a few minutes Unbinding a LUN erases all information contained on the LUN The information is not recoverable Troubleshooting with SAM 8 93 punoouse qnouj Example screen output SAM gt Disks and File Systems gt Disk Devices gt Actions Disk Array Maintenance gt Unbind Confirmation If you unbind this LUN you will not be able to recover any information which it may contain Do you want to continue and unbind this LUN Yes No Figure 8 26 Unbind Screen Changing a LUN s Cache Parameters This action applies to model 10 and model 20 arrays and leads to a dialog window containing two toggles one to enable disable write caching the other to enable disable read caching The dialog window has two buttons which are used to enable or disable write and read caching for the selected LUN Note that enabling or disabling either form of caching only affects the selected LUN not any other LUNs that may have been bound in the array 8 94 Troubleshooting with SAM Example screen output SAM gt Disks and File Systems gt Disk Devices Actions Disk Array
416. prompt to re enable the SP620 SCSI SP Creating Logical Disk Units LUNs Preparing to Create Logical Disk Units Before creating LUNs by binding disk modules identify which disk modules are to be used in each LUN To maximize throughput performance and availability disk modules bound into a LUN must be on separate internal buses Binding Disk Modules as Logical Disk Units The binding logical units function is used to bind disk modules into the following configurations RAID 1 0 group not supported on MPE iX RAID 5 group RAID 0 group not supported on the SP620 SCSI SP or MPE iX RAID 1 mirrored pair Individual disk unit not supported on the SP620 SCSI SP or MPE iX Hot spares RAID 3 not supported on HP UX or MPE AX Note RAID 0 and individual disk units are not supported on models A3549 or A3550 because they do not provide data redundancy For microcode revision 8 56 and later these RAID levels can only be configured in FE mode with the Grid Manager utility Creating Logical Disk Units LUNs 6 43 Jobeue N pus Configuration worksheets have been provided in chapter 1 to assist in planning the configuration Before continuing the disk array should be examined to ensure that there are disk modules in each of the compartments module IDs specified on the worksheet A disk can be bound to a new logical disk unit only if it is unbound UNB Grid Manager does not allow a logical disk to be bound if it is already a
417. put Section 10 DIAGNOSTIC TROUBLE TREE PDEV is an C1300WDR5 Sending a Test Unit Ready Command Beginning Read Read Compare tests Performing First Read Performing Second Read Read Read Compare tests completed End of Section 10 DIAGNOSTIC TROUBLE TREE HARRAY Exiting HARRAY has terminated 7 12 HARRAY Section 10 Diagnostic Trouble Tree HARRAY Section 17 External Exerciser The HARRAY Section 17 External Exerciser is an interactive program that provides access to the set of internal diagnostics logs and utilities at both the LUN and FRU level Note HARRAY on HP UX 9 04 does not support LUNs 6 and 7 HARRAY is accessed from the DUI prompt as shown below When the HARRAY prompt appears the exerciser is waiting for a command Examples are shown for every command DUI gt harray sc 17 pd 24 14 0 kk kk kk ee e KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KK KKK KKK KKK Sk kx kk KKK KKK KKKKKKKK KKKKKKKK RK HIGH AVAILABILITY DISK ARRAY DIAGNOSTIC GRO KKKKK KKKKK XOKOK OK C Copyright Hewlett Packard Co 1992 93 KAKEN K k k x All Rights Reserved KERER KKKKK KKKKK This program may only be used by HP support personnel TA MK and those customers with the appropriate Class license or Node license for systems specified by the license KERER HP shall not be liable for any damages resulting from EEEE misuse or unauthorized use of this program This KERAN program rema
418. qnoJj Key Action Dord Dump the contents of the Unsolicited Event Log to the screen The default displays the Log starting with the first entry through the last entry The Esc key will halt display of entries Gorg Goto a particular Unsolicited Event Log entry Display Faulted CRUs To display a list of faulted CRUs detected by the current SP from the Grid Manager Main menu follow the menu path 4 View Maintenance Information 2 Display Faulted CRUs Use screen to determine what component s have failed when the disk array s front panel FAULT LED is lit Although a disk array can have two SPs the disk array has only one FAULT LED As a result this screen must be checked on both SPs to insure that all failed CRUs are identified failed disk modules typically are reported only by the SP that controls the LUN to which the failed disk has been bound Once a fault has been repaired the Unsolicited Event Log entry is removed If there are no entries then the message This SP has no detected faults is displayed Entries show the CRU ID and a description of the fault 8 46 Viewing Maintenance Information m xterm Ioj xi P A SP SCSI ID 15 Current Faults S N 9621954 CRU Fault fal Disk AO Powerup Wait Figure 8 15 Current Faults Screen Display Verify Results To view the status of a unit Verify and Sniff from the Grid Manager Main menu follow the menu path 4 View Maintenance Information 3
419. r s path LUN ownership and path changes must be reassigned by using Grid Manager and SYSGEN ioconfig Multi bus Two hosts Two SCSI 2 adapters one per host Two SCSI 2 buses one per host Two SPs Resembles two basic configurations side by side and provides modest high availability with the SCSI 2 buses not shared between two hosts With MPE IX supported RAID levels applications can continue to run if a disk module fails If one SP fails the System Administrator can access the LUNs owned by the failed SP by accessing the remaining operational SP with the other host If a host adapter fails the System Administrator can access the LUNs in the array with the other host adapter The System Administrator must change the path and the SP ownership of the LUNs for the failed SP or failed host adapter by using SYS GEN ioconfig and Grid Manager respectively 5 4 Disk Array Configurations Basic Configuration for Model 10 and Model 20 Disk Arrays The basic configuration has one host with one SCSI 2 adapter connected by one external Fast Wide differential SCSI 2 bus to the SP in the array as illustrated in Figure 5 1 The system can survive the failure of a disk module within a redundant RAID group or mirror but cannot continue if the SCSI 2 adapter or SCSI 2 cable fails Host computer Disk array SCSI 2 SCSI 2 bus l SPA adapter RS 232 System console To DTC System port or disk termin
420. r See Enabling Disabling the SP620 SCSI Storage Processor in chapter 6 for information Pull both ejectors until they unlock and the SP starts out of the chassis Pull the SP out of the chassis making sure not to touch any electronic components or circuits Place SP on a static free work surface Remove any installed SIMM modules Model 10 20 SP Removal Replacement 3 Installing an SP Figure 9 10 a Disable the enable disable switch on the SP The SP620 SCSI SP does not have an enable disable switch It is disabled using Grid Manager See Enabling Disabling the SP620 SCSI Storage Processor in chapter 6 for information Install any required SIMM modules Locate the slot where the SP is to be installed Ifinstalling an additional SP remove the SP filler module from the slot Slide the SP into the slot using the slot guides until the ejectors engage with chassis edge Press both ejectors until they lock and the SP seats fully into the chassis Enable the enable disable switch on the SP The SP620 SCSI SP does not have a switch Enable cache Model 10 20 SP Removal Replacement 9 15 jueuie2e deH pue jeAoway Before removing set switch to disable gt 5 S o c 8 e 7 7 TION e before oving Figure 9 8 Setting the Enable Disable Switch 9 16 Model 10 20 SP Removal Replacement
421. rameters type YES then select Enable and press Return for CACHE PAGE SIZE 2 Answer the next question CACHE SIZE xx where xx is the array SP 64 in most cases by typing 0 zero This turns off CACHE and disables the CACHE IMAGE of the BOUND DRIVES which is causing the problem 7 Reboot the SPs The easiest way is while still in Change Storage System Parameters to select the Change Storage System Options and select the current value 2 for HP UX 10 x or A for HP UX 9 04 This causes the SPs to reboot automatically 8 Return to the Main menu and select UNBIND Press the space bar to move to the LUN to be removed Select it by pressing RETURN Answer YES when asked if this LUN is to be unbound The LUN should be removed Then remove any other LUNs that may be required at this time 9 Once all the LUNs are unbound go back through the Grid Manager Main menu and re enable the CACHE to what it originally was 10 Reboot the SPs again as described in step 7 above Once all of the above have been done the units are unbound and can be rebound as needed If you do not want to follow this procedure because data on the disk module will be lost insert the disk modules so as to not cause a rebuild if there is an HA bind on the unit Then go to the Presentation screen and look at the BIND SLOT and BIND SIGNATURE for all the involved disk modules This should determine where the disk modules were originally bound and then allow the
422. rating system maintains synchronization Mirroring is further explained in chapter 1 multi bus configuration A storage system configuration that has two hosts each with one SCSI 2 adapter and two external Fast Wide differential SCSI 2 buses and two SPs physical disk A physical disk unit is one or more disk modules each having a head assembly and spindle bound by Grid Manager into a RAID 5 group independent access array RAID 1 mirrored pair RAID 0 group nonredundant array RAID 1 0 group mirrored RAID 0 group individual disk unit or hot spare The operating system sees the physical disk unit which might include more than one disk module as one contiguous span of disk space path See primary disk array path Glossary G 3 Aiesso 5 primary disk array path or primary path In a disk array that has two SPs each SP connects to each physical disk unit LUN through either a primary or a secondary path The primary path belongs to the SP which owns the physical disk unit and the secondary path that belongs to the other SP The SP on which a physical disk unit is bound is the default owner of that physical disk unit Knowing the paths is important because control may need to be transferred to the secondary path in the event of a SP or SCSI 2 adapter failure RAID Redundant Array of Inexpensive Disks RAID 0 Three or more disk modules bound as striped disks the disk array reads and writes file information with more th
423. ray Time SCSI disk array s time in hh mm ss 24 hour format Cabnt Type Cabinet type for example deskside or rackmount Fan A B Fan status for example OK or NR not ready VSC A B C Power supply units status BBU Battery Backup Unit status Cache Amount of physical memory on selected SP Systm Physical memory reserved for SP s use Read Physical memory reserved for reads and status Write Physical memory reserved for writes and status Firmware Rev SP firmware revision Microcode Rev SP microcode revision Prom Rev Levl PROM revision level Auto Trespass Auto trespass setting Stat Logging Statistics logging setting The physical layout contains information about the status of the disks and LUNs Table 7 8 Disk State for the Map in the Information Log Status field values Description blank removed Off off Pow powering up Unb unbound Rdy ready Reb rebuilding Ena enabled Bin binding Fmt formatting Support Tools Manager STM 7 73 sonsoubeiq pue seninf ISOH Table 7 8 Disk State for the Map in the Information Log cont d Status field values Description Eqz equalizing with a hot spare Uns unsupported Table 7 9 LUN Bind State for the Map in the Information Log Status field values Description blank disk is unbound Ind disk is bound to an individual disk LUN R 0 dis
424. re 7 20 Tests Verify Disks Screen 7 92 Support Tools Manager STM dh Epari Poot Youty Dirba be dows CS Annee on path m Bile Ei For Disk AD on Array Barial H j 7535 8558 First Mock to verity D 17289007 at Last Mock tn verify 17763007 iraz None Lasr bieti to warty misst be first block oo verify Figure 7 21 Tests Verify Disks Select Blocks Screen Support Tools Manager STM 7 93 sonsoubeiq pue seni ISOH Tests Verify LUNs 1 Select Tests Verify LUNs from the Expert Tools main menu Select the type of verification from the dialog box that appears Figure 7 22 2 Select the desired action and click Yes The Select Blocks dialogue box appears Figure 7 23 3 Define the appropriate block parameters and click OK When the system completes the verify LUNs process the Expert Tools main window displays the information eo kere he a The command will Verthy vla the SES VeriPy rmerenand Litt dn Arie Barla d H5 7335 B538 Select deirnd sinn C ew Py ALL los ALL bikeckos ae ed in get emagin gt Veri Pg SELELTITI una ALL Blocks Emam yeu will i iyeu wijl i you wijl tyu si 1 dijou ll E Figure 7 22 Tests Verify LUNs Screen 7 94 Support Tools Manager STM ci Doat Wanie Lurr ee Pe kn For Lun T on Arras Saria H51 753 5 85B First Mock to verity 0 Bononia fat Last block tn verity BIBIT V ion Hinrks par ver
425. re 7 39 I O Read Disks Output Screen Support Tools Manager STM 7 117 sonsouBeiq pue saninn ISOH T O Read LUNs The following steps are required to view LUN information 1 Select I O 5 Read LUNs from the Expert Tools main menu Select LUNs from the dialog box that appears as shown in Figure 7 40 2 Select the desired action and click Yes The Select Blocks dialogue box appears Figure 7 41 3 Define the appropriate block parameters and click OK When the system completes the disk read the output window appears with information about the LUNs and blocks specified Figure 7 42 hpterm Op x Expert Tool Read Luns for device C2300 Array on path 8 4 8 This command will Read blocks sectors of data from Accessible Luns on Array Serial 9621954 Select desired action lt IB Read ALL luns ALL blocks as listed in next selection gt Read SELECTED luns ALL blocks Lun_O Blocks thru 68908031 Luni Blocks 0 thru 17227025 Lun 2 Blocks 0 thru 69665279 Lun 3 Blocks 0 thru 16411391 Lun 4 Blocks 0 thru 4073138 Lun 5 Blocks 0 thru 16411391 lt X gt Read SELECTED luns SPECIFIED blocks Lun O Blocks lt you will be asked Lun 1 Blocks lt you will be asked Lun 2 Blocks lt you will be asked Lun_3 Blocks lt you will be asked Lun_4 Blocks lt you will be asked Lun 5 Blocks lt you will be asked gt v more wv k
426. re is no MUX port a VT100 or EM100 capable terminal can be connected directly to the disk array Log into the Series 700 or 800 as gridmgrx where x is the number for a particular login For example Login gridmgr0 At this point the normal login messages display for example copyrights restricted rights legend etc After these messages the prompt Connected displays Refresh the screen by typing CONTROL L where CONTROL is the control key and L is the letter el not the number one The Grid Manager screen displays After finishing with Grid Manager type the tilde character followed by the period and press RETURN to exit The login prompt displays At the login prompt change the termmode of the terminal back to HP 6 10 Initiating Grid Manager on HP UX Initiating Grid Manager on MPE iX GridMgr iX is a utility program that is part of the MPEAX FOS and interfaces to the disk array s Grid Manager configuration tool Users primarily system managers of the MPE iX operating system run GridMgr iX to access the disk array s Grid Manager software through their logged on user terminal GridMgr iX uses the HP 3000 Series 900 s DTC I O subsystem to communicate through an RS 232 port with the Grid Manager program running on the disk array GridMgr iX can be used with a terminal that supports the VT100 asynchronous terminal communications protocol or any terminal supported on the HP 3000 Series 900 system whether or not it has
427. re not to touch any electronic components or circuits CAUTION Follow ESD procedures when removing the IPCB Model 10 Deskside Component Removal 9 35 jueuie2e deH pue jeAoway Fan unit Figure 9 21 Removing the IPCB 9 36 Model 10 Deskside Component Removal LEDs LED Cable Figure 9 22 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Remove the outer covers Figure 9 18 3 Remove the inner metal cover 11 screws Figure 9 19 4 Lift the LED cable out of the cable retainer The LEDs are soldered to the front of the cable 5 Unplug the LED cable from the midplane 6 Push the LED cable forward out of the bracket See arrow 1 in Figure 9 22 Note The green LED is on the top and the white LED is on the bottom on a deskside model On a rackmount model the green LED is on the left and the white LED is on the right 7 Remove the individual LED bezels The bezels are keyed with tabs that fit into grooves in the LED bodies See arrow 2 in Figure 9 22 8 Thread the LED cable back through the sheet metal opening See arrow 3 in Figure 9 22 9 Remove the LEDs LED cable by lifting up and out through the top of the bracket See arrow 4 in Figure 9 22 Model 10 Deskside Component Removal 9 37 jueuie2e deH pue jeAoway Figure 9 22 Removing the LEDs LED Cable 9 38 Model 10 Deskside Component Removal Fan Cable Figure 9 23 1 Power
428. red action and click Yes The Select Blocks dialogue box appears Figure 7 45 3 Definethe appropriate block parameters and click OK When the system completes the verify Disks process the Expert Tools main window displays the information 7 120 Support Tools Manager STM amp hpterm Expert Tool Verify Disks for device C2300 Array on path 874 9 This command will Verify via the SCSI Verify command Disks on Array Serial amp 9621954 Select desired action dp Verify ALL disks ALL blocks as listed in next selection lt gt Verify SELECTED disks ALL blocks thru 17263907 thru 17263907 thru 17263907 thru 17263907 thru 17263907 thru 17263907 Disk R0 Blocks Disk BO Blocks Disk_C0 Blocks Disk DO Blocks Disk EO Blocks Disk 81 Blocks Disk B1 Blocks 0 thru 17263307 Disk C1 Blocks 0 thru 34870016 VU VVV VY VY VVV VVV VVV VVV VVV VVV VVV VVV XN Verify SELECTED disks SPECIFIED blocks Disk_AO Blocks lt you will be asked gt Disk BO Blocks lt you will be asked Disk C Blocks lt you will be asked more wv UY more oooooooo Oe v Y KRKK v more v KOK Figure 7 44 Tests Verify Disks Screen hpterm iof xi Expert Tool Verify Disks for device C2300 Array on path 874 9 For Disk AO on Array Serial 9621954 First block to verify lt 0 17263907 0 Last block to verify 0 17263907 17263307 l Blocks per verify cmd 1
429. removed from Remove the inner metal cover 11 screws Figure 9 41 Remove the midplane Figure 9 55 Remove the retaining screws from the four connectors two screws per connector Remove the harness Model 20 Deskside Component Removal Back of midplane mounting bracket Connector J27 DC power distribution cable Connector Sigua retaining screw Figure 9 46 Removing the DC Power Distribution Cable Model 20 Deskside Component Removal 9 79 jueuie2e deH pue jeAoway Fan Cable Figure 9 47 Note The fan cable can be removed without removing the midplane Remove the two es Ne oo 9 80 screws at the top of the midplane and tilt the midplane back to gain access to the fan cable connector When reinstalling the fan cable make sure there is enough slack in the cable so the fan cable folds properly when the fan pack is closed Unplug the fan cable from the fan pack Figure 9 6 Remove the fan pack Figure 9 6 Release the fan cable from the cable retainer Remove the four retaining screws and remove the distribution panel Remove the fan cable from the two internal cable retainers Disconnect the fan cable from the midplane Remove the fan cable Model 20 Deskside Component Removal Fan cable we Midplane Za Midplane connector Figure 9 47 Removing the Fan Cable Model 20 Deskside Component Removal 9 81 u w v
430. replaced In summary any LUN can report a disk drive failure even though the disk drive is not a part of the LUN HARAYMON on HP UX 9 0x versus HP UX 10 x HARAYMON is implemented as a daemon process in HP UX 9 0x If HARAYMON was installed correctly in usr diag bin it starts running each time the system boots When ps ef is typed at the HP UX shell prompt HARAYMON is included in the list of processes Its process identification is usr diag bin HARYAMON On HP UX 10 x HARAYMON is started 7 10 High Availability Disk Array Monitor HARAYMON by the Sherlock Diagnostic Platform when a disk array I O event is detected HARAYMON is not a daemon process in HP UX 10 x so it is not included in the list of processes that displays when ps ef is typed Except for this difference HARAYMON functionality is identical on HP UX 9 0x and HP UX 10 x systems HARAYMON is available for HP UX 9 04 from the patch tape shipped with the disk array It is available for HP UX 9 05 and 9 07 on patch PHSS 7587 High Availability Disk Array Monitor HARAYMON 7 11 sonsouDeig pue seni ISOH HARRAY Section 10 Diagnostic Trouble Tree Note To run HARRAY type SUPLICEN and enter the password This section runs the default Diagnostic Trouble Tree The algorithm follows Perform an Inquiry to determine if device is supported Perform a Test Unit Ready to clear all Sense data Perform a Read Read Compare test and report results Example Out
431. revisions 8 29 8 61 and 9 xx 0x00A02003 Target mode addressing used with NCR 53C825 SCSI IC Occurs if a command is issued with target mode addressing enabled with the NCR 53C825 SCSI processor used for the host SCSI interface in the array revision A of this IC resolves this problem Microcode revision 8 20 has been modified to not allow target mode addressing to be enabled unless a 53C825 SCSI processor is installed in the array If running on a Windows NT system and target mode addressing is used a bound unit 7 might conflict with the controller ID of 7 Under target mode addressing the LUN stays at 0 and the individual devices targets are addressed Windows NT only allows for eight devices 0 6 for disks and 7 for the initiator SP If this exists the target disk 7 will have to be unbound 0x00E00000 I O from cache and firmware timing issue The microcode would occasionally panic with a OXO0E00000 0x00E0000C code during normal operating conditions The panic was a result of communication problems between the two SPs ina mirrored caching array Fixed in microcode revision 8 61 Executing I O exclusively to and from cache for an extended period of time gt 5 seconds could result in a panic OxOOE00000 with extended status code 0x00EO0000C Fixed in microcode revisions 8 29 to 8 61 Panic 0x00E00000 and 0x00E0000D were caused by a firmware related timing issue under heavy I O activity Fixed in microcode revisions 8 29 8 6
432. ribution cable model 20 rackmount 9 98 disk module 9 8 1 6 Index fan cable model 10 deskside 9 39 fan cable model 10 rackmount 9 59 fan cable model 20 deskside 9 80 fan cable model 20 rackmount 9 100 fan pack 9 11 internal AC harness model 10 deskside 9 43 internal AC harness model 10 rackmount 9 63 internal AC harness model 20 deskside 9 84 internal AC harness model 20 rackmount 9 103 IPCB model 10 deskside 9 35 IPCB model 10 rackmount 9 55 LEDs LED cable model 10 deskside 9 37 LEDs LED cable model 10 rackmount 9 57 LEDs LED cable model 20 deskside 9 76 LEDs LED cable model 20 rackmount 9 96 midplane model 10 deskside 9 31 midplane model 10 rackmount 9 52 midplane model 20 deskside 9 72 midplane model 20 rackmount 9 91 on off switch model 10 deskside 9 46 on off switch model 10 rackmount 9 66 on off switch model 20 deskside 9 87 on off switch model 20 rackmount 9 106 outer covers model 10 deskside 9 29 outer covers model 20 deskside 9 70 power supply 9 19 RIP PCA model 20 deskside 9 75 RIP PCA model 20 rackmount 9 94 SCSI cables model 10 deskside 9 41 SCSI cables model 10 rackmount 9 62 SIMM 9 24 SP 9 14 replaceable parts model 10 R 14 model 20 R 17 replacement BBU 9 21 disk module 9 8 fan pack 9 11 power supply 9 19 SIMM 9 24 SP 9 14 REQUEST SENSE HARRAY command 7 37 requests completed on Presentation screen 6 35 RESET HARRAY command 7 38 RES
433. right C Clariion Storage Systems Inc 1992 1994 Disk array subsystem controller Model 7300 Firmware revision 07 97 ABCDEFGHIJKLMNOPORS TUVWXYZ During the array boot sequence the alphabet displays one letter at a time as shown on the preceding line to indicate that power up tests are running These characters and any periods dots that display on the screen during the boot sequence indicate the generic activity shown in Table 8 3 In the table the term DDRAM refers to data memory on the SP IDRAM refers to instruction memory on the SP The power up sequence tests specific components on the SP including the SIMMs DRAM The sequence also is responsible for loading microcode to the SPs In the event of a power up failure suspect parts are the SP SIMMs or database disk modules Database disk modules store the microcode image that is loaded to the SP s during power up At least one database disk module must be present in a database slot for the power up sequence to complete Always examine all CRUs in the disk array for fault LEDs if the power up tests fail Table 8 3 Disk Array Power up Tests for the SP only Displayed information Model 10 Model 20 Activity No dots No dots Initial Power up PROM Tests 1 dot 1 dot DDRAM ALL Ones Test 2 dots 2 dots DDRAM ALL Zeros Test 3 dots 3 dots DDRAM Even Crosstalk Test 8 10 Power Up Troubleshooting Table 8 3 Disk Array Power up Tests for the SP only
434. rovider replaces module Power supply Yes If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply The System Administrator restarts the operating system if necessary SCSI 2 adapter cable No If the adapter cable fails LUN ownership and path changes may be made using Grid Manager and SYSGEN ioconfig respectively Can regain access to LUNs associated with failed SP host adapter or SCSI 2 cable by making changes manually in Grid Manager and SYSGEN ioconfig 5 10 Two Adapter Two SP Configuration Multi Bus Configuration The multi bus configuration shown in Figure 5 4 has two hosts each with one SCSI 2 adapter connected by one external Fast Wide differential SCSI 2 bus to an SP in the disk array The SP that binds a physical disk unit LUN is the default owner Each host uses its own disks in the disk array independently The multi bus configuration resembles two basic configuration systems side by side This configuration can be used for either a model 10 or a model 20 disk array One host is connected to the array by way of SP A and the other host is connected to SP B The RS 232 cables are connected to their own array terminals In this configuration one disk array can support two systems Host computer 1 SCSI 2 adapter System console To MUX port or terminal System cons
435. rray caching is enabled Not supported and can be configured for microcode revision 8 56 and later only in FE mode 6 50 Creating Logical Disk Units LUNs Unbinding a Logical Disk Unit LUN Unbinding a logical disk unit LUN divides it into its constituent disk modules allowing the user to rebind the modules in any valid configuration To unbind use Grid Manager Unlike binding unbinding takes only a few minutes A hot spare can be unbound by either SP if it is in a RDY H state A LUN can only be unbound by the SP to which it was bound The only exception is when the owning SP has failed the non owning SP can then unbind the LUN that was owned by the other SP CAUTION Unbinding a LUN destroys all information on the disks An unbound module cannot be used until it is bound into a LUN so the operating system can access it All links to data are lost once disks are unbound The data cannot be recovered To Unbind a Logical Disk Unit 1 Ifthe disk array has two SPs make sure to use the Grid Manager of the SP that currently owns the LUN except if it is a global hot spare From the Grid Manager Main menu choose the Unbind Logical Units option After reading the warning message that says all data will be lost press RETURN The grid appears with all disk modules of the first bound LUN highlighted SOR oe Use the arrow keys or space bar to highlight the logical disk unit to unbind For a RAID group Grid Manager high
436. rrored Cache lt gt Enabled X Disabled Spindle Synchronization X Enabled UV more wv eA v more wv w vy more wv ek S Figure 7 50 Utilities System Configuration Host I F Settings Screen 7 126 Support Tools Manager STM SAM System Administration Management Tool For detailed information on using SAM see the following manuals e How HP UX Works Concepts for the System Administrator e System Administrator Tasks e Solving HP UX Problems The following are basic instructions to assist in configuring the disk array with SAM These instructions apply to HP UX 10 0x Series 800 only Starting SAM 1 At the HP UX prompt type SAM 2 Press RETURN The SAM main menu will appear Making a Menu Selection 1 Use the up and down arrow keys or the TAB to move the cursor to the field that needs to be highlighted 2 Press RETURN to highlight the desired field Binding a Logical Disk Unit Using SAM 1 Press the F4 key to gain access to the top bar on the screen Press the arrow keys or the TAB key to highlight the Actions field 2 Press RETURN to access the Actions menu Highlight the Disk Array Maintenance field 3 Usethe up and down arrow keys to move the cursor to the field that needs to be selected Highlight the Bind LUN field SAM System Administration Management Tool 7 127 sonsoubeig pue seni ISOH Use the TAB key to move around in this screen Move the cursor to the Bind T
437. rrored pair RAID 0 group nonredundant array RAID 1 0 group mirrored RAID 0 group individual disk unit or hot spare Before the physical disk unit can store data it must be made available to the operating system Unbinding reverses the bind process changing a physical disk unit into its original disk module s Unbinding destroys all user data on the physical disk unit The Grid Manager utility is used to bind and unbind disk modules bus A physical connection between printed circuit boards in one or more computer chassis used to communicate data See also SCSI 2 bus and multi bus configuration cache See disk array caching disk array caching A feature of a disk array with a BBU and two SPs each with at least 8 Mbytes of memory Each SP stores caches information temporarily in its local memory which can save time when data is written to and or read from a physical disk unit LUN Aiesso 5 disk drive module A self contained disk drive that slides into a slot in the front of the disk array It consists of the carrier assembly which holds the disk mechanism and the regulator board Also called disk module disk module Synonym for disk drive module disk striping See striping disk striping disk unit A short name for physical disk unit LUN dual adapter dual SP configuration A configuration that has one host with two SCSI 2 adapters for the disk array and a storage system with two SPs dual SP configurat
438. s 1 Change Logical Unit Parameters 1 Change Default SP Ownership of Unit Displays dialogue for changing SP ownership 2 Change Unit Rebuild Time Displays dialogue for changing rebuild time 3 Change Unit Auto Assign Displays dialogue for changing auto assign Grid Manager Menus 6 29 Jobeue N pus 4 Change Unit Caching Parameters Displays dialogue for changing unit caching FE 6 Change Unit Verify Configuration Displays dialogue for setting LUN verification test parameters and optionally launches this test Logical unit to configure hex Sniffing state Enable Disable Enabled Sniffing time in units of 100 milliseconds per IO 30 Start a background verify Yes No No Background verify time in hours 0 to 254 0 Clear the verify reports Yes No No 2 Change Storage System Parameters 1 Change Storage System Caching Parameters Displays dialog for current cache parameters Change caching parameters N Y Cache State Enable Disable Disable Cache Page Size in KB 2 4 8 16 64 Cache Size in MB 64 FE BBU Testing Enable Disable Enable BBU Test Day of the Week 1 7 1 BBU Test Hour 0 23 0 BBU Test Minutes 0 59 0 2 Change System Memory Configuration Displays dialog for current memory parameters Change Memory Configuration N Use Automatic Memory Configuration N FE SP A System Buffer Size in MB 4 MB min 4 FE SP B System Buffer Size in MB 4 MB min 4
439. s console cables 2 33 disk array to a D Series CPU 6 5 power cord 2 42 RS 232 port to an HP 3000 DTC 2 38 SCSI 2 cables 2 25 CONTROLLER REP HARRAY command 7 17 controlling display of event records 8 54 creating logical disk units LUNs 6 43 logical volumes 7 131 physical volumes 7 129 volume groups 7 130 CRUs display faulted 8 46 cstm 7 68 D database drives 1 11 6 45 8 10 disk array failure 8 15 date amp time array 7 85 1 114 DC power distribution cable removal 9 78 9 98 default ownership in STM 7 114 default SP in STM 7 85 DESCRIBE HARRAY command 7 18 deskside disk array unpacking 2 6 DIAG HARRAY command 7 19 diagnostics 7 2 dimensions R 8 direct connect terminal configuration 2 34 disk array compatibility R Z components 1 9 description 1 1 1 2 dimensions R 8 features R 2 l 2 Index installing and reconfiguring on S700 systems 3 2 internal SCSI 2 bus lengths 2 21 models available R 5 options available R 5 power consumption R 10 power input specifications R 9 powering up 2 43 replaceable parts R 14 requirements for HP UX 9 05 9 07 3 2 specifications R 8 troubleshooting 8 20 unsolicited event log 8 53 upgrade and add on products R 6 weight R 8 disk array caching 1 11 1 19 enabling disabling 6 54 6 62 microcode rev 7 23 6 53 microcode rev 8 xx 6 59 disk array configurations HP 3000 Series 900 5 3 HP 9000 Series 700 3 5 HP 9000 Series 800 4 2 Disk Failure Information display 8 50 disk firmware
440. s 0x05123454 An address of zero indicates that the SP was unable to locate a hard parity error The four SIMM modules on the SP are referred to as SIMM 1 SIMM 2 SIMM 3 and SIMM 4 in this description Looking at the SP SIMM 1 is closest to the large voltage regulator module with or without a heat sink SIMM 4 is closest to the edge of the board Voltage Regulator SIMM Sockets Figure 8 11 SP SIMMs Troubleshooting the Disk Array 8 31 punoouse qnouj Two steps can be used to identify which of the four SIMMs has failed 1 Ifthe least significant nibble the rightmost digit of the failing address is either 0x0 through 0x3 or 0x8 through OxB the failing SIMM is either SIMM 1 or SIMM 2 If the least significant nibble of the failing address is either 0x4 through 0x7 or 0xC through OxF the failing SIMM is either SIMM 3 or SIMM 4 In the example above an address of 0x05123454 indicates that the failing SIMM is either SIMM 3 or SIMM 4 since the least significant nibble is 0x4 2 The failing SIMM is identified by determining if the failing address is in the high or low half of the total address space The division between the high and low SIMM depends on the total amount of memory installed on the SP The cross over points from low to high for the different memory configurations are as follows 2 Mbytes 0x04100000 4 Mbytes 0x04200000 8 Mbytes 0x04400000 16 Mbytes 0x04800000 32 Mb
441. s A1 E1 amp AS E3 AO EO amp A2 E2 A0 A4 amp BO B4 BEH Hardware Path Volume Logical Device File Group ID Physica Volumes Figure 1 21 Configuration Management Worksheet for HP UX 9 04 Planning Disk Unit Configurations 1 43 uonduioseg jonpold Host Name Array Name Host Adapter Paths Ai E1 amp A3 E3 AO EO amp A2 E2 A0 A4 amp BO B4 LIT In Out In Out SP A SP B Logical Volumes Figure 1 22 Configuration Management Worksheet for HP UX 10 x and Later 1 44 Planning Disk Unit Configurations To complete the configuration management worksheet follow these steps 1 Determine how many host systems will be attached to the disk array Copy and use one Configuration Management Worksheet per array per host system a Fill in the Host Name and Array Name on each sheet b Correlate LUNs that are accessed by more than one host by the second and third subnumbers in the primary and alternate paths It is critical to set up the LVM structure through the disk array s SPs in such a way that continuous controller ownership transfers are avoided Determine which SCSI bus configuration cabling strategy is most advantageous This decision will determine how many SCSI 2 host adapter cards will be required in the system and how the bus es will be structured through the disk array SPs It will also facilitate planning of other fast wide SCSI peripheral devices on the e
442. s aborted a command due to an unrecoverable error includes retries exhausted on the host SCSI bus 0B 2C 00 The initiator has issued a non tagged CDB during a Contingent Allegiance that was not a Request Sense while a TAGGED I O was present on the board for the T L nexus The SP considers this to be an invalid sequence of commands The initiator is encouraged to reissue the command with a valid tag 0B 43 00 The SP has aborted the command because a SCSI message that the SP needed to complete the command was rejected by the initiator 0B 47 00 The SP has aborted the command due to Parity Errors exhausting their retry count 0B 48 00 The SP has received an SCSI Initiator Detected Error message 8 26 Troubleshooting the Disk Array Table 8 7 Sense Key Additional Sense Code Additional Sense Code Qualifier Values Not a Complete List cont d Sense key Description OB AE 00 The LUN has encountered an Incorrect Initiator Connection Examples include receiving two CDBs for the same l T L Q nexus or the mixing of tagged and untagged CDBs for a given T L Q nexus with no outstanding Contingent Allegiance Per the SCSI 2 specification the SP aborts all outstanding CDBs for thel T L nexus SP Troubleshooting The SP is powered by 24 VDC model 10 and 48 VDC model 20 from the midplane and is converted to 5 VDC by the on board voltage regulators The SP has a power and a FAULT LED that is visible by unlocking and openin
443. s module Power supply Yes If a redundant power supply module is present applications continue running otherwise disk array fails Service provider replaces power supply and System Administrator restarts the operating system if necessary SCSI 2 adapter cable No I O operations fail to all physical disk units configured on SP attached to a failed SCSI 2 adapter cable System Administrator can transfer control to the other SP by changing host s and array configurations Can regain access to LUNs associated with failed SP host adapter or SCSI 2 cable by making changes manually in Grid Manager and LVM Multi Bus Configuration HP UX 10 x and Later 3 11 002 seues 0006 dH uoneunDuuo25 3 12 Multi Bus Configuration HP UX 10 x and Later CONFIGURATION HP 9000 SERIES 800 Disk Array Configurations Basic Configuration for Model 10 and Model 20 Disk Arrays Two SP Configuration Two Adapter Two SP Configuration Multi Bus Configuration Multi Initiator Configuration HP UX 10 x and Later Multi Bus Initiator Configuration HP UX 10 x and Later 008 seues 0006 dH uoHeinByu05 Disk Array Configurations Six hardware configurations are available Table 4 1 A section for each configuration explains the error recovery features and includes examples of a disk configuration for a model 10 and a model 20 array The section for each configuration is modular and self
444. s not supported on MPE X HARAYMON can also be configured to send e mail messages to any other Internet address by placing the destination address in the file usr diag bin haraymon dest for HP UX 9 0x or usr sbin diag haraymon dest for HP UX 10 x The files usr diag bin haraymon sig for HP UX 9 0x and usr sbin diag haraymon sig for HP UX 10 x are free form text files that can be used to attach text to any messages sent by HARAYMON For example contact phone numbers and information about the physical location of the system can be included in the sig file These files can be created with any system editor An example of an entry in the haraymon dest file is doeGthesystem ourcompany com Installing HARRAY and HARAYMON Check to see if HARRAY the disk array diagnostic is installed on any versions of the HP UX operating system by typing ioscan fC disk e Choose a PDEV associated with the disk array Look for NIKE Nike or nike to find one e Type SUPLICEN e On the next line type the password Type SYSDIAG e At the DUI prompt type HARRAY PDEV x x x section 17 where x x x is the PDEV number High Availability Disk Array Monitor HARAYMON sonsouDeig pue seni ISOH If HARRAY installed successfully the PDEV is removed from system use and is identified for example PDEV 8 0 0 is a C1300WDR5 On HP UX 9 x versions of the operating system check to see if HARAYMON is running by typing ps ef grep HARAYMON
445. s their Ldev numbers It then displays a prompt for the Ldev to access Enter an Ldev in the Ldev number field GridMgr iX confirms the selection by displaying both the disk array Ldev number and the Ldev number of the logged on user terminal GridMgr iX then displays information indicating how to exit the program Type CONTROL Y or CONTROL K to exit GridMgr iX CONTROL Y is the normal choice for MPE iX systems CONTROL K can be used when CoNrROL Y does not work for example when logged on from an X terminal workstation whose software intercepts CONTROL Y for its own use without passing it through to the operating system GridMgr iX pauses at the informational display until RETURN is pressed Pressing RETURN causes GridMgr iX to display the disk array s GridMgr Main menu as shown in Figure 6 5 Grid Manager can now be used as described in the rest of this chapter Initiating Grid Manager on MPEAX 6 13 Jobeue N pus e xterm GridHgr Hain Henu 1 Presentation Utilitu 2 Bind Logical Units 3 Unbind Logical Units 4 View Haintenance Information 5 Change Parameters 6 View Cache Statistics Enter or lt number gt for HELP Enter Choice ll Figure 6 5 Grid Manager Main Menu Exiting GridMgr ix To exit GridMgr iX from within Grid Manager type either CONTROL Y or CONTROL K at any Grid Manager Enter Choice prompts When GridMgr iX exits it returns control to the MPE X CI Command Interpreter GridMgr iX Error Con
446. save this new terminal profile but do not exit this Screen To Configure the RS 232 Ports 1 Press the PRIOR SCREEN function key twice to return to the Host Configuration screen 2 Complete the information on the Host Configuration screen if it is not already filled in and press the SAVE FUNCTION key 3 Pressthe Go To DTC function key to go to the DTC Configuration Selection screen 6 8 Configuring RS 232 Ports on HP 3000 Series 900 Systems 10 Go to the DTC Configuration screen for the DTC to which the disk array s serial ports are attached by pressing the Go To function key that matches the DTC type For example press the Go To 72MX function key to go to the DTC 72MX Configuration screen If this DTC type has not already been configured fill in the fields on the DTC Configuration screen then press the SAvE DATA key Move the cursor to the To configure a card field Enter a card number and press CONFIG CARD Fill in the number of the card on the DTC to be configured and press the CONFIG CARD function key to go to the DTC Card Configuration screen Fill in the fields on the DTC Card Configuration screen to configure the disk array s GridMgr iX RS 232 ports For each disk array RS 232 port that is connected to this card ofthis DTC type the corresponding Ldev number in the Ldev field and the profile name GRIDMGR in the Profile Name field When all of this card s GridMgr iX RS 232 ports have been configured press the SAVE D
447. screen 6 34 systerm type array 7 85 7 114 T terminal emulation through a MUX using Grid Manager 2 35 Test Unit Ready 7 84 7 90 7 118 7 119 time amp date array 7 85 7 114 troubleshooting disk array power 8 41 disk module 8 20 general information 8 18 power up 8 10 RIP PCB 8 42 SP 8 27 SP SIMMs 8 31 SP6xx SCSI SP SIMMs 8 32 with Grid Manager 8 86 with SAM 8 90 TTY connection 8 99 two adapter two SP configuration HP 3000 Series 900 5 9 HP 9000 Series 800 4 9 two host configurations 2 27 two SP configuration HP 3000 Series 900 5 7 HP 9000 Series 700 3 8 HP 9000 Series 800 4 7 U UDP on Presentation screen unassigned dirty pages 6 40 UNBIND Nikearry command 7 62 unbinding a LUN in STM 7 85 7 114 a LUN using SAM 8 93 a physical disk unit using SAM 7 128 LUNs 6 51 physical disk units 6 51 unpacking the disk array deskside 2 6 rackmount 2 8 unsolicited event log 8 53 add a filter 8 55 clearing 7 84 7 113 delete all filters of one type 8 56 displaying 8 44 displaying event records 8 54 extended status codes 8 70 filtering event records 8 54 microcode panic 8 75 navigating 8 45 8 54 rebuild aborts 8 65 SP error codes 8 56 viewing 7 84 7 113 UNSOLLOG Nikearry command 7 63 updating disk module firmware 7 82 7 111 SP microcode 7 82 7 110 updating storage processor microcode 7 20 upgrade and add on products R 6 upgrading from an SP to a SP6xx SCSI SP 8 35 using logical volumes
448. scsi nike starting on path 8 4 95 Wed Apr 29 10 59 50 1998 Attempt to retrieve the identification information from the device was successful Wed Apr 29 10 59 50 1998 Tool completed with exit status SUCCESSFUL lt 0 indicating tool completed without errors tty dee swese me 1 meme 1 Figure 7 31 Activity Log Screen If a tool fails refer to the tool Failure Log by selecting Tools Information Failure Log from xstm s main menu for a description of the failure and recommended course of action if any Firmware Update Tool The Firmware Update Tool is an interactive tool that provides information about and allows updating of disk module firmware and SP microcode Access it by selecting the Tools gt Firmware Update Run menu option from xstm s main menu Support Tools Manager STM 7 107 sonsouBeiq pue seni ISOH amp hpterm _ lolx Firmware Update Tool for device type C2300 Array on path 8 4 3 File List Update Help INSTRUCTIONS 1 Use File commands to select firmware file to be used 2 Use List commands to display current information 3 Use Update commands to start update amp controller disk a Controller updates should have peer SP if any enabled I b Disk updates must have peer SP if any disabled Figure 7 32 Firmware Update Tool Top Level Menu Table 7 18 Firmware Update Tool Menu Items and Descriptions
449. sense code additional sense code qualifier values the SP microcode returns The values are hexadecimal For example the code 02 04 01 consists of leftmost digits Sense key 02 in the example middle digits Additional sense code 04 in the example rightmost digits Additional sense code qualifier 01 in the example Troubleshooting the Disk Array 8 23 punoouse qnouj Obtaining Additional Disk Failure Information with Grid Manager If a disk has failed the Unsolicited Event Log within Grid Manager can display specific failure information CRU EVENT CODE and EXTENDED STATUS The failed disk module can be identified by slot number under the CRU column Refer to the Storage Processor Error Code and extended status code tables in Unsolicited Event Log on p 53 of this chapter Hot Spare versus Rebuild If a hot spare is configured in the array it is used for a rebuild as soon as any disk module fails within a RAID 1 RAID 1 0 or RAID 5 LUN When the rebuild completes the hot spare becomes part of the RAID 1 RAID 1 0 or RAID 5 LUN and assumes the LUN number of the LUN with the failed module The hot spare is globally available to all LUNs When the failed disk is replaced a disk to disk copy begins between the original hot spare and the new replacement disk When the disk to disk copy completes the hot spare is returned to its original state and becomes available again with its own original LUN number While the rebuild is occurring on the
450. shooting the Disk Array Identifying Logical Disk Units LUNs and Disk Module Failures e Every bound disk is assigned a LUN number by the array The Presentation Utility can be used to display LUN numbers Every bound disk displays a STATE for example ENA and its physical unit number If a disk module in a LUN fails its FAULT LED may not illuminate The SP may place the disk into an OFF state or the state may be a blank and only the physical unit number is displayed For example the failed disk in slot A0 in Figure 8 7 is part of LUN 0 and has been turned off by the SP LUN 0 comprises two disks AO and A1 The failed disk in slot C1 in Figure 8 8 is part of LUN 1 and has no state displayed it is blank LUN 1 comprises three disks Al B1 and C1 SPA TT x SP SCSI ID 0 Presentation Utility PROM Revision Microcode Revision 7 597 07 23 Statistics Logging ENABLED A OFF ENA BIN BIN BIN ENA BIN 0 OK B A d 2 3 OK Peer SP REMOVED Requests Completed 6K Cache State DISABLED Figure 8 7 Model 10 Mapping Microcode 7 23 Troubleshooting the Disk Array 8 21 punoouse qnouj SPA SP SCSI ID 0 Presentation Utility PROM Revision 7 97 FAN A OK FAN B OK 0 ENA ENA ENA ENA ENA VSC A OK 9 g 9 J p VSC B OK VSC C OK 1 BIN BIN T 1 0 0 0 Figure 8 8 Model 20 Mapping Microcode 7 23 SPA 5 T SP SCSI ID
451. sing Grid Manager and LVM respectively HP UX 10 x If the adapter cable fails LUN ownership transfers and LVM switches to the alternate path automatically and transparently Can regain access to LUNs associated with failed SP host adapter or SCSI 2 cable by making changes manually in Grid Manager and LVM 4 10 Two Adapter Two SP Configuration Multi Bus Configuration The multi bus configuration shown in Figure 4 4 has two hosts each with a SCSI 2 adapter connected by an external Fast Wide differential SCSI 2 bus to an SP in the disk array The SP that binds a physical disk unit LUN is the default owner Each host uses its own disks in the disk array independently The multi bus configuration resembles two basic configuration systems side by side This configuration can be used for sites requiring high availability because either host can continue after failure of any disk module within a disk array or mirror This configuration can be used for either a model 10 or a model 20 disk array One host is connected to the array by way of SP A and the other host is connected to SP B The RS 232 cables are connected to their own array terminals In this configuration one disk array can support two systems Host computer 1 Disk array SCSI 2 adapter Host computer 2 SCSI 2 adapter System console To MUX port or terminal System console To MUX port or ter
452. sk Array Caching for Microcode Revision 7 23 6 53 Disabling and Enabling Disk Array Caching 0205 6 54 Specifying Cache Page Size and Cache Size 00 02 e eee eee 6 56 Setting up Disk Array Caching for Microcode Revision 8 xx 9 xx 6 59 Specifying Memory Configuration for Microcode Revision 8 xx 9 xx 6 60 Disabling and Enabling Disk Array Caching 0205 6 62 Changing Storage System Parameters 00 eee ee eee ee ee eee 6 65 Changing the Storage Systems Option Package 06 6 65 Checking Auto Trespass and Option Package 4 6 67 Ioscan Shows Two LUN 0s for Disk Array on HP UX 9 0x 6 67 Default Host Interface Options for Microcode Revision 8 20 6 68 Default Host Interface Options for Microcode Revision 9 xx 6 69 Grid Manager Passwords 00 cee cece eee eee een e een enee 6 71 Adding a Passwords visis mu ee avi ae see ee CEMUS 6 71 Changing a Password 0 cee eect een eee eee neeees 6 72 Grid Manager Inventory Screens 0 cee eee eee eens 6 73 7 Host Utilities and Diagnostics U9 U09 High Availability Disk Array Monitor HARAYMON 7 5 Installing HARRAY and HARAYMON 02 0 0 eee eee eee ee eee 7 5 HARAYMON Messages 0 cece eect mh 7 7 Disk Module Failure Message 0 00 e
453. sk modules in the array are functioning normally none is being rebuilt A hot spare has not been configured in slots AO Al A2 A3 or A4 for the model 10 and in slots AO BO CO DO or EO for the model 20 Even though disk array caching is enabled the disk array does not use caching with any logical disk unit unless the write and or read cache for that unit has been enabled The cache for a LUN can be enabled when the unit is bound or at any later time with the Change Logical Unit Parameters menu 6 64 Specifying Memory Configuration for Microcode Revision 8 xx 9 xx Changing Storage System Parameters Significant problems will occur if the SP parameters are changed in the Host Interface Options area The host interface options must never be changed individually The options are properly set by selecting either Storage System Option Package A or 2 The default setting for the first item in the host interface options Addressing model is LUN value Problems occur if this value is changed to TARGET When the Addressing model is set to TARGET the SCSI address switches are bypassed The result is that the address switches have no effect on what is seen by the host system or EM100 terminal connection For example if the Fast Wide SCSI 2 host adapter HP 28696A is at hardware address 0 4 and the SCSI address is set to 13 and 14 and LUNs are bound at 0 1 2 and 3 and TARGET model gets set ioscan and autoconfig see the addresses as 0 4 0
454. sk on MPE iX RAID levels 3 0 1 0 and independent mode are not supported on MPE ix RAID levels 5 1 0 1 0 and independent mode are supported on models A3231A A32324A A33884A A3389A A3539A and A3540A 1 18 Disk Array High Availability Features RAID levels 5 1 0 and 1 are supported on the models A3549A and A3550A RAID level 0 and independent mode are not supported on the models A3549A and A3550A Disk Array Caching Disk array caching which is also called storage system caching is available for disk arrays that have Two SPs each with at least 8 Mbytes of memory mirrored on each SP for example 8 Mbytes 2x4 Mbytes on each SP or a total of 16 Mbytes 4x8 Mbytes A BBU Disk modules in slots A0 A1 A2 A3 and A4 the cache vault drives for the model 10 or in slots AO BO CO DO and EO for the model 20 Note SIMMs for cache are available in 4 and 16 Mbyte sizes and must be ordered in pars because the add on option is for a single SIMM Orders must be based on the information contained in the tables in the SIMM Removal Replacement section of chapter 9 Two SPs SIMMs a BBU and disk modules in the slots listed above must be present to enable storage system cache The disk modules listed are cache vaults for the disk arrays During normal operation when disk array caching is enabled each SP temporarily stores the most recently accessed data in its cache memory on the SIMMs To ensure data integrity e
455. snate esis Oooh eret eee eta d br atta i33 7 39 RO MT RANDOM MEDIA TEST 0 2 02 e eee eh 7 40 RUL RETRIEVE UNSOLICITED LOGS 20 00 02 e eee eee 7 41 SHOW MEGHS xung ea og taney Ake ova ie ws m PUR ees 7 42 SUSPEND 2 scp ei hives bbe Means See eee Oe aan Sead eee 7 43 WEI MT reet uriemnabumAenktiemqane eucetimevue5mnx ces 7 44 NIKEARRY 4 I eee ee a duet Bas erg EL ee 7 46 Detailed Functional Specifications 0 cece ee eee eee 7 46 NIKEARRY Operational Commands 0c eee eee eee eens 7 47 Quick Start Instruction 0 0 0 0 000 cc eee n eee 7 48 BIND ie aon ted ord Shi tid cols eu ta th at st cud 7 50 DISPMAP enr vx i eo ai i a dd dace eee as 7 52 DOWNLOAD 2 Seah ae Go be Lhe hd ae Lae ee ean 7 53 FRU Mapping 2 m eee 2 ee e 4e d eek eee dee 7 55 PRUINFO ie ak ee eas tui ER Xu ee ee Ee 7 56 HELB eie otc uM Le ae Laas Rt SY aha ust 7 57 INQUIRY iie thet Get ew ou LIRE ERIATEWEDE EARS Eris 7 57 MODESENS Eien ioe ales esu wx ene be ee eer pow CES 7 58 RESET iba onal ee A aired gute ae ed Acted Scie 7 61 UNBIND Fleer Dile beet we E ta ihe ator kes tl een dena ee 7 62 UNSOLEDOG zd deed LS EM LIU E Eu CM d 7 63 UPDATE iz eEnebDexns etereue ge Oa ae eed 7 64 UTILINEO 2 topi RU E Res UN Re euet ES 7 67 Support Tools Manager STM esses eA 7 68 xstm Interface c ee ek ee eed vbt Lie a eA eas 7 69 Information Toolsi eei cei ccc ccc eee ene eens 7 69 Firmware Update Tool
456. st and last entries of the Unsolicited Event Log respectively The requested entries are displayed one line at a time in a format that allows a screen capture program to capture them in a text file The file can be viewed later with a text editor Pressing the Esc key during the dump terminates the operation Gorg Go to a particular Unsolicited Event Log entry When typed the user is asked for an event number a full log typically contains more than 1000 events The line number must be within the current range of events in the log or an error message is displayed The Unsolicited Event Log is displayed starting at the specified event number Filtering Event Records A filter allows events to be selected by FRU disk module ID or error code For example filter Al describes events about disk module A1 only a filter of 903 describes events from 8 54 Unsolicited Event Log error 0x903 only Disk modules A0 through E4 fans FAN A or FAN B power supplies VSC A VSC B or VSC C and SPs SP A or SP B can be selected As part of a filter an asterisk can be used as a wildcard to represent one character For example A describes events about the A bus only and 3 describes events whose codes end in 3 codes are three hexadecimal digits To add a filter 1 Atthe Event Log prompt press A The Apply Filters menu appears Apply Filters 1 Apply FRU Filter 2 Apply Error Code Filter Enter for HELP to R
457. stems 0 eee eee ee eee 6 3 Connecting the Disk Array to a D Series CPU 0200 0005 6 5 EISA 8 Port MUX Card J2482A 0 0 cece eee eens 6 5 EISA 64 Port MUX Card J2488A 0 eee eee 6 6 Configuring RS 232 Ports on HP 3000 Series 900 Systems 6 7 Initiating Grid Manager on HP UX 00 eee eee eee ee 6 10 Initiating Grid Manager on MPEAX 0 eee eee eee ee eee 6 11 U9 U09 Exiting GridMgrAX csset crede eae beats e oEPbduierbeles Hoa tad 6 14 GridMgr iX Error Conditions and Messages eee eeee 6 14 Displaying the Grid Manager Main Menu esee eese 6 17 Grid Manager Menus 3 ssc cae Ve ER REIR EUR M UE UIS 6 19 Grid Manager Menus for Microcode Revision 7 23 000 00 6 19 Grid Manager Menus for Microcode Revision 8 20 0000 eee 6 22 Grid Manager Menus for Microcode Revision 8 56 6 25 Grid Manager Menus for Microcode Revision 9 Xx ussu 6 28 Presentation Utility Screen lleleeeeeeeeeeeeeee ree 6 33 SP Stabus oai xu teen hid pau ESO ONE NOE SIRE RN UR Ed 6 33 SP Tdentitier c csi heed eh eet E eh 6 33 SP SCSDID ss cesena esai XR iig ucc ad ee ee b Ed 6 34 Serial Number x2 pues eee paw et ana awakes datos 6 34 Heartbeat eR Rer RE ERE I Det PIETRA 6 34 Fault Eield uuum db chile a pXrentUe eC UGUEUEUE 6 34 System Status e ure rx ve
458. t IBM only 0x804 Single Bit Error The SP s tolerance level for single bit errors in the write cache has been exceeded 0x805 Single Bit Error The SP s tolerance level for single bit errors in the read cache has been exceeded 0x820 Soft Media Error The SP successfully Notify the user after three cleared a media defect errors month FRU 0x840 Disk Sector Invalidated The LUN that Event code Description 0x901 Hard SCSI error An abnormal SCSI bus or disk drive event was detected and could not be cleared through retry operations 0x903 Fan Removed A fan pack has been shut down or removed from the disk array 0x904 VSC Removed A VSC power supply has been shut down or removed from the disk array 0x905 Chassis Overtemperature A chassis overtemperature condition has been detected The controller attempts corrective action 8 62 Unsolicited Event Log Table 8 12 Error Codes cont d Event code 0x906 Description Unit Shutdown The Unit of which this drive is a member has been shutdown and can no longer be accessed by the host See Recovered Error Log Reporting on p 66 of this chapter for more information 0x907 Controller Firmware SP microcode Panic A fatal firmware error has occurred as a result the program running in the SP has reset SP The SP was restarted and continued normally 0x908 Hardware Fault Cache disabling Some hardware condition has caused the SP to disable the
459. t Log information for the SP to which the command is issued Info Capacity N A Displays capacity information for LUNs and disk modules on the SCSI disk array See Figure 7 35 Inquiry N A Displays inquiry information for LUNs and disk modules on the SCSI disk array See Figure 7 36 VO Read Disks N A Reads and optionally displays data from specified disk modules on the SCSI disk array See Figure 7 37 Read LUNs N A Reads and optionally displays data from specified LUNs on the SCSI disk array See Figure 7 40 Tests Test Unit Ready N A Performs unit ready testing of LUNs and disk modules See Figure 7 43 Verify Disks N A Performs verification of accessible disk modules See Figure 7 44 Verify LUNs N A Performs verification of accessible LUNs See Figure 7 46 Support Tools Manager STM 7 113 sonsoubeig pue seni ISOH Table 7 19 Expert Tool Menus and Descriptions cont d Menu items Utilities Option Property Description Bind LUN N A Binds selected disk modules into a LUN with a specified RAID level Unbind LUNs N A Unbinds a LUN LUN Configuration Auto assign Enables or disables the auto assign state of a LUN Cache Parms Enables or disables write and read caching for a LUN Default Ownership Sets the default SP for a specified LUN Rebuild Time Sets the maximum amount of time the SCSI disk array should take to rebuild a failed disk module on a LUN System Configur
460. t Number A1 FRU NUMBER 12 FRU Slot Number A2 FRU NUMBER 18 FRU Slot Number A3 Enter FRU NUMBER for testing or 99 to Cancel gt 18 MECH UTILITY 4 23 PM A Read Only Media Test will be performed on FRU 18 Types of Media Tests 1 Selected Area 2 Random Area Enter Test Type 1 2 2 gt 2 Transfer Length Types 1 Random Transfer Length 7 34 HARRAY Section 17 External Exerciser 2 User Transfer Length Enter Transfer Type 1 2 1 2 Enter Blocks per Transfer 1 127 127 3 Enter Loop Count 1 256 1 gt 1 MECH Read Only Media Test in progress please wait MECH Loop Count 1 Completed MECH Read Only Media Test Successful Example 7 MECH command with option 5 HARRAY gt mech The following commands are available at the mech level 0 Test Unit Ready TUR 1 Inquiry 2 Read Capacity 3 Read 4 Read Media Test RO MT 5 Verify via scsi commands 99 Exit Enter number 0 5 or 99 to Cancel 4 gt 5 The following mechs FRU s are available for test FRU NUMBER 0 FRU Slot Number A0 FRU NUMBER 6 FRU Slot Number A1 FRU NUMBER 12 FRU Slot Number A2 FRU NUMBER 18 FRU Slot Number A3 Enter FRU NUMBER for testing or 99 to Cancel 6 MECH UTILITY 4 23 PM A mech VERIFY will be performed on FRU 6 Enter Starting Block Address 0 4121402 0 100 Enter Ending Block Address 100 4121402 4121402 1000 Enter blocks to verify per scsi cmd 1 901 0 to abort 901
461. t Removal 9 55 jueuie2e deH pue jeAoway Fan unit Figure 9 32 Removing the IPCB 9 56 Model 10 Rackmount Component Removal LEDs LED Cable Figure 9 33 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Access the rear of the disk array 3 Remove the following a b Fan pack Figure 9 6 SPs or fillers Figure 9 9 Power supplies or fillers Figure 9 11 Front panel Figure 9 18 BBU located in the lower front by pressing in on the two tabs at the sides of the unit Figure 9 13 All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array 4 Remove the four retaining screws that hold the chassis in the rack Figure 9 29 WARNING The disk array is very heavy Use extreme care when moving or lifting it At least two persons are needed to lift a disk array 5 Remove the empty chassis from the rack Figure 9 29 6 Remove 11 screws and remove the top panel of the chassis Figure 9 30 7 Lift the LED cable out of the cable retainer The LEDs are soldered to the front of the cable Model 10 Rackmount Component Removal 9 57 jueuie2e daH pue jeAoway 8 Unplug the LED cable from the midplane 9 Push the LED cable forward out of the bracket See arrow 1 in Figure 9 33 Note The green LED is on the top and the white LED is on the bottom on a deskside model On a rackmount model the gr
462. t lines should resemble the following For HP UX 9 04 Class LU H W Path disk 22 24 0 0 disk 23 22 0 1 disk 24 24 0 2 disk 25 24 0 3 disk 12 24 1 0 disk 26 24 1 1 disk 27 24 1 2 For HP UX 10 x Class T H W Path disk 3 8 6 0 disk 5 16 3 0 disk 15 16 14 0 disk 16 16 14 1 disk 17 16 14 2 disk 9 52 5 0 disk 10 52 6 0 Driver scsi3 target disc3 scsi3 target disc3 scsi3 target disc3 scsi3 target disc3 scsi3 target disc3 scsi3 target disc3 scsi3 target disc3 Driver S W State disc3 CLAIMED disc3 CLAIMED disc3 CLAIMED disc3 CLAIMED disc3 CLAIMED disc3 CLAIMED disc3 CLAIMED If the entries are not correct check the following 2 40 H W Status S W Status ok 0x20000202 ok ok 0x0 ok ok 0x0 ok ok 0x0 ok ok 0x20000202 ok ok 0x650202 ok ok 0x650202 ok H W Type Description DEVICE HP C2490AM DEVICE HP C2430D DEVICE DGC C2300WDR5 DEVICE DGC C2300WDR5 DEVICE DGC C2300WDR5 DEVICE HP C2474S DEVICE HP C2473S All hardware connections are tight and no pins are bent There are no addressing conflicts on the bus two targets with the same address or a target at the same address as the host Connect Asynchronous Console Cables If the connections are good but the appropriate device files are not listed install all device files by changing to the dev directory and issuing the insf e command Connect Asynchronous Console
463. t supported because it does not provide data redundancy and can be configured for microcode revision 8 56 and later in FE mode only RAID 1 0 group A minimum of four disk modules and an even number of modules must be bound Grid Manager pairs modules into mirrored images in the order that they are selected The first and second modules selected become a pair of mirrored images the third and fourth modules selected become another mirrored pair and so on The first module selected in each pair is the primary image the second module is the secondary image 6 44 Creating Logical Disk Units LUNs Table 6 4 Restrictions and Recommendations for Binding Disks into LUNs cont d Group to bind Individual disk unit Restrictions and recommendations Functions as a stand alone disk Individual disk units are not supported because they do not provide data redundancy and can be configured for microcode revision 8 56 and later in FE mode only Global hot spare Used in RAID 5 RAID 1 and RAID 1 0 The capacity of a disk module bound as a hot spare must be at least as great as the capacity of the largest disk module that it might replace On a model 10 the disk modules in slots AO A1 A3 or BO cannot be bound as a hot spare because they are database drives On a model 20 the disk modules in slots AO BO CO or A3 cannot be bound as a hot spare because they are the database drives that contain the SP microcode In
464. t the desired action and click Yes The Select Blocks dialogue box appears Figure 7 38 3 Definethe appropriate block parameters and click OK When the system completes the disk read the output window appears with information about the disks and blocks specified Figure 7 39 hpterm l x Expert Tool Read Disks for device C2300 Array on path 874 9 This command will Read blocks sectors of data from Disks on Array Serial amp 9621954 Select desired action lt Read ALL disks ALL blocks as listed in next selection lt gt Read SELECTED disks ALL blocks thru 17263907 thru 17263907 thru 17263907 thru 17263907 thru 17263907 thru 17263907 Disk_AO Blocks Disk BO0 Blocks Disk CO Blocks Disk DO Blocks Disk EO Blocks Disk Rd Blocks Disk B1 Blocks thru 17263907 Disk C1 Blocks 0 thru 34870016 VU AAAA AAAA AAAA AAAA AAAA AAAA AAAA AAAA XX Read SELECTED disks SPECIFIED blocks Disk R0 Blocks lt you will be asked Disk BO Blocks lt you will be asked Disk C Blocks lt you will be asked gt KKK v more v KKK v more v eK v more v KKK Figure 7 37 I O gt Read Disks Screen oooooooo 7 116 Support Tools Manager STM amp hpterm P x Expert Tool Read Disks for device C2300 Array on path 8 4 9 For Disk AQ on Array Serial 9621954 First block to read lt 0 17263907 J Last block to read 0 17263907 17263907
465. ta format is as specified in SCSI 2 Other supported features 16 bit wide data transfer Synchronous data transfer Tagged Command Queuing This device responds to a RESET condition with a SOFT RESET alternative Vendor Identification DGC Product Identification C2300WDUNB NIKEARRY 7 57 sonsouDeig pue seni ISOH Product Revision Level HP01 Vendor Specific Information Addr Hexadecimal 36 0x24 0x53 0x41 0x55 0x4e 0x41 0x20 0x43 Ox4f 44 Ox2c Ox4e 0x54 0x52 Ox4f Ox4c Ox4c 0x45 0x52 CONTROLLER 52 0x34 0x20 0x20 0x20 0x20 MODESENSE The MODESENSE command reads the mode sense pages from the device and displays them to the console MODESENSE takes two inputs as parameters a LUN index and a mode sense page If no parameters are given the user is prompted for the LUN index and mode sense page The prompt for the mode sense page includes a list of supported page codes When MODESENSE is issued with one parameter it is taken as the mode sense page and the LUN index is taken from the TESTUNIT environment variable The mode sense parameter is interpreted as a hexadecimal value Example 1 NIKEARRY gt modesense Enter test array cr 0 5 0 Enter Page Code 0x26 Controller Report Page 0x27 LUN Report Page 0x29 FRU Configuration Page 0x2A System Configuration Page Enter selection cr page 0x26 The System Controller Page contains Cabinet ID C1100 or C1300 Rack Controller ID Controller B Pro
466. tained had the disk module not failed User data O Parity data L 512 639 1024 1151 1536 1663 First Module Meas 640 767 1152 1279 2048 2175 Second Module Or iO Els 7 Oss 768 895 1664 1791 2176 2303 Third Module Fourth Module Fifth Module Q Figure 1 17 Distribution of User and Parity Data in a RAID 5 Group For highest availability for a RAID 5 group in a model 20 disk array the disk modules making up the group must be on different internal SCSI 2 buses Because a model 10 disk array has only two internal SCSI 2 buses it does not matter which buses the disk modules are on Figure 1 17 illustrates user and parity data with the default stripe element size of 128 sectors 65 536 bytes in a five module RAID 5 group The stripe size includes all stripe Planning Disk Unit Configurations 1 31 uonduioseg jonpold elements The disk block addresses in the stripe proceed sequentially from the first module to the second third and fourth then back to the first and so on The disk array performs more steps when writing data to a RAID 5 group than to any other disk configuration For each write operation to a RAID 5 group the disk array must perform the following functions e Read data from the sectors being written and parity data for those sectors e Recalculate the parity data e Write the new user and parity data If a disk module fails performance is degraded until the defective module is replaced an
467. talled 0x633 Fan Speed Increased Disk array has detected an increase in a fan s speed perhaps because temperature rose or another fan failed 0x634 Fan Speed Decreased The fan speed has returned to normal 0x635 Inconsistent Check Information The SP has detected an inconsistency in the sector based check codes 0x637 BBU Recharging The Battery Backup Unit in this disk array is recharging 0x638 BBU Enabled The Battery Backup Unit in this array has been enabled for use 0x639 System Cache Enabled Storage system cache has become ready 0x641 Rebuild Aborted A background rebuild operation has aborted prior to completing the entire disk See Rebuild Aborts on p 65 of this chapter for more information 0x643 SP Initializing Peer second SP was detected in the chassis 0x644 Peer SP Inserted This SP has been informed by its peer SP that the peer has completed its power up initialization sequence 8 58 Unsolicited Event Log Table 8 10 Informational Codes cont d Event code Description 0x645 CRU Bound Extended status is unit number This message is logged once for each CRU in a LUN at the completion of a Bind operation 0x646 CRU Unbound Extended status is unit number This message is logged once for each CRU in a LUN at the completion of an Unbind 0x647 Fuse Bad Extended status is fuse number This message is logged once at the completion of each powerup for any bad fuse dete
468. te LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP on model 20 only 4 Gbyte LP 9 Gbyte LP and 18 Gbyte LP disk support for the SP620 SCSI SP only HARRAY gt DOWN PDEV 24 14 0 DOWNLOAD UTI Types of 1 Downl 2 Downl 3 Downl LOAD is an C1300WDR5 LITY 4 29 PM Download oad Controller Firmware Only oad Controller Firmware AND Update Controller Prom Code oad Disk Firmware HARRAY Section 17 External Exerciser 7 21 sonsouDeig pue seni ISOH Enter Download Type Enter Firmware Filename t 3 Download operation started Working Restarting firmware please wait 25 Percent Complete 50 Percent Complete 75 Percent Complete 100 Percent Complete Download operation SUCCESSFUL Download type 2 7 22 HARRAY gt down PDEV 24 14 0 is an C1300WDR5 DOWNLOAD UTILITY 4 34 Types of Download 1 Download Controller Firmware Only PM or 99 to exit hit CR to abort download 1 Son gt firmware controller s 2 Download Controller Firmware AND Update Controller Prom Code 3 Download Disk Firmware Enter Download Type Enter Firmware Filename 1 3 Download operation started Working Restarting firmware 10 20 30 40 50 60 70 80 90 100 Percent Complete Percent Percent Percent Percent Percent Percent Percent Percent Percent please wait Complete Complete Complete Complete Complete Complete Complete Complete Complet
469. ter connected by one external Fast Wide differential SCSI 2 bus to two SPs The two SP configuration provides better performance than the basic configuration For better SP performance with this configuration some logical disk units LUNs can be bound on one SP the rest can be bound on the second SP The SP that binds a logical disk unit is the default owner of that unit The route through the SP that owns a logical disk unit is the primary route the route through the other SP is the secondary route System console Host computer SCSI 2 adapter Disk array SCSI 2 bus SPA RS 232 m SPB RS 232 B To DTC port or terminal Figure 5 2 Model 10 20 Two SP Configuration Two SP Configuration 5 7 006 Seles 000 dH uoneunbyuo Table 5 3 Two SP Configuration Error Recovery Failing component Disk module Continue after failure Yes What happens and how to recover With RAID of any level supported by MPE ix applications continue running System Administrator or service provider replaces module Provides the same features as the basic configuration but with better performance and higher availability With RAID of any supported level applications can continue to run after the failure of a single disk module SP No If one SP fails the System Administrator must reassign ownership of the failed SP s LUNs to the
470. the SCSI ID for SP B Each switch pack has four switches IDO ID1 ID2 and ID3 as shown in Figure 2 11 Using Table 2 5 as a guide set the SCSI ID switches for the configuration being installed Each SCSI ID should be set to a different value Note Addresses should be set according to SCSI 2 bus priority host adapter to 7 other devices to 6 5 etc If adding a second disk array make sure there are no address conflicts on the SCSI 2 bus The address switches are negative true logic as shown in Table 2 5 Set SCSI IDs for SPs in the Disk Array s 2 29 uonejjeisuj SCSIB C a M SCSIA 5 SCSIBO QUIE op op m WOU UUW elo li e ll opi SCSIA Tower Not used ID O ID 1 SPB ID 2 SCSI ID ID 3 ID O ID 1 SPA ID2 SCSI ID ID 3 ON Figure 2 10 SP SCSI ID Model 10 2 30 Set SCSI IDs for SPs in the Disk Array s a a 03 onn n nmnn szs aw mw RSA T0134 T car e NN co Cw N co aaaa eaaeaan d E co M I lt x ceo i xi Figure 2 11 SP SCSI ID Model 20 Table 2 5 SCSI ID Switch Settings Switch number SCSI ID number ID O ID 1 ID 2 ID 3 0 ON ON ON ON 1 OFF ON ON ON 2 ON OFF ON ON 3 OFF OFF ON ON 4 ON ON OFF ON 5 OFF ON OFF ON Set SCSI IDs for SPs in the Disk Array s 2 31 uonejelsu
471. the display orientation switch on the back of the disk array to the Rack position The switch is the rightmost switch in the set of 10 SCSI ID switches located on the back of the disk array Refer to Figure 2 10 for the location of these switches The rightmost switch should be in the up position for rackmount installation 2 16 Install a Rackmount Disk Array in a Cabinet O N o b o oot M N N z Rail O amp e oy o UJ onto rails A Slide disk array Keeper tab Install clip nuts of disk array D Fasten back of disk array Figure 2 4 Mounting the Model 10 Install a Rackmount Disk Array in a Cabinet uonejjeisuj A Install clip nuts o ris 4 o C 20holes S d o z 7 z A i holes uc sos Rail o Do777 Tene J lew uo i B Slide disk array onto rails Keeper tab E C Fasten front of disk array w D Fasten back of disk array Figure 2 5 Mounting the Model 20 2 18 Install a Rackmount Disk Array in a Cabinet Install Additional SCSI 2 Host Adapter Cards on the Host Depending on the configuration one or more Fast Wide differential SCSI 2 host adapter cards may be required in the host computer system The following information will help determine how to connect the disk array to the computer Use the HP Fast Wide SCSI 2 adapter card product n
472. the failing SP or at some point after power up as the SP conducts a Background Verify on the LUN This could also happen on a rebuild 0x00800077 SCSI download or PROM flashed Occurred infrequently during a microcode SCSI download or during the PROM code flash operation Microcode 9 x5 fixes all known occurrences of this panic If this panic occurs with the latest release of microcode obtain the Unsolicited Event Logs in FE mode from both SPs and the memory dump and send them to your HP service representative for failure analysis 0x0080007E Incorrect state reported by peer The peer SP in the disk array declared that this SP has been removed and must reinitialize before continuing 0x00800082 Firmware revision mismatch The peer SP in the disk array advised that this SP is running the wrong revision of firmware and must reinitialize before continuing 0x00800084 Controller not configured This SP should not be present in the current configuration This occurs if a disk array has been set up for single board CACHing and a second SP is started 0x00A00006 Retry stamp panic On very large no disconnect transfers gt 512 blocks the microcode would panic Fixed in microcode revision 8 20 0x00A00016 SCSI errors SCSI errors on the host SCSI bus might cause the microcode to panic with a 0xX00A00016 or 0x00A00047 code Fixed in microcode revision 8 20 0x00A0001C Register bit reset incorrectly
473. the first part of the request has not yet been processed Another instance can be induced by write only I O while the Battery Test was active This has only been seen on RAID 3 LUNs Fixed in microcode revision 9 xx If this panic occurs with the latest revision of microcode obtain the Unsolicited Event Logs in FE mode from both SPs and the memory dump disk and send them to your HP service representative 0x00220060 Miscellaneous panic The microcode might panic with a 0x00220060 under excessive trespass activity coupled with heavy I O traffic Fixed in microcode revisions 8 29 8 61 and 9 xx 8 80 Unsolicited Event Log Table 8 17 Microcode Panic Codes Not a Complete List cont d Value 0x00220064 Description Firmware resources exhausted Under extremely heavy I O conditions with one or more rebuilds or equalizers in process the microcode would intermittently panic Usually this was a result of one or more firmware resources being exhausted because the array was managing a heavy workload The following causes of the 0x00220064 panics have also been fixed Binding more than 12 LUNs concurrently or binding during extremely heavy I O The microcode would issue a large number of resets due to a hung backend bus while I O was still active on the remaining backend buses causing timeouts to occur The microcode has been made more robust in dealing with the shutdown and release of LUNs in support of un
474. the midplane is faulty The power supply distribution cable is an FRU and should be checked The midplane has failed e fadisk drive module even after replacement consistently indicates a fault in the same compartment the problem is the midplane RIP PCB Troubleshooting Model 20 Disk Array Only The primary function of the RIP PCB is to monitor the dual SP modules in a high availability array and report to the host when one of the SP modules has a fault The RIP PCB has no function when a single or dual host is connected to only one SP module in the array The RIP PCB monitors the ground connections of both of the SP modules on the midplane If the RIP PCB does not detect these grounds it sends a reset to the host that causes the host to check the configuration of the array The RIP PCB contains two green LEDs that are not visible during normal operation of the equipment When the upper LED is on it indicates the RIP PCB has power The lower LED is on when the RIP PCB detects the SP module grounds It flashes when it detects a change The RIP PCB is invisible to the array if it fails the array does not report an error to the system log when a SP fails Check for the SP failure message that has been sent to e mail by the array monitor The array monitor message has a time stamp Check the system log to see if an error has been reported at the same time as the array monitor message 8 42 Power Block Diagram If an error has
475. the system Power down the host computer system Remove the slot cover hardware from an available slot Slide in and seat the new card Tighten the card s screws until secure Connect cables from the disk array to the host system with power off Power up the disk array Power up and reboot the HP UX system eode Goog cuo 9 cw Ne a Run ioscan f to verify successful installation For more installation details refer to the manual that was shipped with the card 2 20 Install Additional SCSI 2 Host Adapter Cards on the Host Determine Disk Array SCSI 2 Bus Lengths The total length of the SCSI 2 bus including all internal and external SCSI 2 bus cables for all host s and all disk array s to be connected in the configuration must not exceed 25 m 81 ft If the total length of the SCSI 2 bus in the configuration exceeds 25 m 81 ft shorten the cable lengths between the host s and the disk array or between the disk arrays The total bus length includes the external and internal portions of the SCSI 2 bus The internal synchronous SE SCSI 2 bus lengths for the model 10 and 20 are shown in the following table Table 2 3 Internal SCSI 2 Bus Lengths for Model 10 and Model 20 Model SP installed Internal bus length Model 10 SP A only 6 m 2 ft Model 10 SP B only 6 m 2 ft Model 10 SP A and SP B 1 2 m 4 ft Model 20 SP A only 2 3 m 7 7 ft Model 20 SP B only 1 6 m 5 3 ft Model 20 S
476. this chapter for more information 0x921 Host Vault Load Failed The SP encountered multiple errors while attempting to load the cache image from disk This message may indicate multiple drive failures Unsolicited Event Log 8 63 punoouse qnouj Table 8 12 Error Codes cont d Event code 0x922 Description Host Vault Load Inconsistent The SP found inconsistencies in the card image on the disk This may indicate a failure or abort of the cache dump 0x923 Host Vault Load Failed Bitmap Ok The SP successfully read the control portion of the cache image on disk but found the data portion to be incomplete This means that a failure or abort occurred during the cache dump Probably any LUN with write cached pages will be inaccessible and must be rebound To identify such a LUN read message Ox90A in this table 0x924 Host Vault Disks Scrambled The SP found the vault disks containing the cache image to be in different order than when the cache image was dumped to disk This means the disks were swapped at power down The disks must be restored to their original order before the SP can load the cache image Table 8 13 Fatal Error Codes Event code OxA02 Description Failed SCSI bus An internal SCSI bus has failed The CRU number displayed corresponds to the bus number AO means bus A BO means bus B and so on The failure resulted from a bad cable or cable connection bad terminator ba
477. tion about GridMgr iX see Initiating Grid Manager on MPE iX in chapter 6 Fast Wide SCSI 2 Host Adapter Firmware If one or more HP28696A Fast Wide differential SCSI 2 host adapters on HP 9000 Series 800 or HP 3000 Series 900 systems are being used check that the correct level of firmware is installed on the adapter s If necessary update to the correct firmware version To determine if the SCSI 2 adapter firmware must be updated on HP UX systems 1 2 4 Run ioscan to determine how many SCSI 2 host adapters are installed in the system and their hardware paths by typing ioscan d scsi3 Example output Hardware Path Description Status 8 scsi ok Check that the correct level of firmware is installed on the Fast Wide SCSI host adapter At the SYSDIAG DUI prompt type fwscsipb pdev x section 3 where x is the hardware path s shown by ioscan Active EEPROM Bank lt Look for this information Code revision 0x3621 lt Look for this information The code revision must be greater than 0x3621 Note that throughout this manual numbers prefixed by 0x are hexadecimal numbers Firmware revisions 0x3621 or less do not support the high availability features of the disk array If the revision is not greater than 0x3621 new firmware must be downloaded to each Fast Wide differential SCSI 2 host adapter in order to support the high availability features of the disk array Download the firmware if required by typin
478. ts high availability by using redundant disk resources in disk array and disk mirror configurations Chapters 3 4 and 5 have more information on these configurations A RAID 5 group maintains parity data that lets the disk group survive one disk module failure without losing data In addition in a model 20 disk array the group can survive a single internal SE SCSI 2 bus failure if each disk module in the group was bound on a separate internal SE SCSI 2 bus After the failed disk module is replaced or the bus fault is cleared the SP rebuilds the disk group from the information stored on the remaining disk modules The rebuilt disk group contains an exact replica of the information it would have contained had the disk module or internal SCSI 2 bus never failed A RAID 1 mirrored pair or RAID 1 0 group which uses RAID 1 technology duplicates data on the disks configured in the group If one disk module fails the other module provides continuing access to stored information Similarly a RAID 1 mirrored pair or RAID 1 0 group can survive a single internal SCSI 2 bus failure if each disk module is bound on a separate internal SCSI 2 bus After the failed disk module is replaced or the bus fault is cleared the SP automatically rebuilds the mirror RAID levels 5 1 and 1 0 data redundancy are supported on HP UX RAID level 3 is not supported on HP UX RAID level 5 is supported only as a data disk on MPE iX RAID level 1 is supported as a boot di
479. ue seninf ISOH Support Tools Manager STM Beginning with HP UX 10 10 the Support Tools Manager STM host based utility is available for the High Availability SCSI disk array The tools available for the disk array under STM are the Information Tool supported on the Series 800 only the Expert Tool and the Firmware Update Tool STM has an extensive online help facility as well as man pages STM has three interfaces xstm a X Windows screen based interface mstm a menu based interface and cstm a command based interface The Information Tool provides configuration information for the SCSI disk array s controllers and disk modules The Expert Tool provides LUN and system configuration testing and Unsolicited Event Log access The Firmware Update Tool provides capability to update a SCSI disk array s controller and disk firmware and is functionally similar to the HARRAY diagnostic s DOWNLOAD command The STM tools are a replacement for HARRAY and provide similar functionality STM is available on HP UX 10 10 and later versions STM can be accessed by three interfaces as shown in Table 7 5 Choose the appropriate interface according to the resources available xstm is recommended because it is the easiest to use Table 7 5 STM Interfaces Interface name Intended environment Invoking command xstm X Windows terminals and workstations usr sbin xstm mstm Menu based for non graphics terminals usr sbin mstm
480. ull the BBU out of the chassis CAUTION Be careful when removing the BBU as it is heavy 2 Installing a BBU Figure 9 13 a b Locate the slot where the BBU is to be installed Remove the failed BBU or filler module from the slot Slide the BBU into the slot until it engages the backplane connector Press firmly until the BBU seats fully into the chassis and the latches lock Model 10 20 BBU Removal Replacement 9 21 jueuie2e deH pue jeAoway M 10 deskside and rackmount front view Model 20 deskside and rackmount rear view Shown with the fan packs removed Figure 9 12 Location of BBUs in the Model 10 and Model 20 9 22 Model 10 20 BBU Removal Replacement 2 eLa Thumb holes Figure 9 13 Removing or Installing a BBU bg LILDILIETCIEIEIETCIETCTCILI DICT I TJ I Jj Model 10 20 BBU Removal Replacement 9 23 jueuie2e deH pue jeAoway Model 10 20 SIMM Removal Replacement SIMMs 4 Mbyte and 16 Mbyte capacity per SI
481. ulti initiator configuration has two hosts each with two SCSI 2 adapters two external SCSI 2 buses and two SPs This configuration provides the highest availability and best array performance for dual host configurations because it provides protection against SP SCSI 2 adapter and external SCSI 2 bus cable failures Chapters 3 4 and 5 explain the different configurations in detail including which configurations are supported on the different systems platforms and differences between operating systems For example if one SP in a two adapter two SP configuration on HP UX 10 x fails ownership of the failed SP s LUNs can transfer automatically to the other SP On HP UX 9 04 the transfer cannot take place automatically Instead the system administrator must reassign ownership of the failed SP s LUNs to the operational SP by using Grid Manager Disk Array High Availability Features 1 21 uonduioseg jonpold Auto Trespass Auto trespass allows an SP that does not own a given LUN to gain ownership by accessing the LUN through the other SP Before auto trespass can occur two separate paths a primary path through one SP and an alternate path through the other SP to each LUN must be configured by using LVM If the primary alternate paths have been configured appropriately with LVM controller failures external SCSI 2 bus failures or SCSI 2 host adapter failures do not require immediate manual intervention Normal I O and data processing
482. umber 28696A to connect to the following systems e HP 9000 8x78 e HP 9000 Series 800 F G H I e HP 9000 Model 800 Ex5 If the computer is an HP 9000 K100 K200 or K400 the disk array can be connected in the following three ways e To the F W Differential SCSI 2 connector on the Core I O card GSC HP UX 10 01 only e To the F W Differential SCSI 2 product number 28696A adapter HP UX 10 0 or 10 01 e To F W SCSI 2 card product number 2969A into one of the slots on the HSC Expansion I O card HP UX 10 01 only If the computer is an HP 9000 7xx the disk array can be connected in the following three ways e For an HP 9000 Series 735 or 755 to the F W differential SCSI 2 connector on the Core I O card within the host system e Foran HP 9000 Series 725 100 to the host system using an A4107A F W differential GSC host adapter e For an HP 9000 Series J200 to the F W differential SCSI 2 connector on the Core I O card or to the A4107A F W differential GSC host adapter Install Additional SCSI 2 Host Adapter Cards on the Host 2 19 uonejjeisuj All the SCSI 2 connectors on the Core I O card must either be connected to a device adapter or terminated using a SCSI terminator HP part number A1658 62024 Only service personnel authorized to open the computer and replace parts should install or replace SCSI 2 adapters in the host computer To install a new host adapter card in the HP UX host system Alert all users and shut down
483. ume groups associated with the LUNs whose ownership is to be changed vgchange a n dev vg0n 4 Using LVM remove the information about all the volume groups associated with these LUNs from the system tables vgexport m mapfile dev vg0n 5 Using Grid Manager change the ownership of the LUNs to be moved to the ownership of the replacement SP by choosing the following menu path 5 Change Parameters 1 Change Physical Unit Parameters 1 Change Default Ownership of LUN 6 Power cycle the disk array 7 Rebootthe host system Troubleshooting the Disk Array 8 29 punoouse qnouj 8 Recreate the directories and nodes for the volume group s associated with the path through the failed SP mkdir dev vg0n mknod dev vg0n group c 64 0x0n0000 vgimport m mapfile dev vg0n dev dsk cnd0s2 vgchange a y dev vg0n mount dev vgO0n lvoll mount directory SP Heat Sinks Two different types of power regulators are used on the SPs Due to the difference in the power regulators some SPs have large black heat sinks other SPs have a copper plate heat sink Both are electrically identical and interchangeable Table 8 8 SP Troubleshooting Condition Symptom No LEDs illuminated on SP Solution Is the enable disable switch enabled The SP620 SCSI SP does not have an enable disable switch Both LEDs will be off if the switch is not enabled SP READY LED is illuminated but there is no response from SP
484. upply and System Administrator restarts the operating system if necessary SCSI 2 adapter cable 5 12 No Multi Bus Configuration I O operations fail to all physical disk units configured on SP attached to a failed SCSI 2 adapter cable System Administrator can transfer control to the other SP by changing host s and array configurations Configuring Disk Array SCSI LUNs for MPE iX To configure the disk array for use with the MPE iX operating system follow the example below for each SP and each LUN two or more disk modules owned by a configured SP For example a disk array containing three LUNs owned by a single SP connected to an HP PB Fast Wide SCSI 2 adapter would be configured according to the following example Configure any upper level hardware for example bus converter paths that are required for the model HP 3000 Series 900 being used before configuring the HP PB Fast Wide SCSI 2 adapter For example a 969KS 100 has two levels of bus converters they appear in the path prior to specifying the HP PB slot number as 10 16 They should be configured in SYSGEN as follows io lp 10 PATH 10 ID A2372 60003 PMGR BUS_CONV_MGR LMGR io lp 10 16 PATH 10 16 ID A2372 60003 PMGR BUS CONV MGR LMGR LDEV TYPE BC PMGRPRI MAXIOS LDEV TYPE BC PMGRPRI MAXIOS 2 0 Configure the HP PB Fast Wide SCSI 2 adapter using an ID of HP286964A unless it
485. upply units Figure 9 13 e All disk modules Figure 9 3 Be sure to label the disk modules as they are removed from the disk array Remove the front panel Figure 9 40 Remove the four retaining screws that hold the chassis to the rack Figure 9 52 WARNING The disk array is very heavy Use extreme care when moving or Cogo NS lifting it At least two persons are needed to lift a disk array Remove the empty chassis from the rack Figure 9 52 Remove the top metal panel 11 screws Figure 9 53 Remove the distribution panel Remove the internal AC harness bracket Remove four retaining screws and lift the internal AC bracket away from the chassis Model 20 Rackmount Component Removal 9 103 jueuie2e deaH pue jeAowaY 10 Remove the green wire from the grounding lug behind the on off switch 11 Disconnect the blue and brown wires from the on off switch 12 Pull the cable out 13 Remove the three AC plugs from the internal AC harness bracket Note The plugs must be oriented as shown in Figure 9 62 when they are installed in the bracket If they are not oriented correctly the power supplies cannot be plugged in When reinserting the cable tape the two spade lug wires and the green neutral wire together This makes insertion easier Do not pinch the SCSI 2 cables when installing the distribution panel Bracket retaining screws AC harness Figure 9 61 Removing the AC Harness Bracket 9
486. ure 9 41 4 Lift the LED cable out of the cable retainer The LEDs are soldered to the front of the cable 5 Unplug the LED cable from the midplane 6 Push the LED cable forward out of the bracket See arrow 1 in Figure 9 45 Note The green LED is on the top and the white LED is on the bottom on a deskside model On a rackmount model the green LED is on the left and the white LED in on the right 7 Remove the individual LED bezels The bezels are keyed with tabs that fit into grooves in the LED bodies See arrow 2 in Figure 9 45 8 Thread the LED cable back through the sheet metal opening See arrow 3 in Figure 9 45 9 Remove the LEDs LED cable by lifting up and out through the top of the bracket See arrow 4 in Figure 9 45 9 76 Model 20 Deskside Component Removal LED Bracket LED Cable Connector Midplane Figure 9 45 Removing the LEDs LED Cable Connector Model 20 Deskside Component Removal 9 77 jueuie2e deH pue jeAoway DC Power Distribution Cable Figure 9 46 1 Power off the disk array and disconnect all cables The fans continue to run for a few seconds 2 Remove the outer covers Figure 9 40 3 Remove the following a b SS 9v d 9 78 Fan pack Figure 9 6 SPs Figure 9 9 Power supplies Figure 9 11 BBU located next to the power supply units Figure 9 13 All disk modules Figure 9 3 Each disk module should be labeled with the compartment number it was
487. using the hot spare instead of the failed disk If the hot spare has a different model number than the failed disk module it is replacing performance may degrade For this reason replace the failed disk module as soon as possible After a failed disk has been replaced the SP starts copying the data from the former hot spare onto the replacement disk Note If the failed disk is replaced during the rebuild process the SP finishes rebuilding the disk module on the hot spare before it begins copying data onto the replacement disk When the copy is done the disk group consists of disk modules in the original slots The SP automatically frees the hot spare to serve as a hot spare again A hot spare is most useful when the highest data availability is a requirement It eliminates the time and effort needed for someone to notice that a module has failed find a suitable replacement module and insert the module One or more disk modules per disk array can be configured as hot spares The following modules must not be configured as hot spares because they store the licensed internal microcode or serve as the disk array s cache vault e Model 10 A0 AL BO A2 A3 A4 e Model 20 AO BO CO A3 DO EO Hot Spare for a Model 10 In the following example the modules in slots A0 and BO are a RAID 1 mirrored pair those in Al through A4 and B1 are a RAID 5 group and the module in B2 is a hot spare illustrated in Figure 1 18 Eventually
488. ut firmly by sliding it into the slot Seat the disk module slowly into the connector Note When using RAID 5 RAID 1 0 or RAID 1 a rebuild begins when a disk module is replaced When using RAID 0 or independent disks System Administrator interaction is required to restore data to the disks Disk module handle Disk module Grounding strap Figure 9 3 Removing a Disk Module Model 10 20 Disk Module Removal Replacement 9 9 jueuie2e deH pue jeAoway lt Guide fd slot Disk module s guide Figure 9 4 Installing a Disk Module 9 10 Model 10 20 Disk Module Removal Replacement Model 10 20 Fan Pack Removal Replacement Install a replacement fan pack by performing the following steps CAUTION A failed fan pack can be replaced with the disk array powered on However the process must be completed within two minutes to avoid a thermal shutdown of the disk array 1 Removing a fan pack Figure 9 6 Slide the lock unlock lever to the unlock position Grasp the handle and pull open the fan pack Disconnect the fan pack cable Press down on the latches in the direction shown Pull the fan pack away from the chassis CAUTION Do not lift a fan pack by its handle as this can damage the fan pack 2 Installing a fan pack Figure 9 7 a b R Push the fan pack hinge channels onto the chassis hinge pins Be sure latches seat over hinge pins Reconnect the fan pack
489. v k v more wv vy more wv ek Figure 7 46 Tests Verify LUNs Screen 7 122 Support Tools Manager STM E hpterm igi xi Expert Tool Verify Luns for device C2300 Array on path 874 9 For Lun 0 on Array Serial 9621954 First block to verify 0 68908031 J Last block to verify lt 0 68908031 68908031 Blocks per verify cmd 1 65535 65535 Note Last block to verify must be gt first block to verify Figure 7 47 Tests Verify LUNs Select Blocks Screen Support Tools Manager STM 7 123 sonsouBeiq pue saninn ISOH Utilities gt System Configuration System Cache e hpterm Gf x Change the System Caching Parameters Specify the system cache states Read Cache X Enabled lt gt Disabled Peer Read Cache lt X gt Enabled lt gt Disabled Write Cache lt X gt Enabled lt gt Disabled Statistic Logging lt gt Enabled lt X gt Disabled Low Watermark value lt 0 10025 80 High Watermark value lt 0 10025 96 Cache Page Size 2K 4K 8K 16K 8 l Note Cache Page Size can only be changed if all system caching has been eAEK v more wv X k OY more wv w k vy more wv ek EN Figure 7 48 Utilities System Configuration System Cache Screen 7 124 Support Tools Manager STM Utilities System Configuration System Memory e hpterm Gf x Expert Tool Change System Memory Configuration for
490. ve reported a Unit Attention resulting from the SP microcode issuing a SCSI Bus Reset The drive was idle at time of reset or was just inserted OxOF Mode select occurred This error code should never be seen 0x10 Unit attention The drive reported an unexpected unit attention with unrecognized sense data This error code should never be seen Unsolicited Event Log 8 71 Buljooysajqnoly Table 8 15 Extended Status Codes cont d Extended status code 0x11 Description Device aborted command The disk did not complete a command issued by SP microcode 0x12 Bus reset occurred Similar to OxOE except FRU had a command active at time of reset 0x13 Verify error A read after a write did not verify what was written 0x14 to 0x18 Logical sector data error checksum error The SP internal microcode has detected a logical data error on a sector s of data 0x19 No sense data Drive had a problem with a request sense operation Ox1A Device too small A disk with insufficient capacity has been introduced in a bound array Ox1B Bus shut down Indicates the SP microcode has shut down the SCSI bus disabling any activity or its ability to interrupt Ox1C Unexpected sense key The targeted disk returned a sense key unknown to SP microcode 0x1D Unknown SCSI bus error The SCSI bus has reported an unrecognizable error Ox1E Diagnostic miscompare During the exe
491. ware Product in the configuration in which that Product is sold by HP or subsequently upgraded by HP c Customer s License confers no title or ownership in the Software and no rights in any associated source code and is not construed as a sale of any rights in Software d Customer may not disassemble or decompile the Software unless HP s prior written consent is either obtained or not required by law Upon request Customer will provide HP with reasonably detailed information regarding any disassembly or decompilation e Customer s entire License in Software is transferable subject to HP s prior written authorization and payment to HP of any applicable transfer fees Customer will immediately upon transfer deliver all copies of the Software to the party to whom HP has authorized transfer of Customer s License The transferee must agree in writing to the terms of Customer s License All License terms will be binding on involuntary transferees Customer s License will automatically terminate upon any transfer f Any third party supplier of Software may protect its rights against infringement of its copyright and violations of Customer s Licenses Reference R 21 o2ue49JoH R 22 g h i dD HP may terminate Customer s or any transferee s or any sublicensee s License in Software upon notice for failure to comply with any applicable License terms In the event of termination of Customer s License for any other reason Customer
492. will destroy or return to HP the Software and all copies of the Software immediately on termination Customer will remove and destroy all copies of the Software from any adaptation into which they are merged except for individual pieces of Customer data in a related database With HP s written consent Customer may retain one copy of the Software subsequent to termination for archival purposes In this clause on Licenses to the U S Government the term Customer means HP s direct purchaser any entity sublicensing the Software and the end user 1 IfSoftware is licensed for use in the performance of a U S Government prime contract or subcontract Customer agrees that Software has been developed entirely at private expense Customer agrees that Software and any derivatives or modification is adequately marked when the Restricted Rights Legend below is affixed to the Software or its storage media and is perceptible directly or with the aid of a machine or device Customer agrees to conspicuously put the following legend on the Software media with Customer s name and address added below the notices Restricted Rights Legend Use duplication or disclosure by the U S Government is subject to restrictions as set forth in subparagraph c 1 ii of the Rights in Technical Data and Computer Software clause in DFARS 252 2277 7013 Hewlett Packard Company 3000 Hanover Street Palo Alto CA 94304 U S A Rights for non Ocb U S Government D
493. with two SPs also provides better performance than the basic configuration because some physical disk units LUNs can be bound on one SP and the remaining LUNs can be bound on the second SP The two adapter two SP configuration has one host with two SCSI 2 adapters two external SCSI 2 buses and two SPs This configuration provides the highest availability and best array performance for single host configurations In general if one SP fails the LUNs owned by the failed SP can be accessed by the other SP Likewise if one host adapter fails the array can be accessed through the other adapter In a multi bus configuration two hosts connect to one disk array Each host connects to one of the disk array s SPs over a separate external SCSI 2 bus This configuration resembles two side by side basic configurations and provides modest high availability and better performance than the basic configuration If an SP fails in this configuration the system administrator can manually transfer access from the failed SP s disks to the other SP so information stored on those disk unit s can be accessed by the other host Likewise if a host adapter fails the system administrator can manually transfer access from the failed host adapter to the host The multi initiator configuration has two hosts each with one SCSI 2 adapter one external SCSI 2 bus and two SPs This configuration adds a backup host to a two SP configuration The multi bus m
494. ws to scroll through the screen information The row of v s at the top and bottom of the screen indicates that the window contains more information than can be displayed on the screen Support Tools Manager STM 7 101 sonsouBeiq pue seni ISOH amp hpterm Information Tool Log for C2300 Array on path 8 4 3 PSP Ps PS RAR AA PS PS Ps PS Ps Ps PS PS Ps PSP Ps PS Ps Ps Ps Ps Ps PS PS Ps PS Ps Ps PPS Ps PS Ps Ps PS Ps Ps PS Ps Ps PPS Ps PS PS Ps PS Ps Ps PS Ps Ps Ps Ps Ps PS Ps Ps PS PS Ps PS PS Ps PPS Ps PS PS PS PS IPS PSP Hardware path 874 9 MAP This SP A 9 0 7624 SP gt Array Date 04 29 98 Peer SP B 10 0 Array Timez 10 50 38 Serialtt 9621954 Cabnt Type Rack Mnt Qs SS Sea SS EE gem SSS SSS SS SS Fan A 0K l l l l I l l l l Fan B OK Rdy Rdu Rdu Rdu Rdu Ena Ena Ena Rdu Rdu VSC A OK IRS R 5 R51 R51 R51 R 5 R 5 R 5 R 1 I R 1 I VSC Bs OK I BOIBOIBOIBOIBO IAZIAZIAZIB4IB4l VSC C OK l l l l I l l l l l BBU DK Se ies SSS i quein m A B t D E A B t n E qo ur eee uen De J rum EE LE IC Es I I l l EAA AAAA AAAA AAAA AAAA EE m 1 o Figure 7 28 Information Tool Log Screen Map Section Status fields surround the physical map Table 7 14 xstm Status Fields for the Information Tool Log Status field values Description This SP ID hard address loop ID and SP model of this SP Peer SP ID hard address
495. x02000000 which means the SIMM in slot XUI or XU2 has failed Note Only cache sizes of 8 16 32 and 64 Mbytes are supported because Hewlett Packard does not ship other cache sizes A cache size of 40 Mbytes will work as long as all cache guidelines are followed SCSI 2 Bus Termination Power The SCSI 2 bus termination power term power is supplied by the host SCSI 2 adapter If the term power LED on the SCSI 2 adapter is off there is no termination power on the SCSI 2 bus Possible faults include e A blown fuse on the host adapter e A faulty host adapter e A faulty SCSI device e A faulty SCSI 2 cable connection bent pins or improper seating of cable connector 8 34 Troubleshooting the Disk Array Upgrading from an SP to a SP620 SCSI SP The following describes how to upgrade from an SP running microcode revisions 7 23 or 8 20 to a high performance SP620 SCSI SP running microcode revision 8 56 or later Microcode revisions 8 56 and later are for the SP620 SCSI SP only not the older model SP The SP620 SCSI SP is supported only on the model 20 disk array Each SP620 SCSI SP requires a minimum of one pair of SIMMs with minimum capacity of 8 Mbytes 2 each x 4 Mbytes CAUTION If microcode revision 8 20 is installed on the SP s being upgraded the microcode can be updated directly to revision 8 56 However if microcode revision 7 23 is installed on the SP s the microcode must first be updated to revision 8 20 and t
496. xchange number part number part number Description 1 005 42853 A3232 69107 Model 20 Midplane 2 005 039649 N A LED cable 3 N A N A Chassis 4 005 043512 N A Internal synchronous SE SCSI 2 cables 5 N A N A Internal AC harness bracket 6 113 000634 N A On off switch 7 005 039646 N A Power supply AC cable 8 005 039648 N A Fan cable 9 005 039645 N A SP address cable 10 005 041325 N A RIP PCA N A 5063 5396 A3388 69021 4 Gbyte disk module N A 5063 5398 A3232 69121 2 Gbyte disk module N A 5063 5393 A3666 69021 4 Gbyte LP disk module N A 5063 5335 A5169 69021 4 Gbyte LP disk module N A 5063 5394 A3541 69021 9 Gbyte HH disk module N A 5063 5333 A5132 69021 9 Gbyte LP disk module N A 5063 5334 A5133 69021 18 Gbyte LP disk module N A A3232 60005 A3232 69005 Battery backup unit BBU N A A3232 60003 A3232 69103 Storage processor SP N A A3549 60002 A3549 69102 Storage processor SP620 SCSI SP PCA N A A3232 60004 A3232 69104 Power supply VSC N A A3232 60006 A3232 69006 Fan pack R 18 Reference Table R 2 Replaceable Parts Model 20 cont d N A 5062 9398 A3232 69022 various vendor numbers including MT12D136G 7 MH1M36ADJ 7 HB56D136BR HB56D136BV 7B MH1M36BDJ 7 THM361020ASG 70 MCM36100ASG70 KMM5361000AG KMM5361000BG 7 KMM5361000CG 7 MH1M36BYJ 7 KMM5361003CG 7 KMM5361003BG 7 TM124MBK36B 70 SM5361000W 7DG KMM5361203AWG 7 4 Mbyte SIMM N A 5062 9399 A3232 69020 numerous vendor numbers including
497. xecution features of the Software and none of its program development features b HP grants to Customer a license to Use Software fixed in hardware and not removable in normal operation only when operating an associated Product in the configuration in which that Product is sold by HP or subsequently upgraded by HP Customer may transfer Software fixed in hardware and not removable in normal operation only upon transfer of the associated hardware Product c The following licenses are available for selected Software if so indicated on the Price List and upon payment of the applicable fee 1 Customer may make and Use one copy of that Software licensed directly from HP 2 Customer may sublicense one copy of that Software to an end user for its Use or sublicense one copy of Software to an HP authorized reseller for subsequent distribution to an end user for its Use These sublicenses must incorporate the terms of this License in a written and binding sublicensing agreement which will be made available to HP upon request Use authorization s for Software will apply to copies of that Software made under those Licenses GENERAL LICENSE TERMS a Customer s License does not include the right to updates upgrades or other enhancements HP reserves the right to require an additional license and fee for Use of the Software on upgraded computers processors or controllers b Software bundled with a hardware Product may be used only with that hard
498. xer MUX The HP 3000 Series 900 uses DTCs instead of MUXs For more information refer to Set Up a Connect Asynchronous Console Cables 2 35 uonejjeisuj Disk Array on an HP 3000 Series 900 DTC in this chapter and Initiating Grid Manager on MPE iX in chapter 6 Table 2 6 MUX Components Component Part number Part number Part number HP Terminal Multiplexer Card J2094A J2085A options J2092A required HP ADP Il or MDP 8 port MUX 5062 3054 Option 101 5181 2085 Box DDP Il HP Card to MUX box cable J2094 60004 Option 107 5062 3100 HP MUX box to RS 232 Array 17355D for 17355D for 17355D for Cable 1 per SP Series 800 Series 800 Series 800 Ask the system administrator which ports on the MUX box are available for use Determine the logical unit number of the MUX the configuration type terminal or modem of the ports being used and baud rate of the ports Set up the Disk Array on a Series 800 MUX Port The model 10 and model 20 disk arrays are delivered with a MUX cable 17355D or 24542H and a terminal cable 5061 5003 Use the 17355D MUX cable to connect the SP s RS 232 port on the disk array to the Series 800 s MUX box Use the terminal cable to connect a terminal directly to the SP s RS 232 port Both of these configurations allow access to the Grid Manager utility Install a MUX Card on a Series 800 System T500 or K Series To install a MUX card 1 Power down the Series 800 system
499. xternal Fast Wide differential SCSI 2 bus or buses a Usethe upper left hand section of the worksheet to diagram the cabling configuration b Fora complex multi host multi peripheral setup it may be more useful to draw a picture of the entire bus architecture through all adapters and peripherals on a separate sheet Figure 1 23 and Figure 1 24 illustrate cabling for two adapters with two arrays providing total redundancy of SPs host adapters and cables Planning Disk Unit Configurations 1 45 uonduioseg jonpold Adapter A Adapter B J Terminators p Array 1 Array 2 Figure 1 23 Example Configuration with Two Adapters with Two Arrays HP UX 9 04 Adapter A Adapter A Adapter B Adapter B Figure 1 24 Example Configuration with Two Arrays and Two Adapters in Each of Two Hosts HP UX 10 x and Later 1 46 Planning Disk Unit Configurations c Remember that to take advantage of transparent redundant SP functionality and the LVM dual path capability it is necessary to configure two physical paths to each LUN needing transparent redundant access Determine how many logical units each array is to be divided into and what LUN number and RAID mode will be assigned to each Even though the disk array is a fast wide SCSI device only LUN number values between 0 and 7 are supported on HP UX When binding LUNs Grid Manager assigns LUN numbers in ascending order starting with zero Fill in the LUN c
500. y To enter FE mode type CONnTROL P followed by in upper case FE To return to user mode type CONTROL P followed by in upper case USER Grid Manager Menus for Microcode Revision 8 56 Menu number from the Grid Manager Main Menu 1 Presentation Utility Displays the Presentation screen 2 Bind Physical Units 1 RAID 5 Group Individual Access Array Displays dialogue for making a RAID 5 LUN 2 RAID 3 Group Parallel Access Array Displays dialogue for making a RAID 8 LUN 3 RAID 1 Mirrored Pair Displays dialogue for making a RAID 1 LUN FE 4 RAID 0 Group Non redundant Individual Access Array Displays dialogue for making a RAID 0 LUN 5 RAID 1 0 Group Mirrored RAID 0 Group Displays dialogue for making a RAID 1 0 LUN FE 6 Individual Disk Unit Displays dialogue for making an independent mode disk 7 Hot Spare Displays dialogue for making a hot spare 3 Unbind Physical Units Displays dialogue for unbinding a LUN 4 View Unsolicited Event Log Displays the Unsolicited Event Logs starting at entry number 1 Must use FE mode to see all entries USER mode only shows 0x9xx and OxAxx errors Grid Manager Menus 6 25 Jobeue N pus Allows selection of a filter for either error code or FRU 5 Change Parameters 1 Change Physical Unit Parameters 1 Change Default SP Ownership of Unit Displays dialogue for changing SP ownership 2 Change Unit Rebuild Time Displays dialogue for changing rebuild time 3 Cha
501. y High Availability Features Planning Disk Unit Configurations 1 1 uonduioseg jonpold About the Disk Array The disk array also called storage system provides compact high capacity high availability disk storage on multiple disk modules that can be replaced while power is on The disk array connects to the host in two ways e To allow host access to the disk modules it connects by an external Fast Wide differential SCSI 2 bus to a SCSI 2 adapter in the host e Toallow administration configuration or monitoring of the disk array it connects by an RS 232 cable from the disk array s RS 232 asynchronous console port directly to an EM 100 capable terminal or by a MUX HP UX operating system or DTC port connection MPE iX operating system to the host The disk modules storage control processors SPs power supply units voltage semi regulated converters VSCs fan pack and battery backup unit BBU are hot replaceable which means they can be removed and replaced by the customer while the disk array is powered up Disk arrays and hosts can be configured in a number of ways A disk array can be supported by one host in a basic one SP two SP or two adapter two SP configuration or by two hosts in a multi bus multi initiator or multi bus initiator configuration The different disk array configurations are explained in chapters 3 HP 9000 Series 700 4 HP 9000 Series 800 and 5 HP 3000 Series 900 1 2 About
502. y are valuable to Customer Engineers in helping to troubleshoot the array e Thresholded codes Are normal and require no user action unless they occur at unusually high frequencies e Error codes Typically require user or Customer Engineer action Many of these events use 32 bit extended status information codes The majority of events reported by earlier revisions of microcode use only 8 bit extended status information codes It is recommended that host applications that parse Unsolicited Event Log information to display to users implement the following algorithm when examining an Unsolicited Sense Block The algorithm helps to determine if an 8 bit or 32 bit extended status field should be displayed 1 Extract the 32 bit value contained in bytes 22 through 25 If this value is non zero it should be displayed as the extended status information 8 56 Unsolicited Event Log 2 Ifthe 32 bit value extracted in step 1 is zero extract the 8 bit value contained in byte 17 This 8 bit value should be displayed as the extended status information even if it is zero Table 8 10 Informational Codes Event code Description 0x601 SP Power Up The SP is powered up and ready for I O This field also shows the revision of microcode Licensed Internal Code and the amount of SP memory in Mbytes in hexadecimal For example SP Powerup Rev 0x00080A01 0x00000020 means code revision 8 10 01 and 32 Mbytes
503. ype field This field allows selection of the RAID level of the physical disk group to be bound Choose the appropriate disk modules for the logical disk unit being bound Refer to Planning Disk Unit Configurations in chapter 2 for information Use the Tas key to move the cursor to the Sectors per Stripe Element field This field specifies the amount of disk space consumed by each stripe element in striped arrays If the bind type does not support disk striping this box is absent Press the TAB key until the OK field highlights or press the letter O key to confirm the choice To cancel the binding procedure press the TAB key until the Cancel field highlights or press the C key It takes some time for the system to bind the chosen logical disk unit Unbinding a Logical Disk Unit Using SAM 1 Press the F4 key to gain access to the top bar on the screen Press the arrow keys or the TAB key to highlight the Actions field Press RETURN Press RETURN to access the Actions menu Highlight the Disk Array Maintenance field Use the up and down arrow keys to move the cursor to the field that needs to be highlighted Highlight the Unbind field Press RETURN to access the unbind confirmation screen The screen will display a question asking if this LUN should be unbound Use the Tas key to highlight either Yes or No Press RETURN to confirm or cancel 7 128 SAM System Administration Management Tool LVM Logical Volume Manager For a deta
504. ytes 0x05000000 64 Mbytes 0x06000000 If the failing address is below the crossover address the failing SIMM is either SIMM 1 or SIMM 3 If the failing address is equal to or greater than the crossover address the failing SIMM is either SIMM 2 or SIMM 4 In the above example in a 32 Mbyte disk array the address 0x05123454 falls in the high half of memory and the failing SIMM is SIMM 4 But in a 64 Mbyte disk array the address 0x05123454 falls in the lower half of memory and the failing SIMM is SIMM 3 SIMM Troubleshooting for the SP620 SCSI SP The SP620 SCSI SP has four SIMM single in line memory module sockets that can be populated with 4 Mbyte and or 16 Mbyte SIMMs If a parity error is detected by the disk array during normal operation a panic results The panic code information defines which SIMM pair failed The format of a panic code is error code address for example 0x00000012 0x05123454 8 32 Troubleshooting the Disk Array All SIMM errors report a panic 0x012 parity error The address 0x05123454 references the failing SIMM pair The four SIMM modules occupy slots XU1 XU2 XU3 and XU4 which are clearly labeled on the SP PCA The XU1 and XU2 SIMM slots hold the first pair The XU3 and XU4 SIMM slots hold the second pair Use the following steps to identify which SIMM location has failed 1 Ifthe disk array has one pair of SIMMs both SIMMs should be replaced after receiving the 0x012 panic 2 Ifthe d
505. z equalizing with a hot spare Uns unsupported Support Tools Manager STM 7 103 sonsoubeiq pue seninf ISOH Table 7 16 LUN Bind State for the Map in the Information Tool Log Status field values Description blank disk is unbound Ind disk is bound to an individual disk LUN R 0 disk is bound to a RAID 0 LUN R 1 disk is bound to a RAID 1 LUN R 3 disk is bound to a RAID 3 LUN R 5 disk is bound to a RAID 5 LUN R 10 disk is bound to a RAID 1 0 LUN Hot disk is bound to a hot spare LUN AO B2 etc LUN ownership and LUN number The configuration section displays detailed LUN and disk information e hpterm CE x Information Tool Log for C2300 Array on path 874 9 PPP Ps PS PS Ps PS Ps Ps PS Ps Ps PS Ps Ps Ps Ps Ps PS Ps Ps Ps Ps Ps PS Ps Ps Ps Ps Ps PPS Ps PS Ps Ps PS Ps Ps Ps Ps Ps Ps Ps Ps Ps Ps Ps PPS Ps PS Ps Ps Ps Ps Ps Ps Ps PS Ps Ps Ps PS Ps Ps Ps Ps Ps PS PS Ps PS PSP PPP CONFIGURATION Lun C D Fru BS Cch ATr Tupe Vendor Product ID AprxCap BSz B RZ 0 BA x R M On RRID 5 C2300WDR5 35 39b 512 AO AD IBM DGHSOSY CLRROS 8 8gb BO BO IBM DGHSOSY CLRROS 8 8gb co CO IBM DGHSOSY CLAROS 8 8gb Do DO IBM DGHSOSY CLAROS 8 8gb EO EO IBM DGHSOSY CLAROS 8 8ab 1 ASA x R M On RRID 1 C2300WDR1 8 89b 512 AL AL IBM DCHSOSF CLAROS 8 8ab B1 Bi IBM DCHSOSF CLAROS 8 8ab 2 BAA x R M On RRID 5 C2300MDR5 5 7gb 512 C1 C1 IBM DGHS18Y CLRR18 17 89b D
506. z0dei NAI PII9A x surequoo ebegq 4yzodey NNT YL LZXQ0 ebed z2o uoraoe es de3ug ebeg uoraeanbrguoo we ysks wvzxo ebeg uoraqeanbr ebeg a3 Juo nud 6zx0 odeu NNT LZXO ebeg 3aodew age ioda3uo 9zxg opoo ebeg iequg esuesoepoucANWdVsWuIN c o durex q Ticks with this FRU Busy Hot Spare FRU Number FRU Inquiry Block is Valid Vendor Identification Product Identification Product Revision Level Example 3 NIKEARRY gt modesense Enter Page Code 0x26 Controller Report Page 0x27 LUN Report Page 0x29 FRU Configuration Page Ox2A System Configuration Page Enter selection cr page 0x29 The FRU Configuration Page contains FRU Number 0 FRU Status FRU Type Column Bus Hot Spare FRU 4 Vendor ID Product ID FRU Number 1 FRU Status FRU Type Column 255 Bus Hot Spare FRU 4 Vendor ID Product ID 2399 255 0 x 100ms 0xff Not replaced Seagate ST12550N HP03 FRU Enabled Individual Disk 0 Seagate ST12550N FRU Enabled Individual Disk 0 Seagate ST12550N Note BBUs There are 0 to 50 FRUs FRUs include disks power supply units fans SPs and Example 4 NIKEARRY gt modesense Enter test array cr 0 5 0 7 60 NIKEARRY Enter Page Code 0x26 Controller Report Page 0x27 LUN Report Page 0x29 FRU Configuration Page 0x2A System Configuration Page Enter selection cr page 0x2a The System Configuration Page contains

hp model 10 / 20 service manual

Contents

Download Pdf Manuals

Related Search

Related Contents