Home

PBS Professional User's Guide

Contents

1. 210 11 3 Environmental Requirements for HPCBP 271 11 4 Submitting HPC Basic Profile Jobs 212 vi PBS Professional 10 4 User s Guide Table of Contents 11 5 Managing HPCBP Jobs 11 6 Errors Logging and Troubleshooting 11 7 Advice and Caveats 24240 dese v ood ews ILS See ANO c cued skews neue tare cr eke ce 12 Using Provisioning 120 Definitions sse cke ye Se Rad Parada pea 12 2 How Provisioning Works 12 3 Requirements and Restrictions 12 4 Using Provisioning oc 65 0e8 cae eae es 12 5 Caveats and Errors 6 lt cau nauau ove cs Appendix A Converting NQS to PBS 13 1 Converting Date Specifications Appendix B License Agreement Index PBS Professional 10 4 User s Guide vii Table of Contents Vill PBS Professional 10 4 User s Guide Acknowledgements PBS Professional is the enhanced commercial version of the PBS software originally developed for NASA The NASA version had a number of cor porate and individual contributors over the years for which the PBS devel opers and PBS community is most grateful Below we provide formal legal acknowledgements to corporate and government entities then special thanks to individuals The NASA version of PBS contained software developed by NASA Ames Research Center Lawrence Livermore National Laboratory and MRJ Technology Solutions In addition it included software developed by the NetB
2. qsub 1l select 4 ncpus 1 lplace scatter job script Example 2 Using US mode run a single executable poe job with 4 ranks on hosts spread across the PBS allocated nodes listed in PBS_NODEFILE cat PBS_NODEFILE hostl host2 host3 host4 cat job script poe path mpiprog euilib us qsub 1l select 4 ncpus 1 lplace scatter job script Example 3 Using IP mode run executables prog and prog2 with 2 ranks of prog on host1 2 ranks of prog2 on host2 and 2 ranks of prog2 on host3 PBS Professional 10 4 User s Guide 239 Chapter 10 Multiprocessor Jobs cat PBS_NODEFILE hostl hostl host2 host2 host3 host3 cat job script echo progl gt tmp poe cmd echo progl gt gt tmp poe cmd echo prog2 gt gt tmp poe cmd echo prog2 gt gt tmp poe cmd echo prog2 gt gt tmp poe cmd echo prog2 gt gt tmp poe cmd poe cmdfile tmp poe cmd euilib ip rm tmp poe cmd qsub 1 select 3 ncpus 2 mpiprocs 2 l place scatter job script 240 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 Example 4 Using US mode run executables prog and prog2 with 2 ranks of prog on host1 2 ranks of prog2 on host2 and 2 ranks of prog2 on host3 cat PBS_NODEFILE host1l hostl host2 host2 host3 host3 cat job script echo progl gt tmp poe cmd echo progl gt gt tmp poe cmd echo prog2 gt gt tmp poe cmd echo prog2 gt gt tmp poe cmd echo prog2 gt gt tmp poe
3. An example 12 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 A job script that contains echo off E stepl bat E step2 bat should now be echo off call E stepl bat call E step2 bat Under Windows comments in the job script must be in ASCII characters 4 12 2 Passwords When running PBS in a password protected Windows environment you will need to specify to PBS the password needed in order to run your jobs There are two methods of doing this 1 by providing PBS with a pass word once to be used for all jobs single signon method or 2 by speci fying the password for each job when submitted per job method Check with your system administrator to see which method was configured at your site 4 12 2 1 Single Signon Password Method To provide PBS with a password to be used for all your PBS jobs use the pbs_password command This command can be used whether or not you have jobs enqueued in PBS The command usage syntax is pbs_password s server r d user When no options are given to pbs_password the password credential on the default PBS server for the current user i e the user who executes the command is updated to the prompted password Any user jobs previously held due to an invalid password are not released The available options to pbs_password are PBS Professional 10 4 User s Guide 73 Chapter 4 Submitting a PBS Job r Any user jobs previousl
4. 10 6 3 3 Restrictions on poe Jobs Users submitting poe jobs can run poe outside of PBS but they will see this warning pbsrun poe Warning not running under PBS PBS Professional 10 4 User s Guide 233 Chapter 10 Multiprocessor Jobs Users cannot run poe jobs without arguments inside PBS Attempting to do this will give the following error pbsrun poe Error interactive program name entry not supported under PBS poe will exit with a value of 1 Some environment variables and options to poe behave differently under PBS These differences are described in the next section 10 6 3 4 Options to poe and Environment Variables The usage for poe is poe program program_options poe options Users submitting jobs to poe can set environment variables instead of using options to poe The equivalent environment variable is listed with its poe option All options and environment variables except the follow ing are passed to poe devtype MP_DEVTYPE If InfiniBand is not specified in either the option or the envi ronment variable US mode is not used for the job euidevice MP_EUIDEVICE Ignored by PBS euilib ip jus MP_EUILIB If set to us the job runs in User Space mode If set to any other value that value is passed to IBM poe If the command line option euilib is set it will take precedence over the MP_EUILIB environment variable hosttile hfile MP_HOSTFILE Ignored If this is specified PBS print
5. Placement of jobs on vnodes is changed using the place statement l place modifier modifier where modifier is any combination of group excl and or one of free pack scatter The usage syntax for qalter is qalter job resources job list The following examples illustrate how to use the qalter command First we list all the jobs of a particular user Then we modify two attributes as shown increasing the wall clock time from 20 to 25 minutes and changing the job name from airfoil to engine qstat u barry Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time 51 south barry workq airfoil 930 1 0 16 R 0 01 54 south barry workq airfoil 1 0 20 Q qalter 1 walltime 20 00 N engine 54 qstat a 54 Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time To alter a job attribute via xpbs first select the job s of interest and the click on modify button Doing so will bring up the Modify Job Attributes dialog box From this window you may set the new values for any attribute you are permitted to change Then click on the confirm modify button at the lower left of the window The qalter command can be used on job arrays but not on subjobs or ranges of subjobs When used with job arrays any job array identifiers must be enclosed in double quotes e g qalter l walltime 25 00 1234 south 120 PBS Professional 10 4 User s Guide Working With PBS Jobs Chapter 6 Yo
6. Running Provision ing Held Held 292 PBS Professional 10 4 User s Guide Using Provisioning Chapter 12 12 3 Requirements and Restrictions 12 3 1 Host Restrictions 12 3 1 1 Single vnode Hosts Only PBS will provision only single vnode hosts Do not attempt to use provi sioning on hosts that have more than one vnode 12 3 1 2 Server Host Cannot Be Provisioned The Server host cannot be provisioned a MOM can run on the Server host but that MOM s vnode cannot be provisioned The provision_enable vnode attribute resources_available aoe and current_aoe cannot be set on the Server host 12 3 2 AOE Restrictions Only one AOE can be instantiated at a time on a vnode Only one kind of aoe resource can be requested in a job For example an acceptable job could make the following request l select 1 ncpus 1 aoe suset1 ncpus 2 aoe suse 12 3 2 1 Vnode Job Restrictions A vnode with any of the following jobs will not be selected for provision ing e One or more running jobs e A suspended job e A job being backfilled around PBS Professional 10 4 User s Guide 293 Chapter 12 Using Provisioning 12 3 2 2 Vnode Reservation Restrictions A vnode will not be selected for provisioning for job MyJob if the vnode has a confirmed reservation and the start time of the reservation is before job MyJob will end A vnode will not be selected for provisioning for a job in reservation R1 if the vnode has a c
7. Authorized Users userl south mydomain com server south ctime Mon Apr 28 11 00 00 2008 Mail Users userl mydomain com mtime Mon Apr 28 11 00 00 2008 Variable List PBS_O LOGNAME userl1 PBS O HOST south mydomain com Name 304 south Reserve Name NULL Reserve Owner userl south mydomain com RESV_CONFIRMED reserve substate 2 reserve start Thu May 1 15 00 00 2008 reserve end Thu May 1 16 00 00 2008 reserve duration 3600 reserve state PBS Professional 10 4 User s Guide 189 Chapter 8 Advanced PBS Features queue S304 Resource List ncpus 2 Resource List nodect 1 01 00 00 Resource List select l ncpus 2 Resource List walltime Resource List place free resv_nodes south ncpus 2 FREQ WEEKLY BYDAY MO COUNT 5 reserve _rrule reserve count 5 reserve index 2 Authorized Users userl south mydomain com server south ctime Mon Apr 28 11 01 00 2008 Mail Users userl mydomain com mtime Mon Apr 28 11 01 00 2008 Variable List PBS_O LOGNAME userl1 PBS_O HOST south mydo main com PBS TZID America Los Angeles 8 8 7 Deleting Reservations You can delete an advance or standing reservation by using the pbs_ rdel command For a standing reservation you can only delete the entire reser vation including all occurrences When you delete a reservation all of the jobs that have been submitted to the reservation are also deleted A reser vation can be deleted by its owne
8. Job Array Exit Status on page 223 8 2 Changing UNIX Job umask The W umask nnn option to qsub allows you to specify on UNIX systems what umask PBS should use when creating and or copying your stdout and stderr files and any other files you direct PBS to transfer on your behalf IMPORTANT This feature does not apply to Windows The following example illustrates how to set your umask to 022 i e to have files created with write permission for owner only rw r r qsub W umask 022 my_job PBS W umask 022 8 3 Requesting qsub Wait for Job Completion The W block true option to qsub allows you to specify that you want qsub to wait for the job to complete i e block and report the exit value of the job If job submission fails no special processing will take place If the job is successfully submitted qsub will block until the job terminates or an error occurs If qsub receives one of the signals SIGHUP SIGINT or SIGTERM it will print a message and then exit with the exit status 2 If the job is deleted before running to completion or an internal PBS error occurs an error 158 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 message describing the situation will be printed to this error stream and qsub will exit with an exit status of 3 Signals SIGQUIT and SIGKILL are not trapped and thus will immediately terminate the qsub process leaving the
9. nate when the job is submitted qsub will remain running until the job ter minates is aborted or the user interrupts qsub with a SIGINT the PBS Professional 10 4 User s Guide 91 Chapter 4 Submitting a PBS Job control C key If qsub is interrupted prior to job start it will query if the user wishes to exit If the user responds yes qsub exits and the job is aborted Once the interactive job has started execution input to and output from the job pass through qsub Keyboard generated interrupts are passed to the job Lines entered that begin with the tilde character and contain spe cial sequences are interpreted by qsub itself The recognized special sequences are qsub terminates execution The batch job is also termi nated susp If running under the UNIX C shell suspends the qsub pro gram susp is the suspend character usually CNTL Z asusp If running under the UNIX C shell suspends the input half of qsub terminal to job but allows output to continue to be displayed asusp is the auxiliary suspend character usually control Y 4 13 21 1 Caveats e Interactive batch jobs are not supported on Windows e Interactive batch jobs do not support job arrays 4 14 Failed Jobs Once a job has experienced a certain number of failures PBS holds the job If requeueing a job fails the job is deleted 92 PBS Professional 10 4 User s Guide Chapter 5 Using the xpbs GUI The PB
10. resources default mem 200mb default_chunk mem 100mb A job requesting 1 select 2 ncpus 1 mem 345mb will take 345mb from each of two vnodes and have a job wide limit of 690mb 2 345 The job s Resource _List mem will show 690mb 58 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 A job requesting l1 select 2 ncpus 2 will take 100mb default_chunk value from each vnode and have a job wide limit of 200mb 2 100mb The job s Resource _List mem will show 200mb A job requesting l ncpus 2 will take 200mb inherited from resources default and used to create the select spec from one vnode and a job wide limit of 200mb The job s Resource List mem will show 200mb A job requesting 1 nodes 2 will inherit the 200mb from resources_default mem which will be the job wide limit The memory will be taken from the two vnodes half 100mb from each The gener ated select spec is 2 ncpus l mem 100mb The job s Resource List mem will show 200mb 4 7 Submitting Jobs Using Select amp Place Examples Unless otherwise specified the vnodes allocated to the job will be allocated as shared or exclusive based on the setting of the vnode s sharing attribute Each of the following shows how you would use 1 select and l place 1 A job that will fit in a single host such as an Altix but not in any of the vnodes packed into the fewest vnodes 1 select 1 ncpus 10 mem 20gb l place pack In earlier versions this w
11. Because directives look like comments the scripting language ignores them Under Windows however the command interpreter does not recognize the symbol as a comment and will generate a benign non fatal warning when it encounters each PBS string While it does not cause a problem for the batch job it can be annoying or disconcerting to the user Therefore Windows users may wish to specify a different PBS directive via either the PBS Professional 10 4 User s Guide 71 Chapter 4 Submitting a PBS Job PBS_DPREFIX environment variable or the C option to qsub For example we can direct PBS to use the string REM PBS instead of PBS and use this directive string in our job script REM PBS 1 walltime 1 00 00 REM PBS l1 select mem 400mb REM PBS j oe date t my_application date t Given the above job script we can submit it to PBS in one of two ways set PBS _DPREFIX REM PBS qsub my_job_script or qsub C REM PBS my_job script For additional details on the C option to qsub see section 4 13 Job Submission Options on page 75 4 12 Windows Jobs 4 12 1 Submitting Windows Jobs Any bat files that are to be executed within a PBS job script have to be prefixed with call as in echo off call E stepl bat call E step2 bat Without the call only the first bat file gets executed and it doesn t return control to the calling interpreter
12. Options can be combined To show the state of all running subjobs use t r To show the state only of subjobs not job arrays use t J Table 9 4 Job Array and Subjob Options to qstat Option Result t Shows state of job array object and subjobs Will also show state of jobs J Shows state only of job arrays p Prints the default display with column for Percentage Com pleted For a job array this is the number of subjobs completed or deleted divided by the total number of subjobs For a job it is time used divided by time requested Examples We run an example job and an example job array on a machine with 2 pro cessors demoscript bin sh PBS N JobExample sleep 60 arrayscript bin sh PBS N ArrayExample PBS J 1 5 sleep 60 214 PBS Professional 10 4 User s Guide Job Arrays We run these scripts using qsub qsub arrayscript 1235 host qsub demoscript 1236 host Then qstat Job id Name User 1235 host ArrayExample user1 1236 host JobExample userl qstat J Job id Name User 1235 host ArrayExample user1 qstat p 1235 host ArrayExample user1 1236 host JobExample userl qstat t Job id Name User 1235 host ArrayExample userl 1235 1 host ArrayExample userl PBS Professional 10 4 User s Guide Chapter 9 Time Use S Queue 0 B workg 0 Q workq Time Use S Queue 0 B workg done S Queue Time Use S Queue 0 B workq 00 00
13. Problem with the start and or end 15086 pbs_rsub Bad time time of the reservation such as specification s Given start time is earlier than current date and time Missing start time or end time End time is earlier than start time Reservation duration exceeds 24 15129 pbs_rsub error DAILY hours and the recurrence fre recurrence duration can quency FREQ is set to DAILY not exceed 24 hours Reservation duration exceeds 7 15128 pbs_rsub error days and the frequency FREQ is set to WEEKLY WEEKLY recurrence duration cannot exceed 1 week 194 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 Table 8 5 Reservation Errors Server aa Log Description of Error Error Message Error Code Reservation duration exceeds 1 15130 pbs_rsub error hour and the frequency FREQ is HOURLY recurrence set to HOURLY or the BY rule is duration cannot exceed 1 set to BYHOUR and occurs every hour hour such as BYHOUR 9 10 The PBS _TZID environment vari None pbs _rsub error a valid able is not set correctly at the sub PBS _TZID timezone mission host rejection at environment variable is submission host required The PBS_TZID environment vari 15135 Unrecognized able is not set correctly at the sub PBS_TZID environ mission host rejection at Server ment variable 8 8 9 3 Time Required Between Reservations Leave enough time between re
14. R 00 00 qstat Jrt Req d Req d Elap Job ID Username Queue Jobname SessID NDS TSK Memory Time S Time 1235 5 host userl workq ArrayExamp 048 1 R 00 01 9 7 3 qdel Deleting a Job Array The qdel command will take a job array identifier subjob identifier or job array range The indicated object s are deleted including any currently running subjobs Running subjobs are treated like running jobs Subjobs not running will be deleted and never run Only one email is sent per deleted job array so deleting a job array of 5000 subjobs results in one email being sent 9 7 4 qalter Altering a Job Array The qalter command can only be used on a job array object not on subjobs or ranges Job array attributes are the same as for jobs PBS Professional 10 4 User s Guide 217 Chapter 9 Job Arrays 9 7 5 qorder Ordering Job Arrays in the Queue The qorder command can only be used with job array objects not on sub jobs or ranges This will change the queue order of the job array in associ ation with other jobs or job arrays in the queue 9 7 6 qmove Moving a Job Array The qmove command can only be used with job array objects not with subjobs or ranges Job arrays can only be moved from one server to another if they are in the Q H or W states and only if there are no running subjobs The state of the job array object is preserved in the move The job array will run to completion on the new server As with
15. The PBS man pages UNIX manual entries are installed on SGI systems under usr bsd or for the Altix in usr pbs man In order to find the PBS man pages users will need to ensure that usr bsd is set within their MANPATH The following example illustrates this for the C shell setenv MANPATH usr man usr local man usr bsd SMAN PATH 3 12 Setting Up Your Windows Environment This section discusses the setup steps needed for running PBS Professional in a Microsoft Windows environment including host and file access pass words and restrictions on home directories 24 PBS Professional 10 4 User s Guide Getting Started With PBS Chapter 3 3 12 1 Setting PBS EXEC on Windows In order to make it easier to submit a job script you can set up your envi ronment so that the correct value for PBS EXEC is used automatically Under Windows do the following 1 Look into C Program Files PBS Pro pbs conf and get the value of PBS_ EXEC It will be something like C Program Files PBS Pro exec 2 Set your environment accordingly cmd gt set PBS_EXEC lt path gt For example cmd gt set PBS_EXEC C Program Files PBS Pro exec 3 12 2 Windows User s HOMEDIR Each Windows user is assumed to have a home directory HOMEDIR where his her PBS jobs are initially started If a user has not been explicitly assigned a home directory then PBS will use this Windows assigned default as the base location for the us
16. for placing some type of hold on selected job s for releasing held job s for sending signals to selected job s that are running for writing a message into the output streams of selected job s for moving selected job s into some specified destination for exchanging order of two selected jobs in a queue for running selected job s admin only for requeueing selected job s that are running admin only The middle portion of the Jobs Panel has abbreviated column names indi cating the information being displayed as the following table shows Table 5 3 xpbs Job Column Headings Heading Meaning Job id Job Identifier 102 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 Table 5 3 xpbs Job Column Headings Heading Meaning Name Name assigned to job or script name User User name under which job is running PEs Number of Processing Elements CPUs requested CputUse Amount of CPU time used WalltUse Amount of wall clock time used S State of job Queue Queue in which job resides 5 3 5 xpbs Info Panel The Info panel shows the progress of the commands executed by xpbs Any errors are written to this area The INFO panel also contains a mini mize maximize button for displaying or iconizing the Info panel 5 3 6 xpbs Keyboard Tips There are a number of shortcuts and key sequences that can be used to speed up using xpbs
17. l place free 17 Here is a small job that uses a shared cpuset l select 1 ncpus 1 mem 512kb l place pack shared 18 Request a special resource available on a limited set of nodeboards such as a graphics card 1 select 1 ncpus 2 mem 2gb graphics True 1 ncpus 20 mem 20gb graphics False l place pack excl 19 Align SMP jobs on c brick boundaries 1l select 1 ncpus 4 mem 6gb l place pack group cbrick 20 Align a large job within one router if it fits within a router l select 1 ncpus 100 mem 200gb l place pack group router 21 Fit large jobs that do not fit within a single router into as few available routers as possible Here RES is the resource used for node grouping l select 1 ncpus 300 mem 300gb l place pack group lt RES gt 22 To submit an MPI job specify one chunk per MPI task For a 10 way MPI job with 2gb of memory per MPI task 1l select 10 ncpus 1 mem 2gb 23 To submit a non MPI job including a 1 CPU job or an OpenMP or 62 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 shared memory job use a single chunk For a 2 CPU job requiring 10gb of memory l select 1 ncpus 2 mem 10gb 4 7 1 Examples Using Old Syntax 1 Request CPUs and memory on a single host using old syntax 1l ncpus 5 mem 10gb will be converted into the equivalent l select 1 ncpus 5 mem 10gb l place pack 2 Request CPUs and memory on a named host along with custom resources including a floating l
18. qsub o temp my out job scr qsub e host e Documents and Settings user Desk top output The error output of the above job is to be copied onto the e drive on host using the path Documents and Settings user Desk top output The quote marks are required when arguments to qgsub contain spaces 4 13 3 Exporting Environment Variables The V option declares that all environment variables in the qsub com mand s environment are to be exported to the batch job qsub V my_job PBS V 4 13 4 Expanding Environment Variables The v variable_list option to qsub allows you to specify addi tional environment variables to be exported to the job variable_list names environment variables from the gsub command environment which are made available to the job when it executes These variables and their val ues are passed to the job These variables are added to those already auto matically exported Format comma separated list of strings in the form v variable or v variable value PBS Professional 10 4 User s Guide 79 Chapter 4 Submitting a PBS Job If a variable value pair contains any commas the value must be enclosed in single or double quotes and the variable value pair must be enclosed in the kind of quotes not used to enclose the value For example qsub v DISPLAY myvariable 32 my_job qsub v varl A B C D job sh qsub v a 10 var2 A B c 20 HOME home zzz job sh 4 13 5 Specifying
19. selectJobs List of jobs space separated to automatically select high light in the JOBS listbox selectOwners List of owners checked when limiting the jobs appearing on the Jobs listbox in the main xpbs window Specify value as Owners lt list_of owners gt See u option in qse lect 1B for format of lt list_of_owners gt selectStates List of job states to look for do not space separate when limiting the jobs appearing on the Jobs listbox in the main xpbs window Specify value as Job States lt states_string gt See s option in qselect 1B for for mat of lt states_string gt selectRes List of resource amounts space separated to consult when limiting the jobs appearing on the Jobs listbox in the main xpbs window Specify value as Resources lt res_string gt See 1 option in qse lect 1B for format of lt res_string gt selectExecTime The Execution Time attribute to consult when limiting the list of jobs appearing on the Jobs listbox in the main xpbs window Specify value as Queue Time lt exec_time gt See a option in gselect 1B for format of lt exec_time gt PBS Professional 10 4 User s Guide 113 Chapter 5 Using the xpbs GUI selectAcctName The name of the account that will be checked when limiting the jobs appearing on the Jobs listbox in the main xpbs window Specify value as Account Name lt account name gt See A option in qselect 1B for format of lt account_name
20. so deleting a job array results in one email being sent To delete a job using xpbs first select the job s of interest then click the delete button 6 3 1 Deleting Finished and Moved Jobs 6 3 1 1 Deleting Finished Jobs The qdel command does not affect finished jobs whether this job fin ished at the local server or at the destination server If you try to delete a finished job you will get the following error qdel Job lt jobid gt has finished 124 PBS Professional 10 4 User s Guide Working With PBS Jobs Chapter 6 6 3 1 2 Deleting Moved Jobs A job that has been moved to another server is either finished or still active i e queued or running If the moved job is active at the destination server the qdel command deletes the job 6 4 Sending Messages to Jobs To send a message to a job is to write a message string into one or more output files of the job Typically this is done to leave an informative mes sage in the output of the job Such messages can be written using the qmsg command IMPORTANT A message can only be sent to running jobs The usage syntax of the qmsg command is qmsg E O message_string job_identifier Example qmsg O output file message 54 qmsg O output file message 1234 server Job array identifiers must be enclosed in double quotes The E option writes the message into the error file of the specified job s The O option writes the message into the out
21. vendor s 3 RESTRICTIONS ON USE Notwithstanding the foregoing license grant Licensee shall not do or allow others to do any of the fol lowing a install use copy modify merge or transfer copies of the Prod ucts except as expressly authorized in this Agreement b use any back up copies of the Products for any purpose other than to replace the original copy provided by Altair in the event it is destroyed or damaged c disas semble decompile or unlock reverse translate reverse engineer or in any manner decode the Software or ISV Software for any reason d subli cense sell lend assign rent distribute publicly display or publicly per form the Products or Licensee s rights under this Agreement e allow use outside the Global Zone s or User Sites identified on the Order Form f allow third parties to access or use the Products such as through a service bureau wide area network Internet location or time sharing arrangement except as expressly provided in Section 2 b g remove any Proprietary Rights Notices from the Products h disable or circumvent the LMS pro vided with the Products or 1 link any software developed tested or sup ported by Licensee or third parties to the Products except for Licensee s own proprietary software solely for Licensee s internal use 4 OWNERSHIP AND CONFIDENTIALITY Licensee acknowl edges that all applicable rights in patents copyrights trademarks service marks and trade
22. 32 0 01 H 133 south barry workq trialx 20 0 01 W 154 south barry workq airfoil 930 32 1 30 R 0 32 Windows type the following at the cmd prompt all on one line for F usebackq j in qselect u barry 1 ncpus gt 16 do qstat a j 121 south 133 south 154 south Note This technique of using the output of the gselect command as input to qstat can also be used to supply input to other PBS commands as well PBS Professional 10 4 User s Guide 153 Chapter 7 Checking Job System Status 7 4 Selecting Jobs Using xpbs The xpbs command provides a graphical means of specifying job selec tion criteria offering the flexibility of the qselect command in a point and click interface Above the JOBS panel in the main xpbs display is the Other Criteria button Clicking it will bring up a menu that lets you choose and select any job selection criteria you wish The example below shows a user clicking on the Other Criteria button then selecting Job States to reveal that all job states are currently selected Clicking on any of these job states would remove that state from the selec tion criteria You may specify as many or as few selection criteria as you wish When you have completed your selection click on the Select Jobs button above the HOSTS panel to have xpbs refresh the display with the jobs that match your selection criteria The selected criteria will remain in effect 154 PBS Professional 10 4
23. 6 WARRANTY AND DISCLAIMER Altair warrants for a period of ninety 90 days after Licensee initially receives the Software that the Software will perform under normal use substantially as described in then current Documentation Supplier software included in the Software and ISV Software provided to Licensee shall be warranted as stated by the Sup plier or the ISV Copies of the Suppliers and ISV s terms and conditions of warranty are available on the Altair Support website Support services shall be provided in a workmanlike and professional manner in accordance with the prevailing standard of care for consulting support engineers at the time and place the services are performed ALTAIR DOES NOT REPRESENT OR WARRANT THAT THE PRODUCTS WILL MEET LICENSEE S REQUIREMENTS OR THAT THEIR OPERATION WILL BE UNINTERRUPTED OR ERROR FREE OR THAT IT WILL BE COMPATIBLE WITH ANY PARTICULAR HARDWARE OR SOFTWARE ALTAIR PBS Professional 10 4 User s Guide 307 EXCLUDES AND DISCLAIMS ALL EXPRESS AND IMPLIED WARRANTIES NOT STATED HEREIN INCLUDING THE IMPLIED WARRANTIES OF MERCHANTABILITY FITNESS FOR A PARTICULAR PURPOSE AND NON INFRINGEMENT THE ENTIRE RISK FOR THE PERFORMANCE NON PERFOR MANCE OR RESULTS OBTAINED FROM USE OF THE PROD UCTS RESTS WITH LICENSEE AND NOT ALTAIR ALTAIR MAKES NO WARRANTIES WITH RESPECT TO THE ACCU RACY COMPLETENESS FUNCTIONALITY SAFETY PERFOR MANCE OR ANY OTHER ASPECT OF ANY DESIGN PROTOTYPE OR FINAL PRODUCT DEVELOP
24. 9 6 Job Array Exit Status Exit Status Meaning 0 All subjobs of the job array returned an exit status of 0 No PBS error occurred Deleted subjobs are not consid ered 1 At least 1 subjob returned a non zero exit status No PBS error occurred 2 A PBS error occurred 9 10 9 Scheduling Job Arrays All subjobs within a job array have the same scheduling priority 9 10 9 1 Preemption Individual subjobs may be preempted by higher priority work 9 10 9 2 Peer Scheduling Peer scheduling does not support job arrays 9 10 9 3 Fairshare Subjobs are treated like jobs with respect to fairshare ordering fairshare accounting and fairshare limits If running enough subjobs of a job array causes the priority of the owning entity to change additional subjobs from that job array may not be the next to start PBS Professional 10 4 User s Guide 223 Chapter 9 Job Arrays 9 10 9 4 Placement Sets and Node Grouping All nodes associated with a single subjob should belong to the same place ment set or node group Different subjobs can be put on different place ment sets or node groups 224 PBS Professional 10 4 User s Guide Chapter 10 Multiprocessor Jobs 10 1 Job Placement Placement sets allow partitioning by multiple resources so that a vnode may be in one set that share a value for one resource and another set that share a different value for a different resource See the PBS Professional Administrator s
25. Files cshre 23 hosts equiv 27 login 23 pbs conf 29 112 profile 23 rhosts 27 xpbsre 111 files login 23 logout 23 floating licenses 48 free 56 freq_ spec 182 G Global Grid Forum 5 Graphical user interface 19 Grid 4 5 6 group resource 55 56 grouping 55 GUI 19 H here document 43 hfile 234 Hitchhiker s Guide 238 Hold job 84 or release job 121 Holding a Job Array 218 PBS Professional 10 4 User s Guide Index hostfile 234 HPC Basic Profile 269 HPC Basic Profile Job 270 HPC Basic Profile Server 269 HPCBP Executable location 272 Job resources 276 Job submission requirements 272 Monitoring jobs 277 Password requirement 271 qsub command 274 qsub syntax 274 Submitting jobs 272 Unsupported commands 285 User account 271 HPCBP Job 270 HPCBP MOM 270 HPS IP mode 232 US mode 232 I IBM POE 232 identifier 40 Identifier Syntax 203 InfiniBand 261 262 ensuring use 233 Information Power Grid 5 instance 179 instance of a standing reservation 179 instances option 235 Intel 256 Intel MPI 256 examples 258 Interactive job submission job arrays 204 Interactive batch jobs 91 317 Interdependency 6 interval_spec 182 IP mode HPS 232 J ja 197 Job checkpointable 84 comment 145 dependencies 159 identifier 40 management xi name 81 selecting using xpbs 154 sending messages to 125 sending signals to 126 submission options 75 tracking 155 Job Array Attributes 205 dependencies 221 identifier 20
26. HPS 232 User interfaces 6 name mapping 7 user job accounting 196 username 26 maximum 22 V Veridian 4 Viewing Job Information 141 Vnode Types 32 vnodes provisioning 290 W Wait for Job Completion 158 Widgets 95 PBS Professional 10 4 User s Guide Index Windows 24 26 job arrays staging 211 password 73 staging job arrays 211 Windows 2000 7 Workload management 3 X xpbs 21 107 111 113 115 buttons 106 configuration 111 job arrays 221 usage 93 126 127 152 154 161 xpbsmon 21 xpbsre 111 PBS Professional 10 4 User s Guide 323 Index 324 PBS Professional 10 4 User s Guide
27. Jobs are selected from those owned by a single Server When qselect suc cessfully completes it will have written to standard output a list of zero or more job identifiers which meet the criteria specified by the options Each option acts as a filter restricting the number of jobs which might be listed With no options the gselect command will list all jobs at the Server which the user is authorized to list query status of The u option may be used to limit the selection to jobs owned by this user or other specified users 152 PBS Professional 10 4 User s Guide Checking Job System Status Chapter 7 For a description of the qselect command see qselect on page 205 of the PBS Professional Reference Guide For example say you want to list all jobs owned by user barry that requested more than 16 CPUs You could use the following qselect command syntax qselect u barry l1 ncpus gt 16 121 south 133 south 154 south Notice that what is returned is the job identifiers of jobs that match the selection criteria This may or may not be enough information for your pur poses Many users will use shell syntax to pass the list of job identifiers directly into gst at for viewing purposes as shown in the next example necessarily different between UNIX and Windows UNIX qstat a qselect u barry l ncpus gt 16 Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time 121 south barry workq airfoil
28. Limit However if the job wide limit is modified the corresponding resources in the selection directive are not modified It would be impossible to deter mine where to apply the changes in a compound directive Reducing a job wide limit to a new value less than the sum of the resource in the directive is strongly discouraged This may produce a situation where the job is aborted during execution for exceeding its limits The actual effect of such a modification is not specified A job s walltime may be altered at any time except when the job is in the Exiting state regardless of the initial value If a job is queued requested modifications must still fit within the queue s and server s job resource limits If a requested modification to a resource would exceed the queue s or server s job resource limits the resource request will be rejected Resources are modified by using the 1 option either in chunks inside of selection statements or in job wide modifications using resource _name value pairs The selection statement is of the form l select N chunk N chunk where N specifies how many of that chunk and a chunk is of the form resource_name value resource_name value Job wide resource_name value modifications are of the form l resource_name value resource_name value It is an error to use a boolean resource as a job wide limit PBS Professional 10 4 User s Guide 119 Chapter 6 Working With PBS Jobs
29. MOM Error Condition Message Problems encountered when deleting job Cannot add SOAP HPCBP unable to add SOAP Header to Header MOM the terminate activities request message Problem reading HPCBP HPCBP unable to read HPCBP job iden job ID MOM tifier Bad HPC Basic Profile HPCBP unable to connect to the HPCBP Server connection MOM web service endpoint Problem with user s pass HPCBP unable to terminate job with word MOM user s credentials PBS Server Received malformed HPCBP unable to parse the response response from HPCBP MOM received for job deletion request Server from HPCBP Server Problems encountered when finding status of job Cannot add SOAP HPCBP unable to add SOAP Header to Header MOM the get activity statuses request message Problem reading HPCBP HPCBP unable to read HPCBP job iden JOB ID MOM tifier Bad HPC Basic Profile HPCBP unable to connect to the HPCBP Server connection MOM web service endpoint Received malformed HPCBP unable to parse the job status response from HPCBP MOM response received from HPCBP Server Server Problems encountered when finding node status 282 PBS Professional 10 4 User s Guide HPC Basic Profile Jobs Chapter 11 Table 11 4 Warning and Error Messages Logged by HPCBP MOM Error Condition Message Cannot add SOAP HPCBP unable to ad
30. MUST CLICK I DO NOT ACCEPT AND THEN HAVE AN AUTHORIZED PARTY IN THE ORGANIZATION THAT YOU REPRESENT ACCEPT THESE TERMS IF YOU OR THE ORGANIZATION THAT YOU REPRESENT HAS A MASTER SOFTWARE LICENSE AGREEMENT MASTER SLA ON FILE AT THE CORPORATE HEADQUARTERS OF ALTAIR ENGI PBS Professional 10 4 User s Guide 301 NEERING INC ALTAIR THE MASTER SLA TAKES PRECE DENCE OVER THESE TERMS AND SHALL GOVERN YOUR USE OF THE SOFTWARE MODIFICATION S OF THESE SOFTWARE LICENSE TERMS IS EXPRESSLY PROHIBITED ANY ATTEMTED MODIFICATION S WILL BE NONBINDING AND OF NO FORCE OR EFFECT UNLESS EXPRESSLY AGREED TO IN WRITING BY AN AUTHORIZED COR PORATE OFFICER OF ALTAIR ANY DISPUTE RELATING TO THE VALIDITY OF AN ALLEGED MODIFICATION SHALL BE DETER MINED IN ALTAIR S SOLE DISCRETION Altair Engineering Inc Software License Agreement THIS SOFTWARE LICENSE AGREEMENT including any Additional Terms together with the Agreement shall be effective as of the date of YOUR acceptance of these software license terms and conditions the Effective Date and is between ALTAIR ENGINEERING INC 1820 E Big Beaver Road Troy MI 48083 2031 USA a Michigan corporation Altair and YOU or the organization on whose behalf you have author ity to accept these terms the Licensee Altair and Licensee intending to be legally bound hereby agree as follows 1 DEFINITIONS In addition to terms defined elsewh
31. PBS Resources Resources can be available on the server and queues and on vnodes Jobs can request resources Resources are allocated to jobs and some resources such as memory are consumed by jobs The scheduler matches requested resources with available resources according to rules defined by the administrator PBS can enforce limits on resource usage by jobs PBS provides built in resources and in addition allows the administrator to define custom resources The administrator can specify which resources are available on a given vnode as well as at the server or queue level e g 32 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 floating licenses Vnodes can share resources The administrator can also specify default arguments for qsub These arguments can include resources See the qsub 1B man page Resources made available by defining them via resources_available at the server level are only used as job wide resources These resources e g walltime server_dyn_res are requested using 1 RESOURCE VALUE Resources made available at the host vnode level are only used as chunk resources and can only be requested within chunks using 1 select RESOURCE VALUE Resources such as mem and ncpus can only be used at the vnode level Resources are allocated to jobs both by explicitly requesting them and by applying specified defaults Jobs explicitly request resources either at the vnode level in chunks defined in
32. Professional 10 4 User s Guide HPC Basic Profile Jobs Chapter 11 If you specify HPCBP hosts as part of the job s select statement the list of of HPCBP hosts is passed to the HPCBP Server 11 3 Environmental Requirements for HPCBP 11 3 1 User Account at HPCBP Server You must be able to run commands at the HPCBP Server You must have an account in the Domain Controller at the HPCBP Server 11 3 2 HPCBP Submission Client Architecture You can submit HPCBP jobs only from submission hosts that have the cor rect architecture These are all supported Linux platforms on x86 and x86_ 64 11 3 3 Password Requirement For Job Submission The HPC Basic Profile Server requires a password and a username to per form operations such as job submission status termination etc The PBS Server must pass credential information to the HPCBP MOM at the time of job submission Before submitting an HPCBP job you must run the pos password command to store your password at the PBS server When you submit an HPCBP job you must supply a password This is done in one of two ways e The administrator sets the single_signon_password_enable server attribute to True e You use the Wpwd option to the qsub command to pass credential information to the PBS Server PBS Professional 10 4 User s Guide 271 Chapter 11 HPC Basic Profile Jobs 11 3 4 Location of Executable The executable that your job runs must be available at the HPC Server The f
33. These include Tip 1 All buttons which appear to be depressed in the dialog box subwindow can be activated by pressing the return enter key Tip 2 Pressing the tab key will move the blinking cursor from one text field to another Tip 3 To contiguously select more than one entry left click then drag the mouse across multiple entries PBS Professional 10 4 User s Guide 103 Chapter 5 Using the xpbs GUI Tip 4 To non contiguously select more than one entry hold the control left click on the desired entries 104 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 5 4 Setting xpbs Preferences The Preferences button is in the Menu Bar at the top of the main xpbs window Clicking it will bring up a dialog box that allows you to customize the behavior of xpbs 1 Define Server hosts to query 2 Select wait timeout in seconds 3 Specify xterm command for interactive jobs UNIX only 4 Specify which rsh ssh command to use PBS Professional 10 4 User s Guide 105 Chapter 5 Using the xpbs GUI 5 5 Relationship Between PBS and xpbs xpbs is built on top of the PBS client commands such that all the features of the command line interface are available through the GUI Each task that you perform using xpbs is converted into the necessary PBS com mand and then run Table 5 4 xpbs Buttons and PBS Commands Location Command Button PBS Command Hosts Panel deta
34. User must have access to the resources hosts that the site has configured for PBS e User must have a valid account username and group on the execution hosts e User must be able to transfer files between hosts e g via rcp or scp e User s time zone environment variable must be set correctly in order to use advance and standing reservations See section 8 8 9 1 Setting the Submission Host s Time Zone on page 193 The subsequent sections of this chapter discuss these requirements in detail and provide various setup procedures PBS Professional 10 4 User s Guide 21 Chapter 3 Getting Started With PBS 3 10 Usernames Under PBS By default PBS will use your login identifier as the username under which to run your job This can be changed via the u option to qsub See section 4 13 14 Specifying Job User ID on page 86 The user submitting the job must be authorized to run the job under the execution user name whether explicitly specified or not IMPORTANT PBS enforces a maximum username length of 15 characters If a job is submitted to run under a username longer than this limit the job will be rejected 3 11 Setting Up Your UNIX Linux Environment 3 11 1 Setting PBS EXEC on UNIX Linux In order to make it easier to submit a job script you can set up your envi ronment so that the correct value for PBS_EXEC is used automatically Under sh or bash do the following etc pbs conf 22 PBS Profe
35. V Table of Contents 8 10 Using Comprehensive System Accounting 196 8 11 Running PBS in a UNIX DCE Environment 198 8 12 Running PBS in a UNIX Kerberos Environment 199 8 13 Support for Large Page Mode on AIX 199 8 14 Checking License Availability 200 9 Job Arrays 201 9 1 DeHmtons vawe be cna checnneas bee serps 201 9 2 qsub Submitting a Job Array 204 9 3 Job Artray AUNDUES sess Traein caren 205 9 4 Job Array States nnan anan nannaa 206 9 5 PBS Environmental Variables 4 206 O06 Pile See free geese eee ehewe nes ob 40d 4h kes 206 9 7 PBS Commands anaana naana naana 212 9 8 Other PBS Commands Supported for Job Arrays 220 9 9 Job Arrays and 2P0S s ic c nswwdedeeiasiceres 221 9 10 More on Job Arrays sasanannannn anen 221 10 Multiprocessor Jobs 225 VOL Job Placement esri courte beset e sets E e 225 10 2 Submitting SMP Jobs ana nananana 226 10 3 Submitting MPI VODs lt 0 ee lt ce lt eesaasedee 226 10 4 OpenMP Jobs with PBS wis4iowss naana 228 10 5 Hybrid MPI OpenMP JobS 6i20 i40e0eede ees 228 10 6 MPI Jobs with PBS 4c lt 00 vsecuet eid aana 231 10 7 MPI Jobs on the Alti 0 lt 44vereceede haere 265 LOS PVM Jobs with PBS ou css ooh ens ananuna 266 10 9 Checkpointing SGI MPI Jobs nananana 267 11 HPC Basic Profile Jobs 269 ILI De NitionS 5 o iee dae bo oe riaierirerte rari 269 11 2 How HPC Basic Profile Jobs Work
36. a selection statement or in job wide resource requests See the pbs_resources 7B manual page Jobs are assigned limits on the amount of resources they can use These limits apply to how much the job can use on each vnode per chunk limit and to how much the whole job can use job wide limit Limits are derived from both requested resources and applied default resources Each chunk s per chunk limits determine how much of any resource can be used in that chunk Per chunk resource usage limits are the amount of per chunk resources requested both from explicit requests and from defaults Job resource limits set a limit for per job resource usage Job resource lim its are derived in this order from 1 explicitly requested job wide resources e g l resource value 2 the select specification e g 1 select 3 the queue s resources _default RES 4 the server s resources default RES 5 the queue s resources_max RES 6 the server s resources max RES The server s default_chunk RES does not affect job wide limits The resources requested for chunks in the select specification are summed and this sum is used for a job wide limit Job resource limits from sums of all chunks override those from job wide defaults and resource requests PBS Professional 10 4 User s Guide 33 Chapter 4 Submitting a PBS Job Various limit checks are applied to jobs If a job s job resource limit exceeds queue or server restrictions it will n
37. administrator or by the user MPD daemons are not started inside a PBS prologue script since it won t have the path of mpirun mpd that the user executed GM or MX which would determine the path to the MPD binary 10 6 7 3 Examples Example 1 Run a single executable MPICH GM job with 64 processes spread out across the PBS allocated hosts listed in PBS_NODEFILE PBS NODEFILE pbs host1l pbs host2 pbs host3 qsub 1 select 3 ncpus 1 MPICH GM HOME bin mpirun mpd np 64 path myprog x 1200 D lt job id gt If the GM MPD daemons are not running the PBS interface to mpirun mpd will start them as this user on the allocated PBS hosts The daemons may have been previously started by the administrator or the user PBS Professional 10 4 User s Guide 247 Chapter 10 Multiprocessor Jobs Example 2 Run an MPICH GM job with multiple executables on multiple hosts listed in the process group file procgrp Job script qsub 1 select 3 ncpus 1 echo hostl 1 userl x y a exe argl arg2 gt procgrp echo host2 1 userl x x b exe argl arg2 gt gt procgrp MPICH GM HOME bin mpirun mpd pg procgrp path mypro x 1200 rm f procgrp D lt job id gt When the job runs mpirun mpd will give the warning message warning pg is allowed but it is up to user to make sure only PBS hosts are specified MPI processes spawned are not guaranteed to be under PBS control The warning is issued because if any of the
38. and co chair of the Global Grid Forum s Scheduling Group 1 6 About Altair Engineering Through engineering consulting and high performance computing technol ogies Altair Engineering increases innovation for more than 1 500 clients around the globe Founded in 1985 Altair s unparalleled knowledge and expertise in product development and manufacturing extend throughout North America Europe and Asia Altair specializes in the development of high end open CAE software solutions for modeling visualization opti mization and process automation 1 7 Why Use PBS PBS Professional provides many features and benefits to both the computer system user and to companies as a whole A few of the more important fea tures are listed below to give the reader both an indication of the power of PBS and an overview of the material that will be covered in later chapters in this book PBS Professional 10 4 User s Guide 5 Chapter 1 Introduction Enterprise wide Resource Sharing provides transparent job scheduling on any PBS system by any authorized user Jobs can be submitted from any client system both local and remote crossing domains where needed Multiple User Interfaces provides a graphical user interface for submitting batch and interactive jobs querying job queue and system status and monitoring job progress PBS also provides a traditional command line interface Security and Access Control Lists permit the administrator to allow
39. assign this Agreement without the prior written consent of other party which shall not be unreasonably withheld All terms and conditions of this Agreement shall be binding upon and inure to the benefit of the par ties hereto and their respective successors and permitted assigns Waiver The failure of a party to enforce at any time any of the provisions of this Agreement shall not be construed to be a waiver of the right of the party thereafter to enforce any such provisions Severability If any provision of this Agreement is found void and unenforceable such provision shall be interpreted so as to best accomplish the intent of the parties within the lim its of applicable law and all remaining provisions shall continue to be valid and enforceable Headings The section headings contained in this Agree ment are for convenience only and shall not be of any effect in constructing the meanings of the Sections Modification No change or modification of this Agreement will be valid unless it is in writing and is signed by a duly authorized representative of each party Conflict In the event of any conflict between the terms of this Agreement and any terms and conditions on a Licensee Purchase Order or comparable document the terms of this Agreement shall prevail Moreover each party agrees any additional terms on any Purchase Order or comparable document other than the transaction items of a item s ordered b pricing c quantity d delivery
40. associated job either running or queued If the job runs to com pletion qsub will exit with the exit status of the job See also section 8 1 UNIX Job Exit Status on page 157 for further discussion of the job exit status For job arrays blocking qsub waits until the entire job array is complete then returns the exit status of the job array 8 4 Specifying Job Dependencies PBS allows you to specify dependencies between two or more jobs Depen dencies are useful for a variety of tasks such as 1 Specifying the order in which jobs in a set should execute 2 Requesting a job run only if an error occurs in another job 3 Holding jobs until a particular job starts or completes execution The W depend dependency_list option to qsub defines the dependency between multiple jobs The dependency list has the format type arg_list type arg_list where except for the on type the arg_1list is one or more PBS job IDs in the form jobid jobid There are several types after arg_list This job may be scheduled for execution at any point after all jobs in arg list have started execution afterok arg_list This job may be scheduled for execution only after all jobs in arg list have terminated with no errors See Warning about exit status with csh in EXIT STATUS PBS Professional 10 4 User s Guide 159 Chapter 8 Advanced PBS Features afternotok arg_list This job may be scheduled for
41. attribute shows that vnode s current AOE The scheduler queries each vnode s aoe resource and current_aoe attribute in order to determine which vnodes to provision for each job Provisioning can be used for interactive jobs A job s walltime clock starts when provisioning for the job has finished 12 2 1 Causing Vnodes To Be Provisioned An AOE can be requested for a job or a reservation When a job requests an AOE that means that the job will be run on vnodes running that AOE When a reservation requests an AOE that means that the reservation reserves vnodes that have that AOE available The AOE is instantiated on reserved vnodes only when a job requesting that AOE runs When the scheduler runs each job that requests an AOE it either finds the vnodes that satisfy the job s requirements or provisions the required vnodes For example if SLES is available on a set of vnodes that other 290 PBS Professional 10 4 User s Guide Using Provisioning Chapter 12 wise suit your job you can request SLES for your job and regardless of the OS running on those vnodes before your job starts SLES will be running at the time the job begins execution 12 2 2 Using an AOE When you request an AOE for a job the requested AOE must be one of the AOEs that has been configured at your site For example if the AOEs available on vnodes are rhel and sles you can request only those you cannot request suse You can request a re
42. by the job This value must be greater than zero Ifthe interval specified is less than that set on the execution queue in which the job resides the queue s interval is used Format Integer n No checkpointing is to be performed s Checkpointing is to be performed only when the Server exe cuting the job is shut down u Checkpointing is unspecified thus resulting in the same 6699 behavior as s PBS Professional 10 4 User s Guide 85 Chapter 4 Submitting a PBS Job If c is not specified the checkpoint attribute is set to the value u qsub c c my_job PBS c c 10 Checkpointing is not supported for job arrays 4 13 14 Specifying Job User ID PBS requires that a user s name be consistent across a server and its execu tion hosts but not across a submission host and a server A user may have access to more than one server and may have a different username on each server In this environment if a user wishes to submit a job to any of the available servers the username for each server is specified The wildcard username will be used if the job ends up at yet another server not specified but only if that wildcard username is valid For example our user is UserS on the submission host HostS UserA on server ServerA and UserB on server ServerB and is UserC everywhere else Note that this user must be UserA on all ExecutionA and UserB on all ExecutionB machines Then our user can use qsub
43. control of PBS Related Documents The following publications contain information that may also be useful to the user of PBS PBS Professional Quick Start Guide a short overview of the installation of PBS Professional PBS Professional 10 4 User s Guide xi PBS Professional Installation amp Upgrade Guide Con tains administrator s information on installing and upgrad ing PBS Professional PBS Professional Administrator s Guide Contains administrator s information required to configure and man age PBS as well as a discussion of how PBS components interoperate PBS Professional Programmer s Guide discusses the PBS application programming interface API security within PBS and inter daemon service communication PBS Professional Reference Guide Contains reference material for PBS Professional Ordering Software and Publications To order additional copies of this and other PBS publications or to pur chase additional software licenses contact an authorized reseller or the PBS Sales Department Contact information is included on the copyright page of this document Document Conventions PBS documentation uses the following typographic conventions abbreviation PBS command can be abbreviated such as sub commands to qmgr the shortest acceptable abbreviation is underlined command This fixed width font is used to denote literal commands filenames error messages and program output input Litera
44. core pro cessor system such as to a cache shared between cores or to the memory shared between cores If you want one chunk with 16 CPUs and 8 threads qsub 1 select 1 ncpus 16 ompthreads 8 You might be running an OpenMP application on a host and wish to run more threads than the number of CPUs requested because each thread is I O bound perhaps If you want one chunk with eight CPUs and 16 threads qsub 1 select 1 ncpus 8 ompthreads 16 10 5 Hybrid MPI OpenMP Jobs For jobs that are both MPI and multi threaded the number of threads per chunk for all chunks is set to the number of threads requested explicitly or implicitly in the first chunk except for MPIs that have been integrated with the PBS TM API For these MPIs LAM MPI you can specify the 228 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 number of threads separately for each chunk This means that for most MPIs OMP_NUM_THREADS and NCPUS will default to the number of ncpus requested on the first chunk and for integrated MPIs you can set the ompthreads resource separately for each chunk Should you have a job that is both MPI and multi threaded you can request one chunk for each MPI process or set mpiprocs to the number of MPI processes you want on each chunk For example to request 4 chunks each with 1 MPI process 2 CPUs and 2 threads qsub 1 select 4 ncpus 2 or qsub 1 select 4 ncpus 2 ompthreads 2 To request 4 chunks e
45. dataX and extrax and makes an output file 208 PBS Professional 10 4 User s Guide Job Arrays Chapter 9 newdataX as well as echoing which iteration it is on The dataX and extrax files will be staged from inputs to work then new datax will be staged from work to outputs bin sh PBS N StagingExample PBS J 1 2 PBS W stagein homedir work data array_ index host1 homedir inputs data array index nomedir work extra array_index host1 homedir inputs extra array_index PBS W stageout homedir work newdata array_ index host1 homedir outputs newdata array_index echo Main script index PBS ARRAY INDEX cd homedir work cat data PBS ARRAY INDEX extra PBS ARRAY INDEX gt gt newdata PBS ARRAY INDEX Local path execution directory homedir work Remote host data storage host hostl Remote path for inputs original data files dataX and extrax homedir inputs Remote path for results output of computation newdatax homedir outputs StageScript resides in homedir testdir In that directory we can run it by typing qsub StageScript It will run in homedir our home directory which is why the line cd homedir work PBS Professional 10 4 User s Guide 209 Chapter 9 Job Arrays is in the script Example 3 In this example we have the same script as before but we will run it in a staging and execution directory created by PBS StageScript takes two input files dat
46. deleted by the server Mail will be sent to the job submitter stating the error 160 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 Suppose you have three jobs job1 job2 and job3 and you want job3 to start after job1 and job2 have ended The first example below illustrates the options you would use on the qsub command line to specify these job dependencies qsub jobl 16394 jupiter qsub job2 16395 jupiter qsub W depend afterany 16394 16395 job3 16396 jupiter As another example suppose instead you want job2 to start only if job1 ends with no errors i e it exits with a no error status qsub jobl 16397 jupiter qsub W depend afterok 16397 job2 16396 jupiter Similarly you can use before dependencies as the following example exhibits Note that unlike after dependencies before dependencies require the use of the on dependency qsub W depend on 2 jobl 16397 jupiter qsub W depend beforeany 16397 job2 16398 jupiter qsub W depend beforeany 16397 job3 16399 jupiter You can use xpbs to specify job dependencies as well On the Submit Job window in the other options section far left center of window click on one of the three dependency buttons after depend before depend or concurrency These will launch a Dependency window in which you will be able to set up the dependencies you wish PBS Professional 10 4 User s Guide 161 Chapter 8 Advanced PBS Fe
47. directory as the job s staging and execution directory By default the sandbox attribute is not set The user can set the sandbox attribute via qsub or through a PBS direc tive For example qsub Wsandbox PRIVATE The job s sandbox attribute cannot be altered while the job is executing Table 8 2 Effect of Job s sandbox Attribute on Location of Staging and Execution Directory Job s sandbox attribute Es not set Job s staging and execution directory is the user s home directory HOME Job s staging and execution directory is the user s home directory PRIVATE Job s staging and execution directory is a job spe cific directory created by PBS If the qsub k option is used output and error files are retained on the primary execution host in the staging and execution directory This directory is removed along with all of its contents when the job finishes 8 6 2 2 The Job s jobdir Attribute and the PBS_JOBDIR Environment Variable The job s jobdir attribute is a read only attribute set to the pathname of the job s staging and execution directory on the primary host The user can view this attribute by using qstat f only while the job is executing The value of jobdir is not retained if a job is rerun it is undefined whether jobdir is visible or not when the job is not executing PBS Professional 10 4 User s Guide 165 Chapter 8 Advanced PBS Features The environment vari
48. each state array indices remaining string job array List of indices of subjobs still queued Range or list of ranges e g 500 552 596 1000 array_indices_ submitted string job array Complete list of indi ces of subjobs given at submission time Given as range e g 1 100 PBS Professional 10 4 User s Guide 205 Chapter 9 Job Arrays 9 4 Job Array States See Job Array States on page 438 of the PBS Professional Reference Guide and Subjob States on page 439 of the PBS Professional Reference Guide 9 5 PBS Environmental Variables Table 9 2 PBS Environmental Variables Environment Variable Name Used For Description PBS_ ARRAY INDEX subjobs Subjob index in job array e g 7 PBS_ARRAY ID subjobs Identifier for a job array Sequence number of job array e g 1234 server PBS_JOBID Jobs sub Identifier for a job or a subjob jobs For subjob sequence number and subjob index in brackets e g 1234 7 server 9 6 File Staging File staging for job arrays is like that for jobs with an added variable to specify the subjob index This variable is array_index This is the name of the variable that will be used for the actual array index The stdout and stderr files follow the naming convention for jobs but include the iden tifier of the job array which includes the subscripted index As with jobs the stagein and stageout key
49. end and which days and or hours apply freq_spec This is the frequency with which the reservation repeats Valid values are WEEKLY DAILY HOURLY When using a freq_spec of WEEKLY you may use an interval_spec of BYDAY and or BYHOUR When using a freq_spec of DAILY you may use an interval_spec of BYHOUR When using a freq_spec of HOURLY do not use an interval spec count_spec The exact number of occurrences Number up to 4 digits in length Format integer interval_spec Specifies the interval at which there will be occurrences Can be one or both of BYDAY lt days gt or BYHOUR lt hours gt Valid values are BYDAY MO TU WE TH FR SA SU and BYHOUR 0 1 2 23 When using both separate them with a semicolon Separate days or hours with a comma For example to specify that there will be recurrences on Tuesdays and Wednesdays at 9 a m and 11 a m use BYDAY TU WE BYHOUR 9 11 BYDAY should be used with FREQ WEEKLY BYHOUR should be used with FREQ DAILY or FREQ WEEKLY until_spec Occurrences will start up to but not after this date and time This means that if occurrences last for an hour and nor mally start at 9 a m then a time of 9 05 a m on the day specified in the until_ spec means that an occurrence will start on that day Format YYYYMMDD THHMMSS 182 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 Note that the year month day section is separated from the hour min
50. execution only after all jobs in arg _list have terminated with errors See Warning about exit status with csh in EXIT STATUS afterany arg_list This job may be scheduled for execution after all jobs in arg list have finished execution with or without errors before arg_list Jobs in arg list may begin execution once this job has begun execution beforeok arg_list Jobs in arg list may begin execution once this job termi nates without errors See Warning about exit status with csh in EXIT STATUS beforenotok arg_list If this job terminates execution with errors the jobs in arg list may begin See Warning about exit status with csh in EXIT STATUS beforeany arg_list Jobs in arg list may begin execution once this job termi nates execution with or without errors on count This job may be scheduled for execution after count dependencies on other jobs have been satisfied This type is used in conjunction with one of the before types listed count is an integer greater than 0 Job IDs in the arg list of before types must have been submitted with a type of on To use the before types the user must have the authority to alter the jobs inarg_list Otherwise the dependency is rejected and the new job aborted Error processing of the existence state or condition of the job on which the newly submitted job is a deferred service i e the check is performed after the job is queued If an error is detected the new job will be
51. file Windows timeoutSecs Specify the number of seconds before timing out waiting for a connection to a PBS host xtermCmd The xterm command to run driving an interactive PBS ses sion labelFont Font applied to text appearing in labels fixlabelFont Font applied to text that label fixed width widgets such as listbox labels This must be a fixed width font textFont Font applied to a text widget Keep this as fixed width font backgroundColor The color applied to background of frames buttons entries scrollbar handles foregroundColor The color applied to text in any context activeColor The color applied to the background of a selection a selected command button or a selected scroll bar handle disabledColor Color applied to a disabled widget signalColor Color applied to buttons that signal something to the user about a change of state For example the color of the Track Job button when returned output files are detected shadingColor A color shading applied to some of the frames to emphasize focus as well as decoration 112 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 selectorColor The color applied to the selector box of a radiobutton or checkbutton selectHosts List of hosts space separated to automatically select high light in the HOSTS listbox selectQueues List of queues space separated to automatically select highlight in the QUEUES listbox
52. following example shows the creation of an advance reservation ask ing for 1 vnode 30 minutes of wall clock time and a start time of 11 30 Since an end time is not specified PBS will calculate the end time based on the reservation start time and duration pbs_rsub R 1130 D 00 30 00 PBS returns the reservation ID R226 south UNCONFIRMED The following example shows an advance reservation for 2 CPUs from 8 00 p m to 10 00 p m pbs_rsub R 2000 00 E 2200 00 1 select 1 ncpus 2 PBS returns the reservation ID R332 south UNCONFIRMED 8 8 4 Creating Standing Reservations You create standing reservations using the pbs_rsub command You must specify a start and end date when creating a standing reservation The recurring nature of the reservation is specified using the r option to pbs_rsub The r option takes the recurrence_rule argument which specifies the standing reservation s occurrences The recurrence rule uses iCalendar syntax and uses a subset of the parameters described in RFC 2445 The recurrence rule can take two forms FREQ freq_spec COUNT count_spec interval_spec PBS Professional 10 4 User s Guide 181 Chapter 8 Advanced PBS Features In this form you specify how often there will be occurrences how many there will be and which days and or hours apply FREQ freq_spec UNTIL until_spec interval_spec In this form the user specifies how often there will be occurrences when the occurrences will
53. give stdout for index number 7 the name fixgamma o1234 7 and stderr the name fixgamma e1234 7 The name of the job array can also be given through stdin 9 6 2 Job Array Staging Syntax on Windows In Windows the stagein and stageout string must be contained in double quotes when using array_index Example of a stagein qsub W stagein foo array_index host 1 C WINNT Temp foo array_index J 1 5 stage_script Example of a stageout qsub W stageut C WINNT Temp foo array_index host 1 Q my_username foo array_index _out J 1 5 stage_script PBS Professional 10 4 User s Guide 211 Chapter 9 Job Arrays 9 7 PBS Commands 9 7 1 PBS Commands Taking Job Arrays as Arguments Note Some shells such as csh and tcsh use the square bracket J as a metacharacter When using one of these shells and a PBS command tak ing subjobs job arrays or job array ranges as arguments the subjob job array or job array range must be enclosed in double quotes The following table shows PBS commands that take job arrays subjobs or ranges as arguments The cells in the table indicate which objects are acted upon In the table Array the job array object Array Range the set of subjobs of the job array with indi ces in range given Array Index the individual subjob of the job array with the index given Array RUNNING the set of subjobs of the job array which are currently running Array QU
54. given those subjobs are requeued as jobs would be 9 7 10 qrun Running a Job Array The qrun command takes a subjob or a range of subjobs not a job array object If a single subjob is given as the argument it is run as a job would be Ifa range of subjobs is given as the argument the non running subjobs within that range will be run 9 7 11 tracejob on Job Arrays The tracejob command can be run on job arrays and individual subjobs When tracejob is run on a job array or a subjob the same information is displayed as for a job with additional information for a job array Note that subjobs do not exist until they are running so tracejob will not show any information until they are When tracejob is run on a job array the information displayed is only that for the job array object not the subjobs Job arrays themselves do not produce any MOM log information Running tracejob on a job array will give information about why a subjob did not start 9 7 12 qsig Signaling a Job Array If a job array object subjob or job array range is given to qsig all currently running subjobs within the specified set will be sent the signal PBS Professional 10 4 User s Guide 219 Chapter 9 Job Arrays 9 7 13 qmsg Sending Messages The qmsg command is not supported for job arrays 9 8 Other PBS Commands Supported for Job Arrays 9 8 1 qselect Selection of Job Arrays The default behavior of qselect is to return the job array identifier w
55. gt selectCheckpoint The checkpoint attribute relationship including the logical operator to consult when limiting the list of jobs appearing on the Jobs listbox in the main xpbs window Specify value as Checkpoint lt checkpoint_arg gt See c option in qselect 1B for format of lt checkpoint_arg gt selectHold The hold types string to look for in a job when limiting the jobs appearing on the Jobs listbox in the main xpbs win dow Specify value as Hold _ Types lt hold_string gt See h option in qselect 1B for format of lt hold_string gt selectPriority The priority relationship including the logical operator to consult when limiting the list of jobs appearing on the Jobs listbox in the main xpbs window Specify value as Prior ity lt priority_value gt See p option in qgselect 1B for format of lt priority_value gt selectRerun The rerunnable attribute to consult when limiting the list of jobs appearing on the Jobs listbox in the main xpbs win dow Specify value as Rerunnable lt rerun_val gt See r option in qgselect 1B for format of lt rerun_val gt selectJobName Name of the job that will be checked when limiting the jobs appearing on the Jobs listbox in the main xpbs window Specify value as Job Name lt jobname gt See N option in qselect 1B for format of lt jobname gt iconizeHosts View A boolean value true or false indicating whether or not to iconize the HOSTS regi
56. hosts listed in procgrp are not under the control of PBS then the processes on those hosts will not be under the control of PBS 10 6 8 PBS Jobs with MPICH MX s mpirun Using MPD mpirun mpd PBS provides an interface to MPICH MX s mpirun using MPD If exe cuted inside a PBS job this allows for PBS to track all MPICH MxX pro cesses started by the MPD daemons so that PBS can perform accounting and have complete job control If executed outside of a PBS job it behaves exactly as if standard mpirun ch_mx with MPD was used You use the same command as you would use outside of PBS either mpirun mpd or mpirun Ifthe MPD daemons are not already running the PBS interface will take care of starting them for you 248 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 8 1 Options Inside a PBS job script all of the options to the PBS interface are the same as mpirun ch_gmwith MPD except for the following m lt file gt The file argument contents are ignored and replaced by the contents of the PBS_ NODEFILE Np If not specified the number of entries found in the PBS_NODEFILE is used The maximum number of ranks that can be launched is the number of entries in PBS_NODEFILE pg The use of the pg option for having multiple executables on multiple hosts is allowed but it is up to user to make sure only PBS hosts are specified in the process group file MPI processes spawned on non PBS hos
57. i e it won t be used as an index value the highest index used will be the next below Y For example 1 100 2 gives 1 3 Syaa 99 Blocking qsub waits until the entire job array is complete then returns the exit status of the job array Interactive submission of job arrays is not allowed Examples Example 1 To submit a job array of 10 000 subjobs with indices 1 2 3 10000 qsub J 1 10000 job scr 1234 server domain com Example 2 To submit a job array of 500 subjobs with indices 500 501 502 1000 qsub J 500 1000 job scr 1235 server domain com Example 3 To submit a job array with indices 1 3 5 999 qsub J 1 1000 2 job scr 1236 server domain com 9 2 1 Interactive Job Submission Job arrays do not support interactive submission 204 PBS Professional 10 4 User s Guide Job Arrays 9 3 Job Array Attributes Chapter 9 Job arrays and subjobs have all of the attributes of a job In addition they have the following when appropriate These attributes are read only Table 9 1 Job Array Attributes Name Type Applies To Value array boolean job array True if item is job array array_id string subjob Subjob s job array identifier array_index string subjob Subjob s index num ber array_state_count string job array Similar to state_count attribute for server and queue objects Lists num ber of subjobs in
58. in a UNIX DCE Environment on page 198 W block opt Requesting qsub Wait for Job Completion on page 158 W pwd password Per job Password Method on page 75 and Running PBS in a UNIX DCE Environment on page 198 W sand Staging and Execution Directory User s box lt value gt Home vs Job specific on page 163 W umask nnn Changing UNIX Job umask on page 158 Z Suppressing Job Identifier on page 91 4 13 1 Specifying Queue and or Server The q destination option to qsub allows you to specify a partic ular destination to which you want the job submitted The destination names a queue a Server or a queue at a Server The qsub command will submit the script to the Server defined by the destination argument If the destination is a routing queue the job may be routed by the Server to a new destination If the q option is not specified the qsub command will sub mit the script to the default queue at the default Server See also the dis cussion of PBS_DEFAULT in Environment Variables on page 28 The destination specification takes the following form PBS Professional 10 4 User s Guide 71 Chapter 4 Submitting a PBS Job q queuve host Examples qsub q queue my_job qsub q server my_ job PBS q queueName qsub q queueName serverName my_job qsub q queueName serverName domain com my_job 4 13 2 Redirecting Output and Error Files PBS by
59. in alternate format qstat H Req d Req d Elap Job ID Username Queue Jobname SessID NDS TSK Memory Time S Time 101 S1 userl workq STDIN 5168 11 F 00 00 102 S1 userl Q1 S2 STDIN 1 2 M To see alternate format status for jobs job arrays and subjobs that are fin ished and moved use qstat Ht To see alternate format status for job arrays that are finished or moved use qstat HJ The H option is incompatible with the a i and r options 7 1 22 Viewing Estimated Start Times For Jobs You can view the estimated start times and vnodes of jobs using the qstat command If you use the T option to qstat when viewing job informa tion the Elap Time field is replaced with the Est Start field Running jobs are shown above queued jobs See qstat on page 217 of the PBS Professional Reference Guide If the estimated start time or vnode information is invisible to unprivileged users no estimated start time or vnode information is available via qstat Example output qstat T Req d Req d Est Job ID Username Queue Jobname SessID NDS TSK Memory Time S Start 5 hostl userl workq foojob 12345 1 1 128mb 00 10 R 9 hostl userl workq foojob af 1 128mb 00 10 Q 11 30 10 host1 userl workq foojob 1 1 128mb 00 10 Q Tu 15 7 host1 userl workq foojob 1 1 128mb 00 10 Q Jul 8 hostl userl workq foojob 1 1 128mb 00 10 Q 2010 11 host1 userl workq foojob 1 1 128mb 00 10 Q gt 5yrs PBS Professional 10 4 User s
60. jobs between jobs and job arrays Note Job dependencies are not supported for subjobs or ranges of subjobs PBS Professional 10 4 User s Guide 221 Chapter 9 Job Arrays 9 10 4 The Rerunnable Flag and Job Arrays Job arrays are required to be rerunnable PBS will not accept a job array that is not marked as rerunnable You can submit a job array without spec ifying whether it is rerunnable and PBS will automatically mark it as rerunnable 9 10 5 Accounting Job accounting records for job arrays and subjobs are the same as for jobs When a job array has been moved from one server to another the subjob accounting records are split between the two servers except that there will be no Q records for subjobs 9 10 6 Checkpointing Checkpointing is not supported for job arrays On systems that support checkpointing subjobs are not checkpointed instead they run to comple tion 9 10 7 Prologues and Epilogues If defined prologues and epilogues will run at the beginning and end of each subjob but not for job arrays 222 PBS Professional 10 4 User s Guide Job Arrays Chapter 9 9 10 8 Job Array Exit Status The exit status of a job array is determined by the status of each of the com pleted subjobs It is only available when all valid subjobs have completed The individual exit status of a completed subjob is passed to the epilogue and is available in the E accounting log record of that subjob Table
61. jobs a qstat on the server from which the job array was moved will not show the job array A qstat on the job array object will be redirected to the new server Note The subjob accounting records will be split between the two servers 9 7 1 qhold Holding a Job Array The qhold command can only be used with job array objects not with sub jobs or ranges A hold can be applied to a job array only from the Q B or W states This will put the job array in the H held state If any subjobs are running they will run to completion No queued subjobs will be started while in the H state 9 7 8 qris Releasing a Job Array The qrls command can only be used with job array objects not with sub jobs or ranges If the job array was in the Q or B state it will be returned to that state If it was in the W state it will be returned to that state unless its waiting time was reached it will go to the Q state 218 PBS Professional 10 4 User s Guide Job Arrays Chapter 9 9 7 9 qrerun Requeueing a Job Array The qrerun command will take a job array identifier subjob identifier or job array range Ifa job array identifier is given as an argument it is returned to its initial state at submission time or to its altered state if it has been qaltered All of that job array s subjobs are requeued which includes those that are currently running and completed and deleted Ifa subjob or range is
62. moved to a server that is not your default server When that happens you will need to give the job ID as an argument to qstat If you use only qstat your job will not appear to exist For example you submit a job to ServerA and it returns the jobid as 123 ServerA Then 123 ServerA is moved to ServerB In this case use qstat 123 PBS Professional 10 4 User s Guide 147 Chapter 7 Checking Job System Status or qstat 123 ServerA to get information about your job ServerA will query ServerB for the information To list all jobs at ServerB you can use qstat ServerB If you use qstat without the job ID the job will not appear to exist 7 1 20 Viewing Resources Allocated to a Job The exec_vnode attribute displayed via qstat shows the allocated resources on each vnode The exec_vnode line looks like exec_vnode hostA ncpus 1 For example a job requesting 1l select 2 ncpus 1 mem 1gb 1 ncpus 4 mem 2gb would get an exec_vnode of exec_vnode VNA ncpus 1 mem 1gb VNB ncpus 1 mem 1gb VNC ncpus 4 mem 2gb Note that the vnodes and resources required to satisfy a chunk are grouped by parentheses In the example above if two vnodes on a single host were required to satisfy the last chunk the exec_vnode might be exec_vnode VNA ncpus 1 mem 1gb VNB ncpus 1 mem l1gb VNC1 ncpus 2 mem 1gb VNC2 ncpus 2 mem 1gb You cannot use the qstat command to view any custom resource which has been c
63. number and or HostID on the Order Form License Units A parameter used by the LMS to determine usage of the Software permitted under this Agreement at any one time Licensed Workstations Single user computers located in the same Glo bal Zone s that Licensee owns or leases that are connected to the License Network Server via local area network or Licensee s private wide area network Maintenance Release Any release of the Products made generally avail able by Altair to its Licensees with annual leases or those with perpetual licenses who have an active maintenance agreement in effect that corrects programming errors or makes other minor changes to the Software The fees for maintenance and support services are included in the annual license fee but perpetual licenses require a separate fee Order Form Altair s standard form in either hard copy or electronic for mat that contains the specific parameters such as identifying Licensee s contracting office License Fees Software Support and License Net work Servers of the transaction governed by this Agreement Products Products include Altair Software ISV Software and or Suppli ers software and Documentation related to all of the forgoing Proprietary Rights Notices Patent copyright trademark or other propri etary rights notices applied to the Products packaging or media Software The Altair software identified in the Order Form and any Updates or Maintenance Rel
64. of UNIX and Linux from workstations and servers to supercomputers New platforms are being supported with each new release System Integration allows PBS to take advantage of vendor specific enhancements on different systems such as supporting cpusets on SGI sys tems Job Arrays are a mechanism for containerizing related work making it pos sible to submit query modify and display a set of jobs as a single unit PBS Professional 10 4 User s Guide 7 Chapter 1 Introduction 8 PBS Professional 10 4 User s Guide Chapter 2 Concepts and Components PBS is a distributed workload management system As such PBS handles the management and monitoring of the computational workload on a set of one or more computers Modern workload management solutions like PBS Professional include the features of traditional batch queueing but offer greater flexibility and control than first generation batch systems such as NQS Workload management systems have three primary roles Queuing The collecting together of work or tasks to be run on a com puter Users submit tasks or jobs to the resource manage ment system where they are queued up until the system is ready to run them PBS Professional 10 4 User s Guide 9 Chapter 2 Concepts and Components Scheduling The process of selecting which jobs to run when and where according to a predetermined policy Sites balance competing needs and goals on the system s to maximize e
65. on UNIX and PBS Destination Folder pbs conf on Windows The environment variable PBS_DPREF IX determines the prefix string which identifies directives in the job script The default prefix string is PBS however the Windows user may wish to change this as discussed in section 4 11 Changing the Job s PBS Directive on page 71 3 14 Temporary Scratch Space TMPDIR PBS creates an environment variable TMPDIR which contains the full path name to a temporary scratch directory created for each PBS job The directory will be removed when the job terminates Under Windows IMP will also be set to the value of TMPDIR amp The tem porary directory will be created under either winnt temp or win dows temp unless an alternative directory was specified by the administrator in the MOM configuration file Users can access the job specific temporary space by changing directory to it inside their job script For example UNIX cd STMPDIR Windows cd tTMPDIR PBS Professional 10 4 User s Guide 29 Chapter 3 Getting Started With PBS 30 PBS Professional 10 4 User s Guide Chapter 4 Submitting a PBS Job This chapter describes virtual nodes how to submit a PBS job how to use resources for jobs how to place your job on vnodes job attributes and sev eral related topics 4 1 Vnodes Virtual Nodes A virtual node or vnode is an abstract object representing a set of resources which form a
66. on Windows on page 25 To submit a PBS job type the following UNIX Linux shell script qsub lt name of shell script gt UNIX Linux Python script qsub S PBS_EXEC bin pbs python lt python job script gt Windows command script qsub lt name of job script gt 36 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 Windows Python script qsub S PBS_EXEC bin pbs python exe lt python job script gt If the path contains any spaces it must be quoted for example qsub S PBS_EXEC bin pbs python exe lt python job script gt 4 3 3 The Job Script 4 3 3 1 Python Job Scripts PBS allows you to submit jobs using a Python script You can use the same Python script under Windows or UNIX Linux PBS includes a Python package allowing Python job scripts to run you do not need to install Python To run a Python job script UNIX Linux qsub S PBS_EXEC bin pbs python lt script name gt Windows qsub S PBS_EXEC bin pbs python exe lt script name gt If the path contains any spaces it must be quoted for example qsub S PBS_EXEC bin pbs python exe lt python job script gt You can include PBS directives in a Python job script as you would in a UNIX shell script For example cat myjob py PBS 1 select 1 ncpus 3 mem lgb PBS N HelloJob print Hello PBS Professional 10 4 User s Guide 37 Chapter 4 Submitting a PBS Job Python job scripts can access Win32 APIs including
67. place con currently all copies of the task execute at about the same time The following example shows the posdsh command inside of a PBS batch job The options indicate that the user wants pbsdsh to run the myapp program with one argument app arg1 on all four vnodes allo cated to the job i e the default behavior bin sh PBS 1l select 4 ncpus 1 PBS 1 walltime 1 00 00 pbsdsh myapp app argl The pbsdsh command runs one task for each line in the PBS_NODEFILE Each MPI rank will get a single line in the PBS_NODEFILE so if you are running multiple MPI ranks on the same host you will still get multiple ppsdsh tasks on that host 8 8 Advance and Standing Reservation of Resources 8 8 1 Definitions Advance reservation A reservation for a set of resources for a specified time The reservation is only available to a specific user or group of users Standing reservation An advance reservation which recurs at specified times For example the user can reserve 8 CPUs and 10GB every 178 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 Wednesday and Thursday from 5pm to 8pm for the next three months Occurrence of a standing reservation An instance of the standing reservation An occurrence of a standing reservation behaves like an advance reservation with the following exceptions e while a job can be submitted to a specific advance reser vation it can only be submitted to the standi
68. queue r y n S path u user_ list W otherattributes value v variable list V z cmd argl or qsub version where cmd is the executable and arg1 is the first argument in the list 11 4 2 2 qsub Options for HPCBP Jobs The options to the qsub command set the attributes for the job The fol lowing table shows a list of PBS job attributes and their behavior for HPCBP jobs Table 11 2 Behavior of Job Attributes for HPCBP Jobs PBS Job attribute Behavior interactive Job is rejected with transient error Resource List See section 11 4 3 Requesting Resources on page 276 Output path Standard output is staged out to specified location 274 PBS Professional 10 4 User s Guide HPC Basic Profile Jobs Chapter 11 Table 11 2 Behavior of Job Attributes for HPCBP Jobs PBS Job attribute Behavior Error_path Standard error is staged out to specified location no_stdio_sockets Unsupported Shell_Path_List Unsupported Variable_List User s environment is passed to HPCBP Server alt_id Set to job ID returned by HPC Server exec_host Same as standard Set to list of hosts with num ber of CPUs for each exec_vnode Same as standard Set to list of vnodes with number of CPUs and amount of memory job_state See section 11 5 1 1 Job Status Reporting on page 278 resources_used Set to cputime used and amount of memory r
69. result files and disregard the scratch and other temporary files after the job terminates The result files that are interesting for this example end in dat e UNIX W stageout dat myworkstation project3 data e Windows W stageout dat mypc C project data 3 Stage in all files from an application data directory to a subdirectory e UNIX W stagein jobarea myworkstation crashtest1 e Windows W stagein jobarea mypc E crashtest1 4 Stage in data from files and directories matching wing e UNIX 170 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 W stagein myworkstation 848 wing e Windows W stagein mypc E flowcalc wing 5 Stage in bat and dat files to jobarea e UNIX W stagein jobarea myworkstation users me crash 1 at e Windows W stagein jobarea myworkstation C me crash1 at 8 6 4 3 Caveats When using a job specific staging and execution directory do not use an absolute path in local_path 8 6 4 4 Output Filenames The name of the job defaults to the script name if no name is given via qsub N via a PBS directive or via stdin For example if the sequence number is 1234 PBS N fixgamma gives stdout the name fixgamma o1234 and stderr the name fix gamma e1234 For information on submitting jobs see section 4 4 Submitting a PBS Job on page 40 PBS Professional 10 4 User s Guide 171 Chapter 8 Advanced PBS Features 8 6 5 Example of Using Job s
70. run with too few entries in the machinefile For instance specification of the following in 7 1 PBS 1 select 1 ncpus 1 host hostA 1 ncpus 2 host hostB mpirun np 3 tmp mytask would result in a 7 1 PBS_NODEFILE listing hostA hostB hostB but in 8 0 or later it would be hostA hostB which would conflict with the np 3 specification in mpirun as only 2 MPD daemons will be started The correct way now is to specify either a or b as follows a PBS 1 select 1 ncpus 1 host hostA 2 ncpus 1 host hostB b PBS l select 1 ncpus 1 host hostA 1 ncpus 2 host hostB mpiprocs 2 which will cause PBS_NODEFILE to list hostA hostB hostB and an mpirun np 3 will then be consistent 252 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 9 1 Options If executed inside a PBS job script all of the options to the PBS interface are the same as MPICH2 s mpirun except for the following host ghost For specifying the execution host to run on Ignored machinefile lt file gt The file argument contents are ignored and replaced by the contents of the SPBS_NODEFILE localonly lt x gt For specifying the lt x gt number of processes to run locally Not supported The user is advised instead to use the equiv alent arguments np lt x gt localonly Np If the user does not specify a np option then no default value is provided by the PBS wrapper scripts It is up to the loc
71. secrets embodied in the Products are owned by Altair and or its Suppliers or ISVs Licensee further acknowledges that the Products and all copies thereof are and shall remain the sole and exclusive property of Altair and or its Suppliers and ISVs This Agreement is a license and not a sale of the Products Altair retains all rights in the Products not expressly granted to Licensee herein Licensee acknowledges that the Products are confidential and constitute valuable assets and trade secrets of Altair and or its Suppliers and ISVs Licensee agrees to take the same precautions nec essary to protect and maintain the confidentiality of the Products as it does to protect its own information of a confidential nature but in any event no less than a reasonable degree of care and shall not disclose or make them available to any person or entity except as expressly provided in this Agreement Licensee shall promptly notify Altair in the event any unau thorized person obtains access to the Products If Licensee is required by any governmental authority or court of law to disclose Altair s or its ISV s PBS Professional 10 4 User s Guide 305 or its Suppliers confidential information then Licensee shall immediately notify Altair before making such disclosure so that Altair may seek a pro tective order or other appropriate relief Licensee s obligations set forth in Section 3 and Section 4 of this Agreement shall survive termination of this Agreement f
72. section 4 13 21 Interactive batch Jobs on page 91 5 7 Exiting xpbs Click on the Close button located in the Menu bar to leave xpbs If any settings have been changed xpbs will bring up a dialog box asking for a confirmation in regards to saving state information The settings will be saved in the xpbsrc configuration file and will be used the next time you run xpbs as discussed in the following section 5 8 The xpbs Configuration File Upon exit the xpbs state may be written to the xpbsrc file in the user s home directory See also section 3 12 2 Windows User s HOMEDIR on page 25 Information saved includes the selected host s queue s and job s the different jobs listing criteria the view states i e minimized maximized of the Hosts Queues Jobs and INFO regions and all settings in the Preferences section In addition there is a system wide xpbs con figuration file maintained by the PBS Administrator which is used in the absence of a user s personal xpbsrc file 5 9 xpbs Preferences The resources that can be set in the xpbs configuration file xpbsrc are serverHosts List of Server hosts space separated to query by xpbs A special keyword PBS_DEFAULT_SERVER can be used which will be used as a placeholder for the value obtained PBS Professional 10 4 User s Guide 111 Chapter 5 Using the xpbs GUI from the etc pbs conf file UNIX or PBS Des tination Folder pbs conf
73. that uses the selected entry as a parameter An entry widget is brought into focus with a left click To manipulate this widget simply type in the text value Use of arrow keys and mouse selec tion of text for deletion overwrite copying and pasting with sole use of mouse buttons are permitted This widget has a scrollbar for horizontally scanning a long text entry string A matrix of entry boxes is usually shown as several rows of entry widgets where a number of entries called fields can be found per row The matrix is accompanied by up down arrow buttons for paging through the rows of data and each group of fields gets one scrollbar for horizontally scanning long entry strings Moving from field to field can be done using the lt Tab gt move forward lt Cntrl f gt move forward or lt Cntrl b gt move backward keys PBS Professional 10 4 User s Guide 95 Chapter 5 Using the xpbs GUI A spinbox is a combination of an entry widget and a horizontal scrollbar The entry widget will only accept values that fall within a defined list of valid values and incrementing through the valid values is done by clicking on the up down arrows A button is a rectangular region appearing either raised or pressed that invokes an action when clicked with the left mouse button When the but ton appears pressed then hitting the lt RETURN gt key will automatically select the button A text region is an editor like widget This widget is
74. the OPTIONS subwindow to have PBS send you mail when the job terminates Now that the script is built you have four options of what to do next Reset options to default Save the script to a file Submit the job as a batch job Submit the job as an interactive batch job UNIX only Reset clears all the information from the submit job dialog box allowing you to create a job from a fresh start Use the FILE field in the upper left corner to define a filename for the script Then press the Save button This will cause a PBS script file to be generated and written to the named file Pressing the Confirm Submit button at the bottom of the Submit window will submit the PBS job to the selected destination xpbs will display a small window containing the job identifier returned for this job Clicking OK on this window will cause it and the Submit window to be removed from your screen On UNIX systems not Windows you can alternatively submit the job as an interactive batch job by clicking the nteractive button at the bottom of the Submit Job window Doing so will cause an X terminal window xterm to be launched and within that window a PBS interactive batch job submitted The path for the xterm command can be set via the prefer 110 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 ences as discussed above in section 5 4 Setting xpbs Preferences on page 105 For further details on usage and restrictions see
75. the following mod ules e Win32api e Win32con e Pywintypes 4 3 3 1 1 Windows Python Caveat If you have Python natively installed and you need to use the win32api make sure that you import pywintypes before win32api otherwise you will get an error Do the following cmd gt pbs_ python gt gt import pywintypes gt gt import win32api 4 3 3 2 UNIX Shell Scripts Since the job file can be a shell script the first line of a shell script job file specifies which shell to use to execute the script The user s login shell is the default but you can change this This first line can be omitted if it is acceptable for the job file to be interpreted using the login shell 4 3 3 3 Windows Command Scripts If the job file is a shell script specify the shell in the first line of the job file 4 3 3 3 1 Windows Caveats In Windows if you use not epad to create a job script the last line does not automatically get newline terminated Be sure to put one explicitly otherwise PBS job will get the following error message More when the Windows command interpreter tries to execute that last line 38 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 4 3 3 4 PBS Directives PBS directives are at the top of the script file They are used to request resources or set attributes A directive begins with the default string PBS Attributes can also be set using options to the qsub command which will o
76. the program then encrypted and saved securely for use by the job The password should be enclosed in double quotes Keep in mind that in a multi host job the password supplied will be propa gated to all the sister hosts This requires that the password be the same on the user s accounts on all the hosts The use of domain accounts for a multi host job will be ideal in this case Accessing network share drives resources within a job session also requires that you submit the job with a password viaqsub W pwd The Wpwd option to the qsub command is supported only on Windows and all supported Linux platforms on x86 and x86_ 64 4 13 Job Submission Options There are many options to the qsub command The table below gives a quick summary of the available options the rest of this chapter explains how to use each one Table 4 2 Options to the qsub Command Option Function and Page Reference A account_ string Specifying a Local Account on page 89 a date_time Deferring Execution on page 83 C DPREFIX Changing the Job s PBS Directive on page 71 PBS Professional 10 4 User s Guide TS Chapter 4 Submitting a PBS Job Table 4 2 Options to the qsub Command Option Function and Page Reference c interval Specifying Job Checkpoint Interval on page 84 e path Redirecting Output and Error Files on page 78 h Holding a Job Delaying Execution o
77. the user on the execution host s e The epilogue is run on the primary execution host with its current working directory set to the path of the job s staging and execution 172 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 directory and with PBS_JOBDIR and TMPDIR set in its environment e PBS stages out any files or directories e PBS removes any staged files or directories e PBS removes any job specific staging and execution directories and their contents and all TMPDIRs and their contents e PBS writes the final job accounting record and purges any job informa tion from the Server s database 8 6 7 Detailed Description of Job s Lifecycle 8 6 7 1 Creation of TMPDIR For each host allocated to the job PBS creates a job specific temporary scratch directory for the job If the temporary scratch directory cannot be created the job is aborted 8 6 7 2 Choice of Staging and Execution Directories If the job s sandbox attribute is set to PRIVATE PBS creates job spe cific staging and execution directories for the job If the job s sandbox attribute is set to HOME or is unset PBS uses the user s home directory for staging and execution 8 6 7 2 1 Job specific Staging and Execution Directories If the staging and execution directory cannot be created the job is aborted If PBS fails to create a staging and execution directory see the system administrator You should not depend on any par
78. them share a single processor on each host the user would request lselect 3 ncpus 1 mpiprocs 2 The PBS_NODEFILE would contain the following list VnodeA VnodeA VnodeB VnodeB Vnodec Vnodec If you want 3 chunks each with 2 CPUs and running 2 MPI process use 1 select 3 ncpus 2 mpiprocs 2 The PBS_NODEFILE would contain the following list VnodeA VnodeA VnodeB VnodeB Vnodec Vnodec PBS Professional 10 4 User s Guide 221 Chapter 10 Multiprocessor Jobs 10 4 OpenMP Jobs with PBS PBS Professional supports OpenMP applications by setting the OMP_NUM_THREADS variable automatically based on the resource request of a job in the environment of the job The OpenMP run time will pick up the value of OMP_NUM_THREADS and create threads appropri ately The OMP NUM THREADS value can be set explicitly by using the ompthreads pseudo resource for any chunk within the select statement If ompthreads is not used then OMP_NUM_THREADS is set to the value of the ncpus resource of that chunk If neither ncpus nor ompthreads is used within the select statement then OMP_NUM_THREADS is set to 1 To submit an OpenMP job is as a single chunk for a 2 CPU job requiring 10gb of memory you would use qsub 1 select 1 ncpus 2 mem 10gb You might be running an OpenMP application on a host and wish to run fewer threads than the number of CPUs requested This might be because the threads need exclusive access to shared resources in a multi
79. to 10 30 8 8 4 2 Requirements for Creating Standing Reservations e The user must specify a start and end date See the R and E options to the pbs_rsub command in section 8 8 5 The pbs_rsub Com mand on page 185 e The user must set the submission host s PBS_TZID environment variable The format for PBS_TZID is a timezone location Exam ple America Los Angeles America Detroit Europe Berlin Asia Calcutta See section 8 8 9 1 Set ting the Submission Host s Time Zone on page 193 e The recurrence rule must be one unbroken line See the r option to pbs_rsub in section 8 8 5 The pbs_rsub Command on page 185 e The recurrence rule must be enclosed in double quotes 184 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 e Vnodes that have been configured to accept jobs only from a specific queue vnode queue restrictions cannot be used for advance or standing reservations See your PBS administrator to determine whether some vnodes have been configured to accept jobs only from specific queues 8 8 4 3 Examples of Creating Standing Reservations For a reservation that runs every day from 8am to 10am for a total of 10 occurrences pbs_rsub R 0800 E 1000 r FREQ DAILY COUNT 10 Every weekday from 6am to 6pm until December 10 2008 pbs_rsub R 0600 E 1800 r FREQ WEEKLY BYDAY MO TU WE TH FR UNTIL 20081210 Every week from 3pm to 5pm on Monday Wednesday and Friday for 9 occ
80. to create both advance and standing reservations See section 8 8 Advance and Standing Reservation of Resources on page 178 16 PBS Professional 10 4 User s Guide Getting Started With PBS Chapter 3 3 5 Deprecations The sort_priority option to job_sort_key is deprecated and is replaced with the job_ priority option The I nodes nodespec form is replaced by the 1 select and 1 place statements The nodes resource is no longer used The l resource rescspec form is replaced by the 1 select statement The time shared node type is no longer used and the ts suffix is obsolete The cluster node type is no longer used The resource arch is only used inside of a select statement The resource host is only used inside of a select statement The nodect resource is obsolete The ncpus resource should be used instead Sites which currently have default values or limits based on nodect should change them to be based on ncpus The neednodes resource is obsolete The ssinodes resource is obsolete Properties are replaced by boolean resources The ppn resource is deprecated The a option to the qselect command is deprecated The W delay nnnn option to qdel is deprecated PBS Professional 10 4 User s Guide 17 Chapter 3 Getting Started With PBS 3 6 Backward Compatibility 3 6 1 Job Dependencies Affected By Job History Enabling job history changes the behavior of dependent jobs If a job j1 depends on a finis
81. usable part of a machine This could be an entire host or a nodeboard or a blade A single host can be made up of multiple vnodes Each vnode can be managed and scheduled independently PBS views hosts as being composed of one or more vnodes Jobs run on one or more vnodes See the pbs_node_attributes 7B man page PBS Professional 10 4 User s Guide 31 Chapter 4 Submitting a PBS Job 4 1 1 Relationship Between Hosts Nodes and Vnodes A host is any computer Execution hosts used to be called nodes However some machines such as the Altix can be treated as if they are made up of separate pieces containing CPUs memory or both Each piece is called a vnode Some hosts have a single vnode and some have multiple vnodes PBS treats all vnodes alike in most respects Chunks cannot be split across hosts but they can be split across vnodes on the same host Resources that are defined at the host level are applied to vnodes A host level resource is shared among the vnodes on that host This sharing is managed by the MOM 4 1 2 Vnode Types What were called nodes are now called vnodes All vnodes are treated alike and are treated the same as what were called time shared nodes The types time shared and cluster are deprecated The ts suffix is deprecated It is silently ignored and not preserved during rewrite The vnode attribute nt ype is only used to distinguish between PBS and Glo bus vnodes It is read only 4 2
82. with the exceptions and differences noted here Under PBS the poe command is slightly different See IBM s poe pbsrun poe on page 132 of the PBS Professional Reference Guide 232 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 3 1 The User s Environment In order to ensure that the InfiniBand switch can be used for a job the job must have PBS_GET_IBWINS 1 in its environment This can be han dled either by the administrator or by the job submitter Users submitting poe jobs may choose to set PBS_GET_IBWINS 1 in their shell envi ronment and use the V option to the qsub command csh setenv PBS GET IBWINS 1 bash PBS GET IBWINS 1 export PBS GET IBWINS PBS requires no other changes to the user s environment Do not set the PBS_O_ HOST environment variable See section 10 6 3 5 3 Environment Variable on page 236 10 6 3 2 Using the InfiniBand Switch To ensure that a job uses the InfiniBand switch make sure that the job s environment has PBS_GET_IBWINS set to 1 This can be accomplished the following ways e The administrator sets this value for all jobs e To set the environment variable for each job the job submitter sets PBS_GET_IBWINS 1 in their shell environment and uses the V option to every qsub command See the previous section e To set the environment variable for one job the job submitter uses the v PBS_GET_IBWINS 1 option to the qsub command
83. 00 R workq 215 Chapter 9 1235 2 host 1235 3 host 1235 4 host 1235 5 host 1236 host qstat Jt 1235 1 host 1235 2 host 1235 3 host 1235 4 host 1235 5 host qstat 1235 1 host 1235 2 host 1235 3 1235 4 host host 1235 5 host qstat pt 1235 host 216 Jtp ArrayExample ArrayExample ArrayExample ArrayExample JobExample ArrayExample ArrayExample ArrayExample ArrayExample ArrayExample ArrayExample ArrayExample ArrayExample ArrayExample ArrayExample ArrayExample Job Arrays After the first two subjobs finish userl userl userl userl userl User Time Use userl 00 00 00 userl 00 00 00 userl 0 userl 0 userl 0 User done S userl 100 X userl 100 X userl R userl R user1 Q User done S user1 40 B PBS Professional 10 4 User s Guide 00 00 00 R workq 0 Q workg 0 Q workq 0 Q workq 0 Q workq Job Arrays Chapter 9 1235 1 host ArrayExample userl 100 X workq 1235 2 host ArrayExample userl 100 X workq 1235 3 host ArrayExample userl R workq 1235 4 host ArrayExample userl R workq 1235 5 host ArrayExample userl Q workq 1236 host JobExample user1 Q workq Now if we wait until only the last subjob is still running qstat rt Req d Req d Elap Job ID Username Queue Jobname SessID NDS TSK Memory Time S Time 1235 5 host userl workq ArrayExamp 3048 1 R 00 00 1236 host userl workq JobExample 3042 1
84. 2 Using Provi sioning on page 289Chapter 12 Using Provisioning on page 289 3 2 2 Walltime as Checkpoint Interval Measure PBS allows a job to be checkpointed according to its walltime usage See the pbs_job attributes 7B manual page 14 PBS Professional 10 4 User s Guide Getting Started With PBS Chapter 3 3 2 3 Employing User Space Mode on IBM InfiniBand Switches PBS allows users submitting POE jobs to use InfiniBand switches in User Space mode See section 10 6 3 MPI Jobs Using AIX POE on page 232 3 3 New Features in Version 10 1 3 3 1 Submitting HPCBP Jobs PBS Professional can schedule and manage jobs on one or more HPC Basic Profile compliant servers using the Grid Forum OGSA HPC Basic Profile web services standard You can submit a generic job to PBS so that PBS can run it on an HPC Basic Profile Server This chapter describes how to use PBS for HPC Basic Profile jobs See Chapter 8 Metascheduling Using HPC Basic Profile on page 431 3 3 2 Using Job History Information PBS Professional can provide job history information including what the submission parameters were whether the job started execution whether execution succeeded whether staging out of results succeeded and which resources were used PBS can keep job history for jobs which have fin ished execution were deleted or were moved to another server See sec tion 6 9 Using Job History Information on page 131 3 3 3 Reservatio
85. 2 range 202 States 206 Job Array Run Limits 221 Job Arrays 201 checkpointing 222 deleting 217 exit status 223 interactive submission 204 PBS commands 212 placement sets 224 prologues and epilogues 222 qalter 217 qdel 217 qhold 218 qmove 218 qorder 218 qrerun 219 qrls 218 318 Index qrun 219 qselect 220 run limits 221 starving 221 status 214 submitting 204 tracejob 219 Job Arrays and xpbs 221 job container 196 Job Script 36 Job Submission Description Lan guage 270 Job Submission Options 75 jobs MPI 226 PVM 266 SMP 226 job wide 46 JSDL 270 K Kerberos 199 qsub W cred DCE 198 KRB5 199 krb5 199 L Large Page Mode 199 Limits on Resource Usage 53 Linux job container 196 Listbox 95 LoadLeveler 232 Load Leveling 6 M man pages SGI 24 management xi MANPATH 24 Message Passing Interface 231 PBS Professional 10 4 User s Guide meta computing 5 Modifying Job Attributes 117 MOM 11 Monitoring 10 Moving 218 jobs between queues 128 Moving a Job Array 218 MP_DEVTYPE 234 MP_EUIDEVICE 234 MP_EUILIB 234 MP_HOSTFILE 234 MP_INSTANCES 235 MP_PROCS 235 MPI 231 AIX and POE 232 HP UX and Linux 241 Intel MPI 256 examples 258 MPICH_GM rsh ssh examples 243 MPICH2 251 262 examples 254 263 MPICH GM MPD 246 examples 247 rsh ssh 242 MPICH MX MPD 248 examples 250 rsh ssh 244 examples 245 MVAPICH1 261 examples 262 SGI Altix 267 MPI jobs 226 MPI LAPI 232 MPICH 231 MPICH_GM rsh ssh PBS Prof
86. 617 of the PBS Profes sional Administrator s Guide 8 7 The pbsdsh Command The pbsdsh command allows you to distribute and execute a task on each of the vnodes assigned to your job epbsdsh uses the PBS Task Manager API see tm 3 to distribute the program on the allocated vnodes IMPORTANT The pbsdsh command is not available under Windows Usage of the pbsdsh command is pbsdsh c N o s v program program args pbsdsh n N o s v program program args Note that the double dash must come after the options and before the pro gram and arguments The double dash is only required for Linux The available options are c N The program is spawned on the first V vnodes allocated If the value of N is greater than the number of vnodes it will wrap around running multiple copies on the vnodes This option is mutually exclusive with n n N The program is spawned on a single vnode which is the N th vnode allocated This option is mutually exclusive with c O The program will not wait for the tasks to finish S If this option is given the program is run sequentially on each vnode one after the other PBS Professional 10 4 User s Guide 177 Chapter 8 Advanced PBS Features V Verbose output about error messages and task exit status is produced When run without the c or the n option pbosdsh will spawn the pro gram on all vnodes allocated to the PBS job The execution take
87. 85 Chapter 11 HPC Basic Profile Jobs The following commands and their API equivalents are not supported for jobs that end up running on the HPCBP Server gqalter qsig qmsg pbsdsh pbs report printjob pbs_ rcp tracejob pbs_rsub pbs_rstat pbs_rdel ghold qrls qrerun 11 8 See Also For a description of how job attributes are translated into JSDL see the PBS Professional External Reference Specification 286 PBS Professional 10 4 User s Guide HPC Basic Profile Jobs Chapter 11 11 8 1 References 1 OGSA High Performance Computing Profile Working Group OGSA HPCP WG of the Open Grid Forum https forge gridforum org sf projects ogsa hpcp wg The HPC Basic Profile specification is GFD 114 http www ogf org documents GFD 114 pdf 2 OGSA High Performance Computing Profile Working Group OGSA HPCP WG of the Open Grid Forum https forge gridforum org sf projects ogsa hpcp wg The HPC File Staging Profile Version 1 0 http forge ogf org sf go doc15024 nav 1 3 OGSA Job Submission Description Language Working Group JSDL WG of the Open Grid Forum http www ogf org gf group_info view php group jsdl wg The JSDL HPC Profile Application Extension Version 1 0 is GFD 111 http www ogf org documents GFD 111 pdf 4 OGSA Usage Record Working Group UR WG of the Open Grid Forum The Usage Record Format Recommendation is GFD 98 http www ogf org documents GFD 98 pdf 5 Network Working Group Uni
88. 9 arrangement 55 Attribute account string 89 priority 7 83 rerunnable 81 attributes modifying 117 315 B Batch job 18 block 158 Boolean Resources 46 Built in Resources 35 C Changing order of jobs 127 Checking status of jobs 136 of queues 140 of server 139 Checkpointable 84 Checkpointing interval 84 job arrays 222 SGI MPI 267 checkpointing 122 chunk 45 CLI 19 Command line interface 19 Commands 11 commands and provisioning 295 comment 145 Common User Environment 7 Computational Grid Support 6 count spec 182 credential 199 Cross System Scheduling 7 CSA 196 csh 23 Custom resources 44 D DCE 198 199 Dedicated Time 196 Default Resources 47 316 Index Deleting job array range 217 job arrays 217 subjob 217 Deleting Jobs 124 dependencies job arrays 221 Deprecations 17 Destination specifying 77 devtype 234 directive 18 29 68 69 70 71 109 176 299 300 Directives 39 directives 36 Display nodes assigned to job 144 non running jobs 144 queue limits 146 running jobs 144 size in gigabytes 144 size in megawords 144 user specific jobs 143 Distributed clustering 7 workload management 9 E Email notification 80 Enterprise wide Resource Sharing 6 euidevice 234 euilib 234 exclusive 56 Executor 11 Exit Status job arrays 223 PBS Professional 10 4 User s Guide F Fairshare job arrays 223 File output 162 output and error 89 rhosts 26 specify name of 78 staging 6 163
89. ED BY LICENSEE USING THE PRODUCTS Ts INDEMNITY Altair will defend and indemnify at its expense any claim made against Licensee based on an allegation that the Software infringes a patent or copyright Claim provided however that this indemnification does not include claims which are based on Supplier soft ware or ISV software and that Licensee has not materially breached the terms of this Agreement Licensee notifies Altair in writing within ten 10 days after Licensee first learns of the Claim and Licensee cooperates fully in the defense of the claim Altair shall have sole control over such defense provided however that it may not enter into any settlement bind ing upon Licensee without Licensee s consent which shall not be unrea sonably withheld Ifa Claim is made Altair may modify the Software to avoid the alleged infringement provided however that such modifications do not materially diminish the Software s functionality If such modifica tions are not commercially reasonable or technically possible Altair may terminate this Agreement and refund to Licensee the prorated license fee that Licensee paid for the then current Term Perpetual licenses shall be pro rated over a 36 month term Altair shall have no obligation under this Section 7 however if the alleged infringement arises from Altair s compli ance with specifications or instructions prescribed by Licensee modifica tion of the Software by Licensee use of th
90. EUED the set of subjobs of the job array which are currently queued Array REMAINING the set of subjobs of the job array which are queued or running Array DONE the set of subjobs of the job array which have finished running 212 PBS Professional 10 4 User s Guide Job Arrays Chapter 9 Table 9 3 PBS Commands Taking Job Arrays as Arguments Argument to Command Com mand Array Array Range Array Index qstat Array Array Range Array Index qdel Array amp Array Range where Array Index Array REMAIN Array REMAINING ING galte Array erroneous erroneous r qorde Array erroneous erroneous r qmove Array amp erroneous erroneous Array QUEUED ghold Array amp erroneous erroneous Array QUEUED qrls Array amp erroneous erroneous Array QUEUED qre Array RUNNING Array Range where Array Index run amp Array DONE Array RUNNING qrun erroneous Array Range where Array Index Array QUEUED trace erroneous erroneous Array Index job qsig Array RUNNING Array Range where Array Index Array RUNNING qmsg erroneous erroneous erroneous PBS Professional 10 4 User s Guide 213 Chapter 9 Job Arrays 9 7 2 qstat Status of a Job Array The qstat command is used to query the status of a Job Array The default output is to list the Job Array in a single line showing the Job Array Identi fier
91. Email Notification The m MailOptions defines the set of conditions under which the execution server will send a mail message about the job The MailOptions argument is a string which consists of either the single character n or one or more of the characters a b and e If no email notification is specified the default behavior will be the same as for m a a send mail when job is aborted by batch system b send mail when job begins execution e send mail when job ends execution n do not send mail Examples qsub m ae my _job PBS m b 4 13 6 Setting Email Recipient List The M user_list option declares the list of users to whom mail is sent by the execution server when it sends mail about the job The user_list argument is of the form user host user host 80 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 If unset the list defaults to the submitting user at the qsub host i e the job owner qsub M userl mydomain com my_job 4 13 6 1 Caveats PBS on Windows can only send email to addresses that specify an actual hostname that accepts port 25 sendmail requests For the above example on Windows you will need to specify qsub M userl host mydomain com where host mydomain com accepts port 25 connections 4 13 7 Specifying a Job Name The N name option declares a name for the job The name specified may be up to and including 15 ch
92. Guide If a job requests grouping by a resource i e place group resource then the chunks are placed as requested and complex wide node grouping is ignored If a job is to use node grouping but the required number of vnodes is not defined in any one group grouping is ignored This behavior is unchanged PBS Professional 10 4 User s Guide 225 Chapter 10 Multiprocessor Jobs 10 2 Submitting SMP Jobs To submit a job which should run on one host and which requires a certain number of cpus and amount of memory submit the job with qsub l select ncpus N mem M l place group host When the job is run the PBS_NODEFILE will contain one entry the name of the selected execution host Generally this is ignored for SMP jobs as all processes in the job are run on the host where the job script is run The job will have two environment variables NCPUS and OMP NUM THREADS set to N the number of CPUs allocated 10 3 Submitting MPI Jobs The preferred method for submitting an MPI job is by specifying one chunk per MPI task For example for a 10 way MPI job with 2gb of mem ory per MPI task you would use qsub 1 select 10 ncpus 1 mem 2gb If you have a cluster of small systems with for example 2 CPUs each and you wish to submit an MPI job that will run on four separate hosts then submit qsub 1 select 4 ncpus 1 l place scatter The PBS_NODEFILE file will contain one entry for each of the hosts allo cated to the job In the e
93. Guide 151 Chapter 7 Checking Job System Status 13 hostl userl workq foojob 1 1 128mb 00 10 Q If the start time for a job cannot be estimated the start time behaves as if it is unset Forqstat T the start time appears as a question mark forqstat f the start time appears as a time in the past 7 2 Viewing Job System Status with xpbs The main display of xpbs shows a brief listing of all selected Servers all queues on those Servers and any jobs in those queues that match the selec tion criteria discussed below Servers are listed in the HOST panel near the top of the display To view detailed information about a given Server i e similar to that pro duced by qstat fB select the Server in question then click the Detail button Likewise for details on a given queue i e similar to that produced by qstat f Q select the queue in question then click its cor responding Detail button The same applies for jobs as well i e qstat f You can view detailed information on any displayed job by selecting it and then clicking on the Detail button Note that the list of jobs dis played will be dependent upon the Selection Criteria currently selected This is discussed in the xpbs portion of the next section 7 3 The qselect Command The qselect command provides a method to list the job identifier of those jobs job arrays or subjobs which meet a list of selection criteria
94. Ifthe requested values for ncpus and mpiprocs are different an error message is logged to the HPCBP MOM log file and the job is rejected So for example if you request qsub 1 select 3 ncpus 2 mem 8gb the job is rejected because no mpiprocs were requested 11 4 1 4 Number of ompthreads For a job with more than one chunk that requests ompthreads each chunk must request the same value for ompthreads Otherwise an error mes sage is logged to the HPCBP MOM log file and the job is rejected 11 4 1 5 Restrictions on Requesting arch Resource Requesting a value for arch in an HPCBP job means requesting a node or nodes with that architecture from among the nodes controlled by the HPCBP Server It is not necessary for a job to request a value for arch An HPCBP job can request any arch value that can be satisfied by the HPCBP Server PBS Professional 10 4 User s Guide 273 Chapter 11 HPC Basic Profile Jobs 11 4 2 Using the qsub Command for HPCBP Jobs Job submission for non HPCBP jobs is unchanged However when you submit an HPCBP job you must do the following e Specify only one executable and its arguments e Specify executable and arguments in the qsub command line 11 4 2 1 qsub Syntax for HPCBP Jobs qsub a date_time A account_string c interval C directive prefix e path h I j oe eo J X Y Z k ole oe 1 resource list m mail_options M user_list N jobname o path p priority q
95. PBS Professional 10 4 USER S GUIDE OR PBS GridWorks Enabling On Demand Computing A division of 4 Altair PBS Professional User s Guide Altair PBS Professional 10 4 Updated 4 22 10 Edited by Anne Urban Copyright 2003 2010 Altair Engineering Inc All rights reserved PBS PBS Works PBS GridWorks PBS Professional PBS Analyt ics PBS Catalyst e Compute and e Render are trademarks of Altair Engineering Inc and are protected under U S and international laws and trea ties All other marks are the property of their respective owners ALTAIR ENGINEERING INC Proprietary and Confidential Contains Trade Secret Information Not for use or disclosure outside ALTAIR and its licensed clients Information contained herein shall not be decompiled disassembled duplicated or disclosed in whole or in part for any purpose Usage of the soft ware is only as explicitly permitted in the end user software license agree ment Copyright notice does not imply publication For more information contact Altair at Web www pbsgridworks com Email pbssales altair com Technical Support Location Telephone e mail North America 1 248 614 2425 pbssupport altair com China 86 0 21 6117 1666 es altair com cn France 33 0 1 4133 0992 francesupport altair com Germany 49 0 7031 6208 22 hwsupport altair de India 91 80 66 29 4500 pbs support india altair c
96. PBS Professional 10 4 User s Guide 57 Chapter 4 Submitting a PBS Job 4 6 3 Resources Allocated from a Vnode The resources allocated from a vnode are only those specified in the job s schedselect This job attribute is created internally by starting with the select specification and applying any server and queue default_chunk resource defaults that are missing from the select statement The schedse lect job attribute contains only vnode level resources The exec_vnode job attribute shows which resources are allocated from which vnodes 4 6 3 1 Resources Assigned to a Job The Resource_List attribute is the list of resources requested via qsub with job wide defaults applied Vnode level resources from Resource List are used in the converted select when the user doesn t specify a select state ment The converted select statement is used to fill in gaps in schedselect Values for ncpus or mem in the job s Resource_List come from three places 1 Resources specified via qsub 2 the sum of the values in the select specification not including default_chunk or 3 resources inherited from queue and or server resources _ default Case 3 applies only when the user does not specify l select but uses Inodes or Incpus instead The Resource_List mem is a job wide memory limit which if memory enforcement is enabled the entire job the sum of all of the job s usage cannot exceed Examples The queue has the following
97. PU node locked licenses 49 per host node locked licenses 49 per use node locked licenses 49 place statement 55 placement sets job arrays 224 POE 232 poe examples 239 Preemption job arrays 223 printjob 20 procs 235 PROFILE PATH 26 Prologues and Epilogues job arrays 222 ProPack 196 provision 289 provisioned vnode 290 provisioning 291 allowing time 296 and commands 295 AOE restrictions 293 host restrictions 293 requesting 294 using AOE 291 vnodes 290 PVM 266 PBS Professional 10 4 User s Guide Index Q qalter 20 106 job array 217 qdel 20 106 job arrays 217 qdisable 20 106 qenable 20 106 qhold 20 106 121 123 job arrays 218 qmgr 20 qmove 20 107 128 job array 218 qmsg 20 106 125 220 qorder 20 107 127 128 job arrays 218 qrerun 20 106 job arrays 219 qrls 20 106 122 123 job arrays 218 qrun 20 106 job array 219 qselect 20 113 114 152 153 job arrays 220 qsig 20 106 126 qstart 20 106 qstat 20 106 120 123 127 128 135 136 137 140 141 142 143 144 145 146 152 153 qstop 20 106 qsub 20 22 68 71 72 75 92 106 158 159 199 Kerberos 198 qsub options 75 qterm 21 106 Queuing xi 9 Quick Start Guide xi R rcp 21 recurrence rule 181 PBS Professional 10 4 User s Guide Releasing a Job Array 218 report 197 requesting provisioning 294 Requeuing a Job Array 219 Reservation deleting 190 reservation advance 178 180 degraded 179 instance 179 Setting start
98. S graphical user interface is called xpbs and provides a user friendly point and click interface to the PBS commands xpbs utilizes the tcl tk graphics tool suite while providing the user with most of the same functionality as the PBS CLI commands In this chapter we introduce xpbs and show how to create a PBS job using xpbs 5 1 Using xpbs 5 1 1 Starting xpbs If PBS is installed on your local workstation or if you are running under Windows you can launch xpbs by double clicking on the xpbs icon on the desktop You can also start xpbs from the command line with the fol lowing command PBS Professional 10 4 User s Guide 93 Chapter 5 Using the xpbs GUI UNIX xpbs amp Windows xpbs exe Doing so will bring up the main xpbs window as shown below 5 1 2 Running xpbs Under UNIX Before running xpbs for the first time under UNIX you may need to con figure your workstation for it Depending on how PBS is installed at your site you may need to allow xpbs to be displayed on your workstation However if the PBS client commands are installed locally on your work station you can skip this step Ask your PBS administrator if you are unsure The most secure method of running xpbs remotely and displaying it on your local XWindows session is to redirect the X Windows traffic through ssh secure shell via setting the X11Forwarding yes parameter in the sshd_config file Your local system administrator can provide details on
99. S reads down the shell script until it finds the first line that is not a valid PBS directive then stops It assumes the rest of the script is the list of commands or tasks that the user wishes to run In this case PBS sees lines 6 8 as being user commands 70 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 The section Job Submission Options on page 75 describes how to use the qsub command to submit PBS jobs Any option that you specify to the qsub command line except 1I can also be provided as a PBS directive inside the PBS script PBS directives come in two types resource require ments and attribute settings In our example above lines 2 3 specify the 1 resource list option fol lowed by a specific resource request Specifically lines 2 3 request 1 hour of wall clock time as a job wide request and 400 megabytes MB of memory in a chunk Line 4 requests that PBS join the stdout and stderr output streams of the job into a single stream Finally lines 6 8 are the command lines for executing the program s we wish to run You can specify as many programs tasks or job steps as you need 4 11 Changing the Job s PBS Directive By default the text string PBS is used by PBS to determine which lines in the job file are PBS directives The leading symbol was chosen because it is a comment delimiter to all shell scripting languages in com mon use on UNIX systems
100. SD Foundation Inc and its contributors as well as software devel oped by the University of California Berkeley and its contributors Other contributors to the NASA version of PBS include Bruce Kelly and Clark Streeter of NERSC Kent Crispin and Terry Heidelberg of LLNL John Kochmar and Rob Pennington of Pittsburgh Supercomputing Center and Dirk Grunwald of University of Colorado Boulder The ports of PBS to the Cray T3e and the IBM SP SMP were funded by DoD USAERDC the port of PBS to the Cray SV1 was funded by DoD MSIC PBS Professional 10 4 User s Guide ix No list of acknowledgements for PBS would possibly be complete without special recognition of the first two beta test sites Thomas Milliman of the Space Sciences Center of the University of New Hampshire was the first beta tester Wendy Lin of Purdue University was the second beta tester and holds the honor of submitting more problem reports than anyone else out side of NASA X PBS Professional 10 4 User s Guide Preface Intended Audience PBS Professional is the professional workload management system from Altair that provides a unified queuing and job management interface to a set of computing resources This document provides the user with the information required to use PBS Professional including creating submit ting and manipulating batch jobs querying status of jobs queues and sys tems and otherwise making effective use of the computer resources under the
101. The maximum number of ranks that can be launched is the number of entries in PBS_NODEFILE pg The use of the pg option for having multiple executables on multiple hosts is allowed but it is up to user to make sure only PBS hosts are specified in the process group file MPI processes spawned on non PBS hosts are not guaranteed to be under the control of PBS 242 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 5 2 Examples Example 1 Run a single executable MPICH GM job with 64 processes spread out across the PBS allocated hosts listed in PBS_NODEFILE PBS NODEFILE pbs host1l pbs host2 pbs host3 qsub 1 select 3 ncpus 1 mpirun ch_gm np 64 path myprog x 1200 D lt job id gt Example 2 Run an MPICH GM job with multiple executables on multiple hosts listed in the process group file procgrp qsub 1 select 3 ncpus 1 echo hostl 1 userl x y a exe argl arg2 gt procgrp echo host2 1 userl x x b exe argl arg2 gt gt procgrp mpirun ch_gm pg procgrp path mypro x rm f procgrp D lt job id gt When the job runs mpirun ch_gm will give this warning message warning pg is allowed but it is up to user to make sure only PBS hosts are specified MPI processes spawned are not guaranteed to be under the control of PBS The warning is issued because if any of the hosts listed in procgrp are not under the control of PBS then the processes on those hosts will PBS Professi
102. Under PBS nnnaanannaaaan 22 3 11 Setting Up Your UNIX Linux Environment 22 3 12 Setting Up Your Windows Environment 24 3 13 Environment Variables ona ceed dea ews 28 3 14 Temporary Scratch Space TMPDIR 29 Submitting a PBS Job eal 4 1 Vnodes Virtual Nodes n a cand disexwss 31 4 2 PBS Resources nnuun nunana 32 A PIS MOOS ie ae nee E E EE E ee 35 4 4 Submitting a PBS Job nananana aaaea 40 4 5 Requesting Resources ic 100cee van eneeedeuds 43 4 6 Placing Jobs on Vnodes 2 4 viewed sent saa ee ase 55 4 7 Submitting Jobs Using Select amp Place Examples 59 4 8 Backward Compatibility 65 4 9 How PBS Parses a Job Script 0 68 4 10 A Sample PBS Jobs 2c ccsc sccasddeeveuvaxes 69 4 11 Changing the Job s PBS Directive 71 4 12 Windows Jobs oaaaanaana nanana 72 4 13 Job Submission OPGONS is cs des ga vawn chun 75 4 14 Failed Jobs lt 2 tetce Gis ce tcateaadw eke SEEK eam 92 Using the xpbs GUI 93 5 1 Using xpbs eo ee s 93 5 2 Using xpbs Definitions of Terms 95 5 3 Introducing the xpbs Main Display 96 5 4 Setting xpbs Preferences anananananan 105 PBS Professional 10 4 User s Guide Table of Contents 5 5 Relationship Between PBS and xpbs 106 5 6 How to Submit a Job Using xpbs 107 5 7 ENS MODs ser oo ban Oks oe Rees Erp te ok 111 5 8 The xpbs Con
103. User s Guide Checking Job System Status Chapter 7 until you change them again If you exit xpbs you will be prompted if you wish to save your configuration information this includes the job selection criteria 7 5 Using xpbs TrackJob Feature The xpbs command includes a feature that allows you to track the progress of your jobs When you enable the Track Job feature xpbs will monitor your jobs looking for the output files that signal completion of the job The Track Job button will flash red on the xpbs main display and if you then click it xpbs will display a list of all completed jobs that you were previously tracking Selecting one of those jobs will launch a win dow containing the standard output and standard error files associated with the job IMPORTANT The Track Job feature is not currently available on Win dows To enable xpbs job tracking click on the Track Job button at the top cen ter of the main xpbs display Doing so will bring up the Track Job dialog box shown below PBS Professional 10 4 User s Guide 155 Chapter 7 Checking Job System Status From this window you can name the users whose jobs you wish to monitor You also need to specify where you expect the output files to be either local or remote e g will the files be retained on the Server host or did you request them to be delivered to another host Next click the start reset tracking button and then the close window button Note
104. aX and extrax and makes an output file newdataxX as well as echoing which iteration it is on The datax and extrax files will be staged from inputs to the staging and exe cution directory then newdataX will be staged from the staging and execution directory to outputs bin sh PBS N StagingExample PBS J 1 2 PBS W stagein data array index host1 homedir inputs data array index extra array_ index host1 homedir inputs extra array_index PBS W stageout newdata array index host1 homedir outputs newdata array_index echo Main script index PBS ARRAY INDEX cat dataSPBS ARRAY INDEX extra PBS ARRAY INDEX gt gt newdata PBS ARRAY INDEX Local path execution directory created by PBS we don t know the name Remote host data storage host hostl Remote path for inputs original data files dataX and extrax homedir inputs Remote path for results output of computation newdatax homedir outputs 210 PBS Professional 10 4 User s Guide Job Arrays Chapter 9 StageScript resides in homedir testdir In that directory we can run it by typing qsub StageScript It will run in the staging and execution directory created by PBS See sec tion 8 6 Input Output File Staging on page 163 9 6 1 2 Output Filenames The name of the job array will default to the script name if no name is given via qsub N For example if the sequence number were 1234 PBS N fixgamma would
105. able PBS_JOBDTIR is set to the pathname of the staging and execution directory on the primary execution host PBS _JOBDIR is added to the job script process any job tasks and the prologue and epilogue 8 6 3 Attributes and Environment Variables Affecting Staging The following attributes and environment variables affect staging and exe cution Table 8 3 Attributes and Environment Variables Affecting Staging Job s Attribute or Environment Effect Variable sandbox attribute Determines whether PBS uses user s home direc tory or creates job specific directory for staging and execution User settable per job via qsub W or through a PBS directive stagein attribute Sets list of files or directories to be staged in User settable per job via qsub W or through a PBS directive stageout Sets list of files or directories to be staged out attribute User settable per job via qsub W or through a PBS directive Keep Files Determines whether output and or error files attribute remain on execution host User settable per job via qsub k or through a PBS directive If the Keep Files attribute is set to o and or e out put and or error files remain in the staging and execution directory and the job s sandbox attribute is set to PRIVATE standard out and or error files are removed when the staging direc tory is removed at job end along with its con tents 166 PBS Professional 10 4 User s Gui
106. ach with 2 CPUs and 4 threads qsub l1 select 4 ncpus 2 ompthreads 4 To request 16 MPI processes each with 2 threads on machines with 2 pro cessors qsub 1 select 16 ncpus 2 To request two chunks each with 8 CPUs and 8 MPI tasks and four threads qsub l select 2 ncpus 8 mpiprocs 8 ompthreads 4 Example qsub 1 select 4 ncpus 2 This request is satisfied by 4 CPUs from VnodeA 2 from VnodeB and 2 from VnodeC so the following is written to the PBS_NODEFILE VnodeA VnodeA VnodeB Vnodec PBS Professional 10 4 User s Guide 229 Chapter 10 Multiprocessor Jobs The OpenMP environment variables are set for the 4 PBS tasks corre sponding to the 4 MPI processes as follows e For PBS task 1 on VnodeA OMP_NUM_THREADS 2 NCPUS 2 e For PBS task 2 on VnodeA OMP_NUM_THREADS 2 NCPUS 2 e For PBS task 3 on VnodeB OMP_NUM_THREADS 2 NCPUS 2 e For PBS task 4 on VnodeC OMP_NUM_THREADS 2 NCPUS 2 Example qsub l select 3 ncpus 2 mpiprocs 2 ompthreads 1 This is satisfied by 2 CPUs from each of three vnodes VnodeA VnodeB and VnodeC so the following is written to the PBS_ VNODEFILE VnodeA VnodeA VnodeB VnodeB Vnodec Vnodec The OpenMP environment variables are set for the 6 PBS tasks corre sponding to the 6 MPI processes as follows e For PBS task 1 on VnodeA OMP NUM THREADS 1 NCPUS 1 e For PBS task 2 on VnodeA OMP NUM THREADS 1 NCPUS 1 e For PBS task 3 on VnodeB OMP_NUM_THREADS 1 NCPUS 1 e For PBS tas
107. aging and execution directory and with PBS_JOBDIR and TMPDIR set in their environment 8 6 7 7 Standard Out and Standard Error The job s stdout and stderr files are created directly in the job s staging and execution directory on the primary execution host 8 6 7 7 1 Job specific Staging and Execution Directories Ifthe qsub k option is used the stdout and stderr files will not be auto matically copied out of the staging and execution directory at job end they will be deleted when the directory is automatically removed 174 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 8 6 7 7 2 User s Home Directory as Staging and Execution Directory If the k option to qsub is used standard out and or standard error files are retained on the primary execution host instead of being returned to the submission host and are not deleted after job end 8 6 7 8 Running the Epilogue PBS runs the epilogue on the primary host as root The epilogue is exe cuted with its current working directory set to the job s staging and execu tion directory and with PBS_ JOBDIR and TMPDTR set in its environment 8 6 7 9 Staging Files Out and Removing Execution Directory When PBS stages files out it evaluates local _path and remote path relative to PBS JOBDIR Files that cannot be staged out are saved in PBS HOME undelivered See section 12 5 6 Non delivery of Output on page 617 of the PBS Professional Administrator s Gui
108. al mpirun to decide what the reasonable default value should be which is usually 1 The maximum number of ranks that can be launched is the number of entries in PBS_NODEFILE 10 6 9 2 MPD Startup and Shutdown The interface ensures that the MPD daemons are started on each of the hosts listed in the PBS_NODEFILE It also ensures that the MPD dae mons are shut down at the end of MPI job execution PBS Professional 10 4 User s Guide 253 Chapter 10 Multiprocessor Jobs 10 6 9 3 Examples Example 1 Runa single executable MPICH2 job with 6 processes spread out across the PBS allocated hosts listed in PBS_NODEFILE PBS NODEFILE pbs hostl pbs host2 pbs host3 pbs hostl pbs host2 pbs host3 Job script mpirun runs 6 processes mapped to each host listed in PBS NODEFILE mpirun np 6 path myprog x 1200 Run job script qsub l select 3 ncpus 2 job script lt job id gt Example 2 Run an MPICH2 job with multiple executables on multiple hosts using PBS_NODEFILE and mpiexec arguments in mpirun PBS NODEFILE hostA hostA hostB hostB hostc hostc 254 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 Job script PBS 1 select 3 ncpus 2 mpirun np 2 tmp mpitestl np 2 tmp mpitest2 np 2 tmp mpitest3 Run job qsub job script Example 3 Run an MPICH2 job with multiple executables on multiple hosts using mpirun configfile option and PBS_NODEFILE PBS NODEFILE hostA hos
109. annot be moved in the H state if any subjobs are running Checkpointing is not supported for job arrays Even on systems that sup port checkpointing no subjobs will be checkpointed they will run to completion Similarly the qr1s command releases a hold on a job However the user executing the qr 1s command must have the necessary privilege to release a given hold The same rules apply for releasing a hold as exist for setting a hold The qrls command can only be used with job array objects not with sub jobs or ranges The job array will be returned to its pre hold state which can be either Q B or W The usage syntax of the qr 1s command is qrls h hold list job_identifier For job arrays the job identifier must be enclosed in double quotes 122 PBS Professional 10 4 User s Guide Working With PBS Jobs Chapter 6 The following examples illustrate how to use both the ghold and qrls commands Notice that the state S column shows how the state of the job changes with the use of these two commands qstat a 54 Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time 54 south barry workq engine 1 0 20 Q qhold 54 qstat a 54 Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time 54 south barry workq engine 1 0 20 H qrls h u 54 qstat a 54 Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time 54 south barry workq
110. ans that the reservation will not con flict with currently running jobs other confirmed reservations or dedicated time and that the requested resources are available for the reservation A reservation request that fails these tests is rejected All occurrences of a standing reservation must be acceptable in order for the standing reserva tion to be confirmed The pbs_rsub command returns a reservation ID which is the reserva tion name For an advance reservation this reservation ID has the format R lt unique integer gt lt server name gt For a standing reservation this reservation ID refers to the entire series and has the format S lt unique integer gt lt server name gt The user specifies the resources for a reservation using the same syntax as for a job Jobs in reservations are placed the same way non reservation jobs are placed in placement sets The xpbs GUI cannot be used for creation querying or deletion of reser vations The time for which a reservation is requested is in the time zone at the sub mission host 8 8 3 Creating Advance Reservations You create an advance reservation using the pbs_rsub command PBS must be able to calculate the start and end times of the reservation so you must specify two of the following three options 180 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 D Duration E End time R Start time 8 8 3 1 Examples of Creating Advance Reservations The
111. apter 4 Each kind of resource plays a specific role which is either inside chunks or outside of them but not both Some resources e g ncpus can only be used at the host chunk level The rest e g walltime can only be used at the job wide level Therefore no resource can be requested both inside and outside of a selection statement Keep in mind that requesting for exam ple Incpus is the old form which cannot be mixed with the new form 4 5 1 Allocation Resources are allocated to jobs both because jobs explicitly request them and because specified default resources are applied to jobs Jobs explicitly request resources either at the vnode level in chunks defined in a selection statement or in job wide resource requests outside of a selection state ment An explicit resource request can appear in the following in order of precedence 1 qalter 2 qsub 3 PBS job script directives 4 5 2 Requesting Resources in Chunks A chunk declares the value of each resource in a set of resources which are to be allocated as a unit to a job It is the smallest set of resources that will be allocated to a job All of a chunk must be taken from a single host A chunk request is a vnode level request Chunks are described in a selec tion statement which specifies how many of each kind of chunk A selec tion statement has this form 1 select N chunk N chunk If N is not specified it is taken to be 1 No spaces are allowed b
112. aracters in length It must consist of print able non whitespace characters with the first character alphabetic or numeric and contain no special characters If the N option is not speci fied the job name will be the base name of the job script file specified on the command line If no script file name was specified and the script was read from the standard input then the job name will be set to STDIN qsub N myName my_job PBS N myName 4 13 8 Marking a Job as Rerunnable or Not The r y n option declares whether the job is rerunnable To rerun a job is to terminate the job and requeue it in the execution queue in which the job currently resides The va ue argument is a single character either 66 99 oo 99 y or n If the argument is y the job is rerunnable If the argument is 66 29 n the job is not rerunnable The default value is y rerunnable qsub r n my_job PBS r n PBS Professional 10 4 User s Guide 1 Chapter 4 Submitting a PBS Job Marking your job as non rerunnable will not affect how PBS treats it in the case of startup failure If a job that is marked non rerunnable has an error during startup before it begins execution that job is requeued for another attempt The purpose of marking a job as non rerunnable is to prevent it from running twice and using data that undergoes a change during execu tion However if the job never actually starts execution the data is
113. are only scheduled to run from the reservation once the reservation period arrives The jobs in a reservation are not allowed to use in aggregate more resources than the reservation requested A reservation job is started only if its requested walltime will fit within the reservation period So for example if the reservation runs from 10 00 to 11 00 and the job s walltime is 4 hours the job will not be started When an advance reservation ends any running or queued jobs in that res ervation are deleted PBS Professional 10 4 User s Guide 191 Chapter 8 Advanced PBS Features When an occurrence of a standing reservation ends any running jobs in that reservation are killed Any jobs still queued for that reservation are kept in the queued state They are allowed to run in future occurrences When the last occurrence of a standing reservation ends all jobs remaining in the reservation are deleted whether queued or running A job in a reservation cannot be preempted 8 8 8 1 1 Reservation Fault Tolerance If one or more vnodes allocated to an advance reservation or to the soonest occurrence of a standing reservation become unavailable the reservation s state becomes DG or RESV_DEGRADED A degraded reservation does not have all the reserved resources to run its jobs PBS attempts to reconfirm degraded reservations This means that it looks for alternate available vnodes on which to run the reservation The reser vation s re
114. ation job is pbs_rsub l walltime time W qmove job_identifier Example pbs_rsub W qmove 54 pbs_rsub W qmove 1234 server The R and E options to pbs_ rsub are disabled when using the W qmove option For more information see Advance and Standing Reservation of Resources on page 178 and the pbs_rsub 1B qsub 1B and qmove 1B manual pages A job s default walltime is 5 years Therefore an ASAP reservation s start time can be in 5 years if all the jobs in the system have the default wall time You cannot use the pbs_ rsub command or any other command to request a custom resource which has been created to be invisible or unre questable See section 4 5 14 Resource Permissions on page 54 130 PBS Professional 10 4 User s Guide Working With PBS Jobs Chapter 6 6 9 Using Job History Information 6 9 1 Introduction PBS Professional can provide job history information including what the submission parameters were whether the job started execution whether execution succeeded whether staging out of results succeeded and which resources were used PBS can keep job history for jobs which have finished execution were deleted or were moved to another server 6 9 2 Definitions Moved jobs Jobs which were moved to another server Finished jobs Jobs whose execution is done for any reason e Jobs which finished execution successfully and exited e Jobs terminated by PBS while running e Job
115. atures 8 4 1 Job Array Dependencies Job dependencies are supported e Between jobs and jobs e Between job arrays and job arrays e Between job arrays and jobs e Between jobs and job arrays Note Job dependencies are not supported for subjobs or ranges of subjobs 8 5 Delivery of Output Files To transfer output files or to transfer staged in or staged out files to from a remote destination PBS uses either rcp or scp depending on the configu ration options The version of rcp used by PBS always exits with a non zero exit status for any error Thus MOM knows if the file was delivered or not The secure copy program scp is also based on this version of rcp and exits with the proper exit status If using rcp the copy of output or staged files can fail for at least two reasons e The user lacks authorization to access the specified system See discus sion in User s PBS Environment on page 21 e Under UNIX if the user s cshrc outputs any characters to standard output e g contains an echo command the copy will fail If using Secure Copy scp then PBS will first try to deliver output or stagein out files using scp If scp fails PBS will try again using rcp assuming that scp might not exist on the remote host If rcp also fails the above cycle will be repeated after a delay in case the problem is caused by a temporary network problem All failures are logged in MOM s log and an email containing the error
116. ault bin pbs python i lt return gt gt gt print hello hello 4 3 4 Job Script Names It is recommended to avoid using special characters in a job script name If you must use them on UNIX Linux you must escape them using the back slash character 4 4 Submitting a PBS Job There are a few ways to submit a PBS job using the command line The first is to create a job script and submit it using qsub 4 4 1 Submitting a Job Script For example with job script myjob the user can submit it by typing qsub myjob 16387 f00 exampledomain PBS returns a job identifier e g 16387 f00 exampledomain in the example above Its format will be sequence number servername 40 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 or for a job array sequence number servername domain You ll need the job identifier for any actions involving the job such as checking job status modifying the job tracking the job or deleting the job If my_job contains the following the user is naming the job testjob and running a program called myprogram bin sh PBS N testjob myprogram The largest possible job ID is the 7 digit number 9 999 999 After this has been reached job IDs start again at zero 4 4 1 1 Overriding Directives PBS directives in a script can be overridden by using the equivalent options to qsub For example to override the PBS dir
117. b history is still being preserved The x option to the qselect command allows you to list job identifiers for all jobs whether they are running queued finished or moved The H option to the gselect command allows you to list job identifiers for fin ished or moved jobs only See section 7 3 The qselect Command on page 152 6 9 6 1 Listing Jobs by Time Attributes You can use the gselect command to list queued running finished and moved jobs job arrays and subjobs according to their time attributes The t option to the qselect command allows you to specify how you want to select based on time attributes You can also use the t option twice to bracket a time period See section 7 3 The qselect Command on page 152 Example 1 Select jobs with end time between noon and 3PM qselect te gt 09251200 te 1t 09251500 Example 2 Select finished and moved jobs with start time between noon and 3PM qselect x s MF ts gt 09251200 ts 1t 09251500 Example 3 Select all jobs with creation time between noon and 3PM qselect x tc gt 09251200 tc 1t 09251500 Example 4 Select all jobs including finished and moved jobs with qtime of 2 30PM default relation is eq qselect x tq09251430 134 PBS Professional 10 4 User s Guide Chapter 7 Checking Job System Status This chapter introduces several PBS commands useful for checking status of jobs queues and PBS Servers Examples for use are included as are in
118. ble 5 1 xpbs Server Column Headings Heading Meaning Max Maximum number of jobs permitted Tot Count of jobs currently enqueued in any state Que Count of jobs in the Queued state Run Count of jobs in the Running state Hld Count of jobs in the Held state Wat Count of jobs in the Waiting state Tm Count of jobs in the Transiting state Ext Count of jobs in the Exiting state Status Status of the corresponding Server PEsInUse Count of Processing Elements CPUs PEs Vnodes in Use PBS Professional 10 4 User s Guide 99 Chapter 5 Using the xpbs GUI 5 3 3 xpbs Queues Panel The Queues panel is composed of a leading horizontal QUEUES bar a list box and a set of command buttons The QUEUES bar lists the hosts that are consulted when listing queues the bar also contains a minimize maxi mize button for displaying or iconizing the Queues panel The listbox dis plays information about queues managed by the Server host s selected from the Hosts panel each listbox entry can be selected as described above for the Hosts panel To the right of the Queues Panel area are buttons for actions that can be performed on selected queue s detail provides information about selected queue s This func tionality can also be achieved by double clicking on a Queue listbox entry stop for stopping the selected queue s admin only start for starting the selected queue s admin only disable for disabling
119. brought into focus with a left click To manipulate this widget simply type in the text Use of arrow keys backspace delete key mouse selection of text for deletion or overwrite and copying and pasting with sole use of mouse buttons are per mitted This widget has a scrollbar for vertically scanning a long entry 5 3 Introducing the xpbs Main Display The main window or display of xpbs is comprised of five collapsible sub windows or panels Each panel contains specific information Top to bot tom these panels are the Menu Bar Hosts panel Queues panel Jobs panel and the Info panel 5 3 1 xpbs Menu Bar The Menu Bar is composed of a row of command buttons that signal some action with a click of the left mouse button The buttons are Manual Update forces an update of the information on hosts queues and jobs Auto Update sets an automatic update of information every user specified number of minutes Track Job for periodically checking for returned output files of jobs 96 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 Preferences for setting parameters such as the list of Server host s to query Help contains some help information 03 08 05 17 56 23 usr local pbs 1ib xpbs pbs_bin xpbs_datadump t 30 u agu dhcp115 done xpbs_datadump alarm ti About gives general information about the xpbs GUI Close for exiting xpbs plus saving the current setup information PBS Profess
120. ccounting report produced the user issues the command ja lt options gt where the options specify that a report will be written and what kind To end user job accounting the user issues the command ja t the t option can be included in the previous set of options See the manpage on ja for details The starting and ending ja commands must be used before and after any other commands the user wishes to monitor Here are examples of com mand line and a script On the command line qsub N myjobname l ncpus 1 ja myrawfile sleep 50 ja c gt myreport ja t myrawfile ctrl D Accounting data for the user s job sleep 50 is written to myreport If the user creates a file foo with these commands PBS N myjobname PBS 1 ncpus 1 ja myrawfile sleep 50 ja c gt myreport ja t myrawfile The user could run this script via qsub qsub foo This does the same thing via the script foo PBS Professional 10 4 User s Guide 197 Chapter 8 Advanced PBS Features 8 11 Running PBS in a UNIX DCE Environment PBS Professional includes optional support for UNIX based DCE By optional we mean that the customer may acquire a copy of PBS Profes sional with the standard security and authentication module replaced with the DCE module There are two W options available with qsub which will enable a dcel ogin context to be set up for the job when it eventually executes The user may specify either an encrypted password
121. cmd echo prog2 gt gt tmp poe cmd poe cmdfile tmp poe cmd euilib us rm tmp poe cmd qsub 1 select 3 ncpus 2 mpiprocs 2 l place scatter job script 10 6 4 PBS MPI Jobs on HP UX and Linux PBS is tightly integrated with the mpirun command on HP UX so that resources can be tracked and processes managed When running a PBS MPI job you can use the same arguments to the mpirun command as you PBS Professional 10 4 User s Guide 241 Chapter 10 Multiprocessor Jobs would outside of PBS The h host and 1 user options will be ignored and the np number option will be modified to fit the available resources 10 6 5 PBS Jobs with MPICH GM s mpirun Using rsh ssh mpirun ch_gm PBS provides an interface to MPICH GM s mpirun using rsh ssh If executed inside a PBS job this lets PBS track all MPICH GM processes started via rsh ssh so that PBS can perform accounting and have com plete job control If executed outside of a PBS job it behaves exactly as if standard mpirun ch_gmhad been used You use the same command as you would use outside of PBS either mpirun ch_gm or mpirun 10 6 5 1 Options Inside a PBS job script all of the options to the PBS interface are the same asmpirun ch_gmexcept for the following machinefile lt file gt The file argument contents are ignored and replaced by the contents of the PBS_NODEFILE np If not specified the number of entries found in the S PBS_ NODEFILE is used
122. cution name pbs_server All commands and the other daemons services communicate with the Server via an Internet Protocol IP network The Server s main function is to provide the basic batch services such as receiving creating a batch job modifying the job and run ning the job Normally there is one Server managing a given set of resources However if the Server Failover fea ture is enabled there will be two Servers Job Executor MOM The Job Executor or MOM is the daemon service which actually places the job into execution This process pbs_mom is informally called MOM as it is the mother of all executing jobs MOM is a reverse engineered acronym that stands for Machine Oriented Mini server MOM places a job into execution when it receives a copy of the job from a Server MOM creates a new session that is as identical to a user login session as is possible For example under UNIX if the user s login shell is csh then MOM creates a session in which login is run as well as cshrc MOM also has the responsibility for returning the job s output to the user when directed to do so by the Server One MOM runs on each computer which will execute PBS jobs PBS Professional 10 4 User s Guide 11 Chapter 2 Concepts and Components Scheduler The Job Scheduler daemon service pbs_ sched implements the site s policy controlling when each job is run and on which resources The Scheduler communicates with the var ious MOMs to q
123. d accurate usage data or to set system wide resource priorities As a result many computing resources are left under utilized while others are over utilized At the same time users are confronted with an ever expanding array of operating sys tems and platforms Each year scientists engineers designers and ana lysts must waste countless hours learning the nuances of different computing environments rather than being able to focus on their core pri orities PBS Professional addresses these problems for computing inten sive industries such as science engineering finance and entertainment Now you can use the power of PBS Professional to better control your computing resources This allows you to unlock the potential in the valu able assets you already have while at the same time reducing dependency on system administrators and operators freeing them to focus on other actives PBS Professional can also help you effectively manage growth by tracking real usage levels across your systems and enhancing utilization of future purchases PBS Professional 10 4 User s Guide 3 Chapter 1 Introduction 1 4 History of PBS In the past UNIX systems were used in a completely interactive manner Background jobs were just processes with their input disconnected from the terminal However as UNIX moved onto larger and larger machines the need to be able to schedule tasks based on available resources increased in importance The advent of ne
124. d SOAP Header to Header MOM the get factory attributes docu ment request message Problem with reading the HPCBP unable to parse node status infor node status information MOM mation received from the HPC Basic Profile Server HPC Basic Profile Server HPCBP unable to connect to the HPCBP Connection MOM web service endpoint mpiprocs related error unequal ncpus amp HPCBP can t submit job to the HPC mpiprocs MOM Basic Profile Server total num ber of ncpus and mpiprocs requested are not equal ompthreads error ompthreads are not HPCBP can t submit job to the HPC equal across chunks MOM Basic Profile Server number of ompthreads are not equal in multi chunk job request Generic Problems No reply from HPCBP HPCBP unable to receive response from Server MOM hpcbp web service endpoint OpenSSL library issues Cannot find OpenSSL HPCBP unable to find openssl libraries libraries on system MOM on the system PBS Professional 10 4 User s Guide 283 Chapter 11 HPC Basic Profile Jobs 11 6 5 Job State Transition Log Messages See the following table for a list of the job transitions in the HPCBP Server and the associated actions by the HPCBP MOM Table 11 5 Job Transitions in HPCBP Server and Associated Actions by HPCBP MOM Job Transitions in HPC Basic Profile Server Message Logged By HPCBP MOM Start State End State Pending Running job transitioned from pending to runnin
125. d as you would use outside of PBS either mpirun mpd or mpirun If the MPD daemons are not already running the PBS interface will take care of starting them for you 10 6 7 1 Options Inside a PBS job script all of the options to the PBS interface are the same as mpirun mpd with MPD except for the following m lt file gt The file argument contents are ignored and replaced by the contents of the S PBS_NODEFILE np If not specified the number of entries found in the PBS_ NODEFILE is used The maximum number of ranks that can be launched is the number of entries in PBS_NODEFILE pg The use of the pg option for having multiple executables on multiple hosts is allowed but it is up to user to make sure only PBS hosts are specified in the process group file MPI processes spawned on non PBS hosts are not guaranteed to be under the control of PBS 246 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 7 2 MPD Startup and Shutdown The script starts MPD daemons on each of the unique hosts listed in PBS NODEFILE using either the rsh or ssh method based on the value of the environment variable RSHCOMMAND The default is rsh The script also takes care of shutting down the MPD daemons at the end of a run If the MPD daemons are not running the PBS interface to mpirun mpd will start GM s MPD daemons as this user on the allocated PBS hosts The MPD daemons may have been started already by the
126. de 8 6 7 9 1 Job specific Staging and Execution Directories If PBS created job specific staging and execution directories for the job it cleans up at the end of the job The staging and execution directory and all of its contents are removed on all execution hosts 8 6 7 10 Removing TMPDIRs PBS removes all TMPDIRs along with their contents 8 6 8 Staging with Job Arrays File staging is supported for job arrays See File Staging on page 206 PBS Professional 10 4 User s Guide 175 Chapter 8 Advanced PBS Features 8 6 9 Using xpbs for File Staging Using xpbs to set up file staging directives may be easier than using the command line On the Submit Job window in the miscellany options sec tion far left center of window click on the file staging button This will launch the File Staging dialog box shown below in which you will be able to set up the file staging you desire The File Selection Box will be initialized with your current working direc tory If you wish to select a different directory double click on its name and xpbs will list the contents of the new directory in the File Selection Box When the correct directory is displayed simply click on the name of the file you wish to stage in or out Its name will be written in the File Selected area Next click either of the Add file selected buttons to add the named file to the stagein or stageout list Doing so will write the file name into the corre
127. de Advanced PBS Features Chapter 8 Table 8 3 Attributes and Environment Variables Affecting Staging Job s Attribute or Environment Effect Variable jobdir attribute Set to pathname of staging and execution direc tory on primary execution host Read only viewable viaqstat f PBS JOBDIR Set to pathname of staging and execution direc environment vari tory on primary execution host Added to envi able ronments of job script process job tasks and prologue and epilogue TMPDIR environ Location of job specific scratch directory ment variable 8 6 4 Specifying Files To Be Staged In or Staged Out You can specify files to be staged in before the job runs and staged out after the job runs by using Ww stagein file list and W stage out file list Youcan use these as options to qsub or as directives in the job script The file_list takes the form local_path hostname remote_path for both stagein and stageout The name ocal_path is the name of the file in the job s staging and execu tion directory on the execution host The local_path can be relative to the job s staging and execution directory or it can be an absolute path The character separates the local specification from the remote specifi cation PBS Professional 10 4 User s Guide 167 Chapter 8 Advanced PBS Features The name remote_path is the file name on the host specified by hostname For stagein this is th
128. de 87 Chapter 4 Submitting a PBS Job Usernames containing spaces are allowed as long as the username length is no more than 15 characters and the usernames are quoted when used in the command line Table 4 4 Requirements for Admin User to Submit Job Location Action Submission host username Server host username Different UserS Same UserS UserS seek PROFILE PATH For UserS on ServerA For UserA on Serv thosts contains add lt HostS gt UserS a add lt HostS gt UserS set ServerA s qmgr gt set server qmgr gt set server acl_roots UserS acl_roots UserA acl roots attribute Table 4 5 Requirements for Non admin User to Submit Job File Submission host username Server host username Same UserS UserS Different UserS UserA hosts equiv on Serv lt HostS gt lt HostS gt UserS erA 4 13 15 Specifying Job Group ID The W group_list g_list option defines the group name under which the job is to run on the execution system The g list argument is of the form group host group host Only one group name may be given per specified host Only one of the group specifications may be supplied without the corresponding host spec ification That group name will used for execution on any host not named in the argument list If not set the group_list defaults to the primary group 88 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4
129. declares which signal is sent to the job The signal argument is either a signal name e g SIGKILL the signal name without the STG prefix e g KILL or an unsigned signal number e g 9 The signal name SIGNULL is allowed the Server will send the signal 0 to the job which will have no effect Not all signal names will be recognized by qsig If it doesn t recognize the signal name try issuing the signal number instead The request to signal a batch job will be rejected if e The user is not authorized to signal the job e The job is not in the running state e The requested signal is not supported by the execution host e The job is exiting Two special signal names suspend and resume note all lower case are used to suspend and resume jobs When suspended a job continues to occupy system resources but is not executing and is not charged for wall time Manager or operator privilege is required to suspend or resume a job The three examples below all send a signal 9 SIGKILL to job 34 qsig s SIGKILL 34 qsig s KILL 34 126 PBS Professional 10 4 User s Guide Working With PBS Jobs Chapter 6 IMPORTANT On most UNIX systems the command kill 1 that s minus ell will list all the available signals To send a signal to a job using xpbs first select the job s of interest then click the signal button Doing so will launch the Signal Running Job dialog box From this windo
130. default always copies the standard output stdout and standard error stderr files back to PBS_O WORKDIR on the submission host when a job finishes When qsub is run it sets PBS_O _WORKDIR to the current working directory where the qsub command is executed The o path and e path options to qsub allows you to specify the name of the files to which the stdout and the stderr file streams should be written The path argument is of the form host name path_name where hostname is the name of a host to which the file will be returned and path_name is the path name on that host You may specify relative or absolute paths If you specify only a file name it is assumed to be relative to your home directory Do not use variables in the path The following examples illustrate these various options PBS o u userl myOutputFile PBS e u userl myErrorFile qsub o myOutputFile my_job qsub o u userl myOutputFile my_job qsub o myWorkstation u userl myOutputFile my_job qsub e myErrorFile my_job qsub e u userl myErrorFile my_job qsub e myWorkstation u userl myErrorFile my_job 78 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 Note that if the PBS client commands are used on a Windows host then special characters like spaces backslashes and colons can be used in command line arguments such as for specifying pathnames as well as drive letter specifications The following are allowed
131. default you will get email when the reservation is termi nated or confirmed If you want to receive email about events other than those set the reservation s Mail Points attribute via the m lt mail events gt option For more information see the pbs_rsub 1B and pbs_resv_attributes 7B man pages 8 8 6 Viewing the Status of a Reservation The following table shows the list of possible states for a reservation The states that you will usually see are CO UN BD and RN although a reser vation usually remains unconfirmed for too short a time to see that state See Reservation States on page 442 of the PBS Professional Reference Guide To view the status of a reservation use the pbs_rstat command It will display the status of all reservations at the PBS server For a standing res ervation the pbs_rstat command will display the status of the soonest occurrence Duration is shown in seconds The pbs_rstat command will not display a custom resource which has been created to be invisible See section 4 5 14 Resource Permissions on page 54 This command has three options Table 8 4 Options to pbs_rstat Command Option Meaning Description B Brief Lists only the names of the reservations 186 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 Table 8 4 Options to pbs_rstat Command Option Meaning Description Short Lists in table format the name queue name own
132. e workq server south Checkpoint u ctime Thu Aug 23 10 11 09 2004 Error Path south u susan tns3d e89 exec_host south 0 Hold Types n Join Path oe Keep Files n Mail Points a mtime Thu Aug 23 10 41 07 2004 Output_Path south u susan tns3d 089 Priority 0 qtime Thu Aug 23 10 11 09 2004 Rerunnable True Resource List mem 300mb Resource List ncpus 1 Resource List walltime 00 20 00 session_id 2083 PBS Professional 10 4 User s Guide Checking Job System Status Chapter 7 Variable List PBS_O HOME u susan PBS O LANG en US PBS O LOGNAME susan PBS O PATH bin usr bin PBS O SHELL bin csh PBS O HOST south PBS _O WORKDIR u susan PBS O SYSTEM Linux PBS_O QUEUE workg euser susan egroup myegroup queue _type E comment Job run on host south started at 10 41 etime Thu Aug 23 10 11 09 2004 7 1 6 List User Specific Jobs The u option to gst at displays jobs owned by any of a list of user names specified The syntax of the list of users is user_name host user_name host Host names are not required and may be wild carded on the left end e g mydomain com user_name without a host is equivalent to user_name that is at any host qstat u userl Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time 16 south userl workq aims14 1 0 01 H 18 south userl workq aims14 1 0 01 W 52
133. e Software in combination with other software not provided by Altair and which use is not specifically described in the Documentation and if Licensee is not using the most cur rent version of the Software if such alleged infringement would not have occurred except for such exclusions listed here This section 7 states Altair s entire liability to Licensee in the event a Claim is made No indem nification is made for Supplier and or ISV Software 308 PBS Professional 10 4 User s Guide 8 LIMITATION OF REMEDIES AND LIABILITY Licensee s exclusive remedy and Altair s sole liability for Software that does not meet the warranty set forth in Section 6 shall be at Altair s option either 1 to correct the nonconforming Software within a reasonable time so that it conforms to the warranty or 11 to terminate this Agreement and refund to Licensee the license fees that Licensee has paid for the then current Term for the nonconforming Software provided however that Licensee notifies Altair of the problem in writing within the applicable Warranty Period when the problem first occurs Any corrected Software shall be warranted in accordance with Section 6 for ninety 90 days after delivery to Lic ensee The warranties hereunder are void if the Software has been improp erly installed misused or if Licensee has violated the terms of this Agreement Altair s entire liability for all claims arising under or related in any way to this Agreement re
134. e X is the number of MPI processes on hostA Y is the number of MPI processes on hostB etc See the definition of the resources mpiprocs and ompthreads in Built in Resources on page 336 of the PBS Professional Reference Guide See also The mpiprocs Resource on page 227 56 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 4 6 2 PBS _ NODEFILE The file containing the vnodes allocated to a job lists vnode names This file s name is given by the environment variable PBS_NODEFILE For jobs which request vnodes via the lselect option the nodes file will con tain the names of the allocated vnodes with each name repeated M times where M is the number of mpiprocs specified for that vnode For example qsub 1 select 3 ncpus 2 lplace scatter will result in this PBS NODEFILE vnodeA vnodeB vnodeC And qsub l select 3 ncpus 2 mpiprocs 2 will result in this PBS NODEFILE vnodeA vnodeA vnodeB vnodeB vnodeC vnodeC For jobs which requested a set of nodes via the Inodes nodespec option to qsub each vnode allocated to the job will be listed N times where N is the total number of CPUs allocated from the vnode divided by the number of threads requested For example qsub Inodes 4 ncpus 3 ppn 2 will result in each of the four vnodes being written twice 6 CPUs divided by 3 from ncpus The file will contain the name of the first vnode twice followed by the second vnode twice etc
135. e as standard software Unsupported vmem Same as standard vnode Same as standard walltime Supported cpupercent Unsupported custom resources Unsupported 11 5 Managing HPCBP Jobs 11 5 1 Monitoring HPCBP Jobs You can use qstat f lt job ID gt to see a listing of your job s execut able and its argument list For example if your job request was qsub ping n 100 127 0 0 1 The output of qgstat f lt job ID gt will be executable lt jsdl hpcpa Executable gt ping lt jsdl hpcpa Executable gt argument list lt jsdl hpcpa Argument gt n lt jsdl hpcpa Argument gt lt jsdl hpcpa Argument gt 100 lt jsdl hpcpa Argument gt lt jsdl hpcpa Argument gt 127 0 0 1 lt jsdl hpcpa Argument gt PBS Professional 10 4 User s Guide 211 Chapter 11 HPC Basic Profile Jobs 11 5 1 1 Job Status Reporting PBS provides status reporting for HPC Basic Profile jobs via the qstat command The HPCBP MOM contacts the HPC Basic Profile Server and returns status information to the PBS Server The only information avail able is via the HPC Basic Profile The job states returned from HPC Basic Profile Server can be one of the following e Pending e Running e Failed e Finished e Terminated However the only states that are reported by qstat are e Running e Exiting The HPCBP Server reports that the job is in Running state whether the job is waiting to run or is running Once a job transitions to any of the states Termina
136. e dedicated non dedicated time borders PBS uses walltime to schedule within and around dedicated time Ifa job is submitted without a walltime to a non dedicated time queue it will not be started until all dedicated time periods are over Ifa job is submitted to a dedicated time queue without a walltime it will never run To submit a job to be run during dedicated time use the q lt queue name gt option to qsub and give the name of the dedicated time queue you wish to use as the queue name Queues are created by the administrator see your administrator for queue name s 8 10 Using Comprehensive System Accounting PBS supports Comprehensive System Accounting CSA on SGI Altix machines that are running SGI s ProPack 4 0 or greater and have the Linux job container facility available CSA provides accounting information about user jobs called user job accounting CSA works the same with and without PBS To run user job accounting either the user must specify the file to which raw accounting information will be written or an environment variable must be set The environment variable is ACCT _TMPDIR This is the directory where a temporary file of raw accounting data is written 196 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 To run user job accounting the user issues the CSA command ja lt filename gt or if the environment variable ACCT TMPDIR is set ja In order to have an a
137. e location where the input files come from For stage out this is where the output files end up when the job is done You must specify a hostname The name can be absolute or it can be relative to the user s home directory on the remote machine IMPORTANT It is advisable to use an absolute pathname for the remote_path Remember that the path to your home direc tory may be different on each machine and that when using sandbox PRIVATE you may or may not have a home directory on all execution machines For stagein the direction of travel is from remote_path to local_path For stageout the direction of travel is from local_path to remote_path The following example shows how to use a directive to stagein a file named grid dat located in the directory u user1 on the host called serverA The staged in file is copied to the staging and execution direc tory and given the name dat1 Since local_path is evaluated relative to the staging and execution directory it is not necessary to specify a full pathname for dat1 Always use a relative pathname for local_path when the job s staging and execution directory is created by PBS PBS W stagein datl serverA u userl grid dat To use the qsub option to stage in the file residing on myhost in Users myhome mydata datal calling it input_datal in the staging and execution directory qsub W stagein input_datal myhost Users myhome mydata datal To stage more than one file or dir
138. e locked licenses are typically set up as either a boolean resource on the vnode s that are licensed for the application The resource request should include one license for each host To request a host with a per host node locked license for AppA in one chunk qsub l select 1 runsAppA 1 lt jobscript gt Per use node locked licenses are typically set up so that the host s that run the application have the number of licenses that can be used at one time The number of licenses the job requests should be the same as the number of instances of the application that will be run To request a host with a per use node locked license for AppB where you ll run one instance of AppB on two CPUs in one chunk qsub l select 1 ncpus 2 AppB 1 Per CPU node locked licenses are set up so that the host has one license for each licensed CPU You must request one license for each CPU To request a host with a node locked license for AppC where you ll run a job using two CPUs in one chunk qsub l select 1 ncpus 2 AppC 2 PBS Professional 10 4 User s Guide 49 Chapter 4 Submitting a PBS Job 4 5 7 Requesting Scratch Space Scratch space on a machine is set up as a host level dynamic resource The resource will have a name such as dynscratch To request 10MB of scratch space in one chunk a resource request would include 1 select 1 ncpus N dynscratch 10MB 4 5 8 Note About Submitting Jobs The default for walltime is 5 years The schedu
139. e reservation s queue by using the qstat command See qstat on page 217 of the PBS Professional Refer ence Guide and the qstat 1B man page 8 8 6 1 Examples of Viewing Reservation Status Using pbs_rstat In our example we have one advance reservation and one standing reserva tion The advance reservation is for today for two hours starting at noon The standing reservation is for every Thursday for one hour starting at 3 00 p m Today is Monday April 28th and the time is 1 00 so the advance reservation is running and the soonest occurrence of the standing reservation is Thursday May 1 at 3 00 p m Example brief output pbs_rstat B Name R302 south Name S304 south Example short output pbs_rstat S Name Queue User State Start Duration End R302 south R302 userl RN Today 12 00 7200 Today 14 00 304 south S304 userl CO May 1 2008 15 00 3600 May 1 2008 16 00 Example full output pbs_rstat F 188 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 Name R302 south Reserve Name NULL Reserve Owner userl south mydomain com reserve state RESV_RUNNING reserve substate 5 reserve start Mon Apr 28 12 00 00 2008 reserve _end Mon Apr 28 14 00 00 2008 reserve duration 7200 queue R302 Resource List ncpus 2 Resource List nodect 1 02 00 00 Resource List select 1 ncpus 2 Resource List walltime Resource List place free resv_nodes south ncpus 2
140. e server or queue is treated as if it were infinite 4 2 0 2 Resource Names and Values The resource name is any string made up of alphanumeric characters where the first character is alphabetic Resource names must start with an alphabetic character and can contain alphanumeric underscore _ and dash characters If a string resource value contains spaces or shell metacharacters enclose the string in quotes or otherwise escape the space and metacharacters Be sure to use the correct quotes for your shell and the behavior you want If the string resource value contains commas the string must be enclosed in 34 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 an additional set of quotes so that the command e g qsub galter will parse it correctly If the string resource value contains quotes plus signs equal signs colons or parentheses the string resource value must be enclosed in yet another set of additional quotes 4 2 1 Resource Types See Resource Types on page 333 of the PBS Professional Reference Guide for a description of resource types 4 2 2 Built in Resources See Built in Resources on page 336 of the PBS Professional Reference Guide for a list of built in resources 4 3 PBS Jobs 4 3 1 Rules for Submitting Jobs e The place specification cannot be used without the select specifica tion See section 4 6 Placing Jobs on Vnodes on page 55 e A selec
141. e_path to local_path For stageout the direction of travel is from local_path to remote_path When staging more than one filename separate the filenames with a comma and enclose the entire list in double quotes PBS Professional 10 4 User s Guide 207 Chapter 9 Job Arrays Examples Remote_path store film Data files used as input framel frame2 frame3 Local_path pix Executable a out For this example a out produces frame2 out from frame2 PBS W stagein pix in frame array index store film frame array_index PBS W stageout pix out frame array_index out store film frame array index out PBS J 1 3 a out frameSPBS ARRAY INDEX in out Note that the stageout statement is all one line broken here for readability The result will be that the user s directory named film contains the origi nal files framel frame2 frame3 plus the new files framel out frame2 out and frame3 out 9 6 1 1 Scripts Example 1 In this example we have a script named ArrayScript which calls scriptlet1 and scriptlet2 All three scripts are located in homedir testdir bin sh PBS N ArrayExample PBS J 1 2 echo Main script index PBS ARRAY INDEX homedir testdir scriptlet PBS ARRAY INDEX In our example scriptlet and scriptlet2 simply echo their names We run ArrayScript using the qsub command qsub ArrayScript Example 2 In this example we have a script called StageScript It takes two input files
142. eases PBS Professional 10 4 User s Guide 303 Subsidiary Subsidiary means any partnership joint venture corporation or other form of enterprise in which a party possesses directly or indi rectly an ownership interest of fifty percent 50 or greater or manage rial or operational control Suppliers Any person corporation or other legal entity which may pro vide software or documents which are included in the Software Support The maintenance and support services provided by Altair pursu ant to this Agreement Templates Human readable ASCII files containing machine interpretable commands for use with the Software Term The term of licenses granted under this Agreement Annual licenses shall have a 12 month term of use unless stated otherwise on the Order Form Perpetual licenses shall have a term of twenty five years Maintenance agreements for perpetual licenses have a 12 month term Update A new version of the Products made generally available by Altair to its Licensees that includes additional features or functionalities but is substantially the same computer code as the existing Products 2 LICENSE GRANT Subject to the terms and conditions set forth in this Agreement Altair hereby grants Licensee and Licensee hereby accepts a limited non exclusive non transferable license to a install the Products on the License Network Server s identified on the Order Form for use only at the sites identified on the Orde
143. ective naming the job and name it newjob the user could type qsub N newjob my_job 4 4 1 2 Submitting a Simple Job Jobs can also be submitted without specifying values for attributes The simplest way to submit a job is to type qsub myjobscript lt ret gt If myjobscript contains bin sh myapplication the user has simply told PBS to run myapplication PBS Professional 10 4 User s Guide 41 Chapter 4 Submitting a PBS Job 4 4 1 3 Passing Arguments to Job Scripts If you need to pass arguments to a job script you can either use the v option to qsub where you set and use environment variables or use stan dard input When using standard input any PBS directives in the job script will be ignored You can replace directives with the equivalent options to qsub To use standard input you can either use this form echo jobscript sh a foo b bar qsub 1 select or you can use this form qsub option option lt ret gt jobscript sh foo lt d gt 152 mymachine With this form you can type the PBS directives on lines the name of the job script If you do not use the n option to qsub or specify it via a PBS directive second form only the job will be named STDIN 4 4 2 Jobs Without a Job Script There are two ways to submit PBS jobs without using a job script You can run a PBS job by specifying an executable and its arguments instead of a job script You can also specify that qsub read inpu
144. ectory use a comma separated list of paths and enclose the list in double quotes For example to stage two files datal and data2 in qsub W stagein input1 hostA myhome datal input2 hostA myhome datal e Under Windows special characters such as spaces backslashes colons and drive letter specifications are valid pathnames For 168 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 example the following will stagein the grid dat file on drive D at hostB to a local file dat1 on drive C qsub W stagein dat1 hostB D Documents and Set tings grid dat 8 6 4 1 Copying Directories Into and Out Of the Staging and Execution Directory You can stage directories into and out of the staging and execution direc tory the same way you stage files The remote_path and local_path for both stagein and stageout can be a directory If you stagein or stageout a directory PBS copies that directory along with all of its files and subdirectories At the end of the job the directory including all files and subdirectories is deleted This can create a problem if multiple jobs are using the same directory 8 6 4 2 Wildcards In File Staging You can use wildcards when staging files and directories according to the following rules e The asterisk matches one or more characters e The question mark matches a single character e All other characters match only themselves e Wildcard
145. ed by the range 9 1 2 Identifier Syntax Job arrays have three identifier syntaxes e The job array object itself 1234 server or 1234 e A single subjob of a job array with index M 1234 M server or 1234 M e A range of subjobs of a job array 1234 X Y Z server or 1234 X Y Z Examples 1234 server domain com Full job array identifier 1234 Short job array identifier 1234 73 Subjob identifier of the 73rd index of job array 1234 1234 Error if 1234 is a job array 1234 server domain com Error 1f 1234 server domain com is a job array The sequence number 1234 in 1234 server is unique so that jobs and job arrays cannot share a sequence number Note Since some shells for example csh and tesh read and J as shell metacharacters job array names and subjob names will need to be enclosed in double quotes for all PBS commands Example qdel 1234 myhost 5 qdel 1234 myhost PBS Professional 10 4 User s Guide 203 Chapter 9 Job Arrays Single quotes will work except where you are using shell variable substitu tion 9 2 qsub Submitting a Job Array To submit a job array qsub is used with the option J range where range is of the form X Y Z X is the starting index Y is the ending index and Z is the optional stepping factor X and Y must be whole numbers and Z must be a positive integer Y must be greater than X If Y is not a multiple of the stepping factor above X
146. ed environment variable names start with the characters PBS_ Some are then followed by a capital O PBS_O_ indicating that the variable is from the job s originating envi ronment i e the user s Appendix A gives a full listing of all environment variables provided to PBS jobs and their meaning The following short example lists some of the more useful variables and typical values PBS_O HOME u userl PBS _O LOGNAME user1 PBS_O PATH usr new bin usr local bin bin PBS _O SHELL sbin csh PBS _O HOST crayl PBS_O WORKDIR u user1 PBS_O QUEUE submit PBS_JOBID 16386 crayl PBS _QUEUE crayq PBS_ENVIRONMENT PBS_ INTERACTIVE 28 PBS Professional 10 4 User s Guide Getting Started With PBS Chapter 3 There are a number of ways that you can use these environment variables to make more efficient use of PBS In the example above we see PBS_ ENVIRONMENT which we used earlier in this chapter to test if we were running under PBS Another commonly used variable is PBS_O_WORKDIR which contains the name of the directory from which the user submitted the PBS job There are also two environment variables that you can set to affect the behavior of PBS The environment variable PBS_DEFAULT defines the name of the default PBS Server Typically it corresponds to the system name of the host on which the Server is running If PBS_DEFAULT is not set the default is defined by an administrator established file usually etc pbs conf
147. elect statement The place statement has this form l place arrangement sharing grouping where arrangement is one of free pack scatter sharing is one of excl shared grouping can have only one instance of group resource PBS Professional 10 4 User s Guide 55 Chapter 4 Submitting a PBS Job and where Table 4 1 Placement Modifiers Modifier Meaning free Place job on any vnode s pack All chunks will be taken from one host scatter Only one chunk will be taken from a host exclusive Only this job uses the vnodes chosen shared This job can share the vnodes chosen group resource Chunks will be grouped according to a resource All vnodes in the group must have a common value for the resource which can be either the built in resource host or a site defined vnode level resource Note that vnodes can have sharing attributes that override job placement requests See the pbs_node_attributes 7B man page Grouping by resource name will override node_group_key Toruna job on a single host use lplace pack 4 6 1 Vnodes Allocated to a Job The nodes file contains the names of the vnodes allocated to a job The nodes file s name is given by the environment variable PBS_NODEFILE The order in which hosts appear in the file is the order in which chunks are specified in the selection directive The order in which hostnames appear in the file is hostA X times hostB Y times wher
148. en scatter is used the 4 chunks are on different hosts so each host has 1 hps resource qsub 1 select 4 ncpus 2 hps 1 Using place pack When pack is used all the chunks are put on one host so a chunk with no resources and one hps must be specified qsub 1 select 4 ncpus 2 1 ncpus 0 hps 1 This ensures that the hps resource is only counted once You could also use this qsub 1 select 1 ncpus 8 hps 1 For two chunks of 4 CPUs one on one machine and one on another you would use qsub 1 select 2 ncpus 4 l place scatter 10 6 3 6 Useful Information 10 6 3 6 1 IBM Documentation e IBM has documentation describing an InfiniBand cluster Clustering systems using InfiniBand hardware available from IBM at http publib boulder ibm com infocenter systems scope hw index jsp topic iphau referenceinfini band htm e IBM offers guidance on how to use the IB switch in US mode without LoadLeveler PE v4 3 2 doc Configuring InfiniBand for User Space without LoadLeveler PE for AIX only available from IBM at http publib boulder ibm com infocenter clresctr vxrx index jsp topic com ibm clus ter pe432 install doc am101 configusib html e IBM offers a programming API for communicating with and configur ing the InfiniBand switch NRT API Programming Guide PBS Professional 10 4 User s Guide 237 Chapter 10 Multiprocessor Jobs 10 6 3 6 2 Sources for Sample Code When installing the p
149. engine 1 0 20 Q If you attempted to release a hold on a job which is not on hold the request will be ignored If you use the qr 1s command to release a hold on a job that had been previously running and subsequently checkpointed the hold will be released and the job will return to the queued Q state and be eli gible to be scheduled to run when resources come available To hold or release a job using xpbs first select the job s of interest then click the hold or release button The qrls command does not run the job it simply releases the hold and makes the job eligible to be run the next time the scheduler selects it PBS Professional 10 4 User s Guide 123 Chapter 6 Working With PBS Jobs 6 3 Deleting Jobs PBS provides the qdel command for deleting jobs The qdel command deletes jobs in the order in which their job identifiers are presented to the command A batch job may be deleted by its owner a PBS operator or a PBS administrator Example qdel 51 qdel 1234 server Job array identifiers must be enclosed in double quotes Mail is sent for each job deleted unless you specify otherwise Use the fol lowing option to qde1 to prevent more email than you want from being sent Wsuppress_ email lt N gt N must be a non negative integer Make N the largest number of emails you wish to receive per qdel command PBS will send one email for each deleted job up to N Note that a job array is one job
150. equested session_id Returns process ID of process started by the HPCBP MOM for job management not of HPCBP job itself stime Reported start time of job may be inexact substate The job substate may not be same in HPC Basic Profile Server and PBS group_list Unsupported stagein Specified files are staged in stageout Specified files are staged out umask Unsupported PBS Professional 10 4 User s Guide 2135 Chapter 11 HPC Basic Profile Jobs 11 4 3 Requesting Resources The following table shows the behavior for of PBS resources HPCBP jobs Table 11 3 PBS Resources and Their Behavior for HPCBP Jobs PBS Resource Behavior arch Same as standard cput Amount of disk space for job file Same as standard host Same as standard mem Same as standard mpiprocs Number of CPUs to be allocated to job mppwidth Unsupported mppdepth Unsupported mppnppn Unsupported mppnodes Unsupported mpplabels Unsupported mppmem Unsupported mpphost Unsupported mpparch Unsupported ncpus Same as standard nice Unsupported nodect Unsupported ompthreads Must specify equal number of ompthreads in all chunks of multi chunk job pcput Same as standard pmem Same as standard 276 PBS Professional 10 4 User s Guide HPC Basic Profile Jobs Chapter 11 Table 11 3 PBS Resources and Their Behavior for HPCBP Jobs PBS Resource Behavior pymem Sam
151. er s default home directory More specifically the actual home path will be PROFILE PATH My Documents PBS Pro For instance if a userA has not been assigned a home directory it will default to a local home directory of Documents and Settings userA My Documents PBS Pro UserA s job will use the above path as its working directory PBS Professional 10 4 User s Guide 25 Chapter 3 Getting Started With PBS Note that Windows can return as PROF ILE_PATH one of the following forms Documents and Settings username Documents and Settings username local host name Documents and Settings username local host name 00N where N is a number Documents and Settings username domain name 3 12 3 Windows Usernames and Job Submission A PBS job is run from a user account and the associated username string must conform to the POSIX 1 standard for portability That is the user name must contain only alphanumeric characters dot underscore _ and or hyphen The hyphen must not be the first letter of the username If appears in the username then it will assumed to be in the context of a Windows domain account username domainname An exception to the above rule is the space character which is allowed If a space character appears in a username string then it will be displayed quoted and must be specified in a quoted manner The following example requests the job to run under account Bob Jones qs
152. er state and start duration and end times of each reservation Full Lists the name and all non default value attributes for each reservation lt none gt Default Default is S option The full listing for a standing reservation is identical to the listing for an advance reservation with the following additions A line that specifies the recurrence rule reserve rrule FREQ WEEKLY BYDAY MO COUNT 5 An entry for the vnodes reserved for the soonest occurrence of the standing reservation This entry also appears for an advance reserva tion but will be different for each occurrence resv_nodes vnode_name A line that specifies the total number of occurrences of the standing res ervation reserve count 5 The index of the soonest occurrence reserve index 1 The timezone at the site of submission of the reservation is appended to the reservation Variable list For example in California Variable List lt other variables gt PBS_TZID America Los_Angeles To get the status of a reservation at a server other than the default server set the PBS_ SERVER environment variable to the name of the server you wish to query then use the pbs_rstat command Your PBS commands will treat the new server as the default server so you may wish to unset this environment variable when you are finished PBS Professional 10 4 User s Guide 187 Chapter 8 Advanced PBS Features You can also get information about th
153. er party of its obligations under this Agree ment except for a breach by Altair of the warranty set forth in Section 8 for which a remedy is provided under Section 10 or a breach by Licensee of Section 5 or Section 6 for which no cure period is provided and Altair may terminate this Agreement immediately if such default or breach continues for more than thirty 30 days after receipt of such notice Upon termina tion of this Agreement Licensee must cease using the Software and at Altair s option return all copies to Altair or certify it has destroyed all such copies of the Software and Documentation PBS Professional 10 4 User s Guide 311 15 GENERAL PROVISIONS Export Controls Licensee acknowl edges that the Products may be subject to the export control laws and regu lations of the United States and other countries and any amendments thereof Licensee agrees that Licensee will not directly or indirectly export the Products into any country or use the Products in any manner except in compliance with all applicable U S and other countries export laws and regulations Notice All notices given by one party to the other under this Agreement shall be sent by certified mail return receipt requested or by overnight courier to the respective addresses set forth in this Agreement or to such other address either party has specified in writing to the other All notices shall be deemed given upon actual receipt Assignment Neither party shall
154. ere in this Agreement the following terms shall have the meanings defined below for purposes of this Agreement Additional Terms Additional Terms are those terms and conditions which are determined by an Altair Subsidiary to meet local market conditions Documentation Documentation provided by Altair or its resellers on any media for use with the Products Execute To load Software into a computer s RAM or other primary mem ory for execution by the computer Global Zone Software is licensed based on three Global Zones the Amer icas Europe and Asia Pacific When Licensee has Licensed Workstations located in multiple Global Zones which are connected to a single License Network Server a premium is applied to the standard Software License pricing for a single Global Zone 302 PBS Professional 10 4 User s Guide ISV Independent Software Vendor A software company providing its products ISV Software to Altair s Licensees through the Altair License Management System using Altair License Units License Log File A computer file providing usage information on the Software as gathered by the Software License Management System The license management system LMS that accompanies the Software and limits its use in accordance with this Agreement and which includes a License Log File License Network Server A network file server that Licensee owns or leases located on Licensee s premises and identified by machine serial
155. erver log e The HPCBP MOM logs an error message PBS Professional 10 4 User s Guide 219 Chapter 11 HPC Basic Profile Jobs 11 6 4 Error Log Messages at Job Submission Querying and Deletion The HPCBP MOM logs a warning message in the MOM log file whenever it gets any error or warning at the time of e Job submission e Contacting the HPC Basic Profile Server to find job status e Job deletion The HPCBP MOM logs job errors in the file lt PBS job ID gt log The HPCBP MOM stages this file out to the location specified for stdout and stderr files The HPCBP MOM generates log messages depending on their event type and event class You can use the tracejob command to see these log messages The following table shows the warning and error messages logged by the HPCBP MOM and the PBS Server Table 11 4 Warning and Error Messages Logged by HPCBP MOM Logged Error Condition by Message Password related issues Bad user credential at the HPCBP lt username gt unable to termi time of qdel MOM nate the job with user s creden PBS tials Server Cannot determine job HPCBP lt pbsnobody gt unable to deter state when finding status MOM mine the state of the job of jobs running at HPC Basic Profile Server Conversion of PBS job request to JSDL Problem with parsing job HPCBP unable to parse the job request request MOM 280 PBS Professional 10 4 User s Guide HPC Basic P
156. eservation will remain unconfirmed 12 5 4 Held and Requeued Jobs The job is held with a system hold for the following reasons e Provisioning fails due to invalid provisioning request or to internal sys tem error e After provisioning the AOE reported by the vnode does not match the AOE requested by the job The hold can be released by the PBS Administrator after investigating what went wrong and correcting the mistake 296 PBS Professional 10 4 User s Guide Using Provisioning Chapter 12 The job is requeued for the following reasons e The provisioning hook fails due to timeout e The vnode is not reported back up 12 5 5 Conflicting Resource Requests The values of the resources arch and vnode may be changed by provision ing Do not request an AOE and either arch or vnode for the same job PBS Professional 10 4 User s Guide 297 Chapter 12 Using Provisioning 298 PBS Professional 10 4 User s Guide Appendix A Convert ing NQS to PBS For those converting to PBS from NQS or NQE PBS includes a utility called nqs2pbs which converts an existing NQS job script so that it will work with PBS In fact the resulting script will be valid to both NQS and PBS The existing script is copied and PBS directives PBS are inserted prior to each NQS directive either QSUB or Q in the original script nqs2pbs existing NQS script new PBS script Section Setting Up Your UNIX Linux Environment on
157. essional 10 4 User s Guide Index examples 243 MPICH2 251 262 examples 254 263 MPICH GM MPD 246 examples 247 rsh ssh 242 MPICH MX 244 MPD 248 examples 250 rsh ssh 244 examples 245 MPI OpenMP 228 mpirun 231 Intel MPI 256 MPICH2 251 MPICH GM MPD 246 MPICH GM rsh ssh 242 MPICH MX MPD 248 MPICH MxX rsh ssh 244 MVAPICH1 261 MVAPICH2 262 mpirun ch_gm 242 mpirun ch_mx 244 mpirun mpd 246 248 MRJ Technology Solutions ix MRJ Veridian 4 MVAPICHI1 261 examples 262 N name 81 NASA Ames Research Center 4 and PBS ix 3 Information Power Grid 5 Metacenter 5 Network Queueing System NQS 4 nqs2pbs 299 319 Index Node Grouping job arrays 224 Node Specification Conversion 66 Node specification format 66 nqs2pbs 20 O OpenMP 228 Ordering job arrays 218 Ordering Job Arrays in the Queue 218 Ordering Software and Publications xii override 39 P pack 56 Parallel job support 6 Virtual Machine PVM 266 password single signon 73 Windows 73 PBS availability 7 PBS commands job arrays 212 PBS Environmental Variables 206 PBS_ARRAY_ID 206 PBS_ARRAY INDEX 206 PBS_DEFAULT 29 PBS DEFAULT SERVER 111 PBS_DPREFIX 29 PBS ENVIRONMENT 23 29 pbs_hostn 20 PBS_JOBID 206 pbs_migrate_users 20 PBS_O WORKDIR 29 pbs_password 20 73 74 pbs_probe 20 320 pbs_rdel 20 pbs_rstat 20 pbs_rsub 20 185 pbs_tclsh 20 pbsdsh 20 177 pbsfs 20 pbsnodes 20 pbs report 20 Peer Scheduling job arrays 223 per C
158. essional 10 4 User s Guide Submitting a PBS Job Chapter 4 The next example Windows script shows a job wide request for walltime and a chunk request for ncpus and memory PBS 1 walltime 1 00 00 PBS 1 select ncpus 4 mem 400mb PBS j oe date t my_application date t Keep in mind the difference between requesting a vnode level boolean and a job wide boolean qsub 1 select 1 green True will request a vnode with green set to True However qsub l1 green True will request green set to True on the server and or queue 4 5 5 Default Resources Jobs get default resources both job wide and per chunk with the follow ing order of precedence from 1 Default qsub arguments 2 Default queue resources 3 Default server resources For each chunk in the job s selection statement first queue chunk defaults are applied then server chunk defaults are applied If the chunk request does not specify a resource listed in the defaults the default is added Fora resource RESOURCE a chunk default is called default_chunk RESOURCE PBS Professional 10 4 User s Guide 47 Chapter 4 Submitting a PBS Job For example if the queue in which the job is enqueued has the following defaults defined default_chunk ncpus 1 default_chunk mem 2gb a job submitted with this selection statement select 2 ncpus 4 1 mem 9gb will have this specification after the default_ chunk elements are applied select 2 ncpus 4 mem 2gbt 1 ncpus 1 me
159. etween chunks PBS Professional 10 4 User s Guide 45 Chapter 4 Submitting a PBS Job A chunk is one or more resource _name value statements separated by a colon e g ncpus 2 mem 10GB host Host1 ncpus 1 mem 20GB arch linux Example of multiple chunks in a selection statement l select 2 ncpus 1 mem 10GB 3 ncpus 2 mem 8GB arch solaris Each job submission can have only one I select statement Host level resources can only be requested as part of a chunk Server or queue resources cannot be requested as part of a chunk They must be requested outside of the selection statement 4 5 3 Requesting Job wide Resources A job wide resource request is for resource s at the server or queue level Job wide resources are requested outside of a selection statement in this form l keyword value keyword value where keyword identifies either a consumable resource or a time based resource such as walltime Job wide resources are used for requesting floating licenses or other resources not tied to specific vnodes such as cput and walltime Job wide resources can only be requested outside of chunks 4 5 4 Boolean Resources A resource request can specify whether a boolean resource should be true or false For example if some vnodes have green true and some are red true a selection statement for two vnodes each with one CPU all green and no red would be 1l select 2 green true red false ncpus 1 46 PBS Prof
160. fficient use of resources both computer time and people time Monitoring The act of tracking and reserving system resources and enforcing usage policy This includes both software enforcement of usage limits and user or administrator moni toring of scheduling policies to see how well they are meet ing stated goals 2 1 PBS Components PBS consist of two major component types user level commands and sys tem daemons services A brief description of each is given here to help you understand how the pieces fit together and how they affect you PBS Commands Server Scheduler Batch Job 10 PBS Professional 10 4 User s Guide Concepts and Components Chapter 2 Commands PBS supplies both command line programs that are POSIX 1003 2d conforming and a graphical interface These are used to submit monitor modify and delete jobs These cli ent commands can be installed on any system type sup ported by PBS and do not require the local presence of any of the other components of PBS There are three command classifications user commands which any authorized user can use operator commands and manager or administrator commands Operator and man ager commands which require specific access privileges are discussed in the PBS Professional Administrator s Guide Server The Job Server daemon service is the central focus for PBS Within this document it is generally referred to as the Server or by the exe
161. figuration File 111 5 9 pbs Preferences os ces ewes dda duwen ds aue exes 111 6 Working With PBS Jobs 117 6 1 Modifying Job Attributes 117 6 2 Holding and Releasing Jobs 121 Oo Delenne IOUS earne sexx aed fen pede 5 oe ees 124 6 4 Sending Messages to Jobs 004 125 6 5 Sending Signals to JODS i 6cs0css vees eaaeee 126 6 6 Changing Order of Jobs vic kocewde chee ee sks 127 6 7 Moving Jobs Between Queues 128 6 8 Converting a Job into a Reservation Job 130 6 9 Using Job History Information 131 7 Checking Job System Status 135 7 1 The qstatCommands o3 4 40saveacwadadaae es 135 7 2 Viewing Job System Status with xpbs 152 7 3 The qselect Command 5 idsicsccaccedu ces be Ex 152 7 4 Selecting Jobs Using xpbs nananana 154 7 5 Using xpbs TrackJob Feature ananan 155 8 Advanced PBS Features 157 8 1 UNIX Job Exit Statues 20 bcos Adaw eves g wed en 157 8 2 Changing UNIX Job umask 24054 400 158 8 3 Requesting qsub Wait for Job Completion 158 8 4 Specifying Job Dependencies 159 8 5 Delivery of Output Files lt 2 den nananana 162 8 6 Input Output File Staging 163 8 7 The pbsdsh Command 0 cctecdesvandee ds 177 8 8 Advance and Standing Reservation of Resources 178 8 9 Dedicated Time 52 4 02 048edhes aor eke rharade 196 PBS Professional 10 4 User s Guide
162. file Server This MOM is a client side implementation of the HPC Basic Profile Specification and acts as a proxy for and interface to an HPC Basic Profile compliant server HPC Basic Profile Job HPCBP Job Generic job that can run either on vnodes managed by PBS or on nodes managed by HPC Basic Profile Server Job Submission Description Language JSDL Language for describing the resource requirements of jobs 11 2 How HPC Basic Profile Jobs Work 11 2 1 Introduction PBS automatically schedules jobs on vnodes managed by PBS Professional or on nodes managed by an HPC Basic Profile Server without the need for you to specify destination specific parameters Whether the jobs run on PBS Professional or on an HPC Basic Profile Server is based only on site policies and resource availability You can use the qstat command for status reporting and the qdel com mand to cancel a job regardless of where the job runs Jobs eligible to run on the HPCBP Server must specify only a single exe cutable and its arguments and must do so via the qsub command line The job specification must be valid for both PBS and the HPCBP Server A job that is eligible to run on the HPCBP Server is called an HPCBP job in this document 11 2 2 Assigning Nodes and Resources to Jobs The HPCBP MOM does not control the resources assigned from each node for a job The HPC Basic Profile Server assigns resources to the job according to its scheduling policy 270 PBS
163. form Resource Identifier URI Generic Syntax http www rfc editor org rfc rfc3986 txt PBS Professional 10 4 User s Guide 287 Chapter 11 HPC Basic Profile Jobs 288 PBS Professional 10 4 User s Guide Chapter 12 Using Provisioning PBS provides automatic provisioning of an OS or application on vnodes that are configured to be provisioned When a job requires an OS that is available but not running or an application that is not installed PBS provi sions the vnode with that OS or application 12 1 Definitions AOE The environment on a vnode This may be one that results from provisioning that vnode or one that is already in place Provision To install an OS or application or to run a script which per forms installation and or setup PBS Professional 10 4 User s Guide 289 Chapter 12 Using Provisioning Provisioned Vnode A vnode which through the process of provisioning has an OS or application that was installed or which has had a script run on it 12 2 How Provisioning Works Provisioning can be performed only on vnodes that have provisioning enabled shown in the vnode s provision_enable attribute Provisioning can be the following e Directly installing an OS or application e Running a script which may perform setup or installation Each vnode is individually configured for provisioning with a list of avail able AOEs in the vnode s resources_available aoe attribute Each vnode s current_aoe
164. g Pending Terminated job transitioned from pending to termi nated Running Terminated job transitioned from running to terminated Running Failed job transitioned from running to failed Running Finished job completed successfully Pending Finished job transitioned from pending to finished Pending Failed job transitioned from pending to failed none Failed job first appeared in Failed state Whenever a job is submitted to the HPC Basic Profile Server the HPCBP MOM logs the following message job submitted to HPCBP Server as jobid lt hpcbp jobid gt in state lt state gt 284 PBS Professional 10 4 User s Guide HPC Basic Profile Jobs Chapter 11 11 7 Advice and Caveats 11 7 1 Differences Between PBS and HPCBP e The stime attribute in the PBS accounting logs may not represent the exact start time for an HPCBP job e The HPCBP MOM does not use the pbs_ rcp command for staging operations regardless of whether the PBS_SCP environment variable has been set in the configuration file 11 7 2 PBS Features Not Supported With HPCBP e Peer Scheduling e Job operations Suspend resume Checkpoint 11 7 2 1 Unsupported Commands If the user or administrator runs the ppsdsh command for a job running on the HPCBP Server the HPCBP MOM logs an error message to the MOM file and rejects the job PBS Professional 10 4 User s Guide 2
165. g ProPack 4 or greater The PBS mpiexec has the standard mpiexec interface The PBS mpiexec does require proper configuration of the Altix See your administrator to find out whether your system is configured for the PBS mpiexec You can launch an MPI job on a single Altix or across multiple Altixes PBS will manage and track the processes You can use CSA if it is config ured to collect accounting information on your jobs PBS will run the MPI tasks in the cpusets it manages PBS Professional 10 4 User s Guide 265 Chapter 10 Multiprocessor Jobs You can run MPI jobs in the placement sets chosen by PBS When a job is finished PBS will clean up after it For MPI jobs across multiple Altixes PBS will manage the multihost jobs For example if you have two Altixes named Alt1 and Alt2 and want to run two applications called mympil and mympi2 on them you can put this in your job script mpiexec host Alt1 n 4 mympil host Alt2 n 8 mympi2 You can specify the name of the array to use via the PBS_MPI SGIARRAY environment variable To verify how many CPUs are included in a cpuset created by PBS use gt cpuset d lt set name gt egrep cpus This will work either from within a job or not The alt_id returned by MOM has the form cpuset lt name gt lt name gt is the name of the cpuset which is the PBS_JOBID Jobs will share cpusets if the jobs request sharing and the cpusets sharing attribute is not set to force_exc
166. g error malfunction and other problems in accordance with Altair s then current problem reporting procedure If Altair believes that a problem reported by Licensee may not be due to an error in the Software Altair will so notify Licensee Questions must be directed to Altair s specially designated telephone support numbers and email addresses Support will also be available via email at Internet addresses designated by Altair Support is available Monday through Fri day excluding holidays from 8 00 a m to 5 00 p m local time in the Glo bal Zone where licensed unless stated otherwise on the Order Form Exclusions Altair shall have no obligation to maintain or support a altered damaged or Licensee modified Software or any portion of the Software incorporated with or into other software not provided by Altair b any version of the Software other than the current version of the Soft ware or the immediately prior release of the Software c problems caused by Licensee s negligence abuse or misapplication of Software other than as specified in the Documentation or other causes beyond the reasonable control of Altair or d Software installed on any hardware operating sys tem version or network environment that is not supported by Altair Sup port also excludes configuration of hardware non Altair Software and networking services consulting services general solution provider related services and general computer system maintenance
167. gardless of legal theory shall be limited to direct damages and shall not exceed in the aggregate for all claims the license and maintenance fees paid under this Agreement by Licensee in the 12 months prior to the claim on a prorated basis except for claims under Sec tion 7 ALTAIR AND ITS SUPPLIERS AND ISVS SHALL NOT BE LIABLE TO LICENSEE OR ANYONE ELSE FOR INDIRECT SPE CIAL INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING HEREUNDER INCLUDING LOSS OF PROFITS OR DATA DEFECTS IN DESIGN OR PRODUCTS CREATED USING THE SOFTWARE OR ANY INJURY OR DAMAGE RESULTING FROM SUCH DEFECTS SUFFERED BY LICENSEE OR ANY THIRD PARTY EVEN IF ALTAIR OR ITS SUPPLIERS OR ITS ISVS HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES Licensee acknowledges that it is solely responsible for the adequacy and accuracy of the input of data including the output generated from such data and agrees to defend indemnify and hold harmless Altair and its Suppliers and ISVs from any and all claims including reasonable attorney s fees resulting from or in connection with Licensee s use of the Software No action regardless of form arising out of the transactions under this Agreement may be brought by either party against the other more than two 2 years after the cause of action has accrued except for actions related to unpaid fees 9 UNITED STATES GOVERNMENT RESTRICTED RIGHTS This section applies to all acquisitions of the Products by or for the United States gover
168. given job is not running or when the job began executing A text string of is printed for jobs whose com ment has not yet been set The example below illustrates the different type of messages that may be displayed qstat s Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time 16 south userl workq aims14 1 0 01 H Job held by userl on Wed Aug 22 13 06 11 2004 18 south userl workq aims14 1 0 01 W PBS Professional 10 4 User s Guide 145 Chapter 7 Checking Job System Status Waiting on user requested start time 51 south barry workg airfoil 930 1 0 13 R 0 01 Job run on host south started Thu Aug 23 at 10 56 52 south userl workq my_job 1 0 10 Q Not Running No available resources on nodes 57 south susan workq solver 2 0 20 Q 7 1 13 Display Queue Limits The q option to qstat displays any limits set on the requested or default queues Since PBS is shipped with no queue limits set any visible limits will be site specific The limits are listed in the format shown below qstat q server south Queue Memory CPU Time Walltime Node Run Que Im State 7 1 14 Show State of Job Job Array or Subjob The t option to qstat will show the state of a job a job array object and all non X subjobs In combination with J qstat will show only the state of subjobs 7 1 15 Viewing Job Start Time There are two ways you can find the j
169. he recurrence rule override those specified in the arguments to the R and E options Therefore if you specify R 0730 E 0830 BYHOUR 9 PBS Professional 10 4 User s Guide 183 Chapter 8 Advanced PBS Features the duration is one hour but the hour element 9 00 in the recurrence rule has overridden the hour element specified in the argument to R 7 00 The offset is still 30 minutes after the interval start Your reservation will run from 9 30 to 10 30 Similarly ifthe 16th is a Monday and you specify R 160800 E 170900 BYDAY TU BYHOUR 11 the duration 25 hours but both the day and the hour elements have been overridden Your reservation will run on Tuesday at 11 for 25 hours end ing Wednesday at 12 However if you specify R 160810 E 170910 BYDAY TU BYHOUR 11 the duration is 25 hours and the offset from the interval start is 10 minutes Your reservation will run on Tuesday at 11 10 for 25 hours ending Wednesday at 12 10 The minutes in the offset weren t overridden by any thing in the recurrence rule The values specified for the arguments to the R and E options can be used to set the start and end times in a standing reservation just as they are in an advance reservation To do this don t override their elements inside the recurrence rule If you specify R 0930 E 1030 BYDAY MO TU you haven t overridden the hour or minute elements Your reservation will run Monday and Tuesday from 9 30
170. hed job j2 for which PBS is maintaining history than j1 will go into the held state If job j1 depends on a finished job j3 that has been purged from the historical records than j1 will be rejected just as in previous versions of PBS where the job was no longer in the system 3 6 2 PBS path information no longer saved in AUTOEXEC BAT Any value for PATH saved in AUTOEXEC BAT may be lost after installa tion of PBS If there is any path information that needs to be saved AUTOEXEC BAT must be edited by hand after the installation of PBS PBS path information is no longer saved in AUTOEXEC BAT 3 7 Using PBS From the user s perspective a workload management system allows you to make more efficient use of your time You specify the tasks you need exe cuted The system takes care of running these tasks and returning the results to you If the available computers are full then the workload man agement system holds your work and runs it when the resources are avail able With PBS you create a batch job which you then submit to PBS A batch job is a file a shell script under UNIX or a cmd batch file under Windows containing the set of commands you want to run on some set of execution machines It also contains directives which specify the characteristics attributes of the job and resource requirements e g memory or CPU 18 PBS Professional 10 4 User s Guide Getting Started With PBS Chapter 3 time that your job needs Once you crea
171. iBand If executed inside a PBS job this allows for PBS to track all MVAPICH1 processes so that PBS can perform accounting and have complete job control If executed outside of a PBS job it behaves exactly as if standard MVAPICH1 mpirun had been used You use the same mpirun command as you would use outside of PBS 10 6 11 1 Options If executed inside a PBS job script all of the options to the PBS interface are the same as MVAPICH1 s mpirun except for the following map The map option is ignored machinefile lt file gt The machinefile option is ignored exclude The exclude option is ignored Np If the user does not specify a np option then PBS uses the number of entries found in the PBS_NODEFILE The maximum number of ranks that can be launched is the num ber of entries in PBS_ NODEFILE PBS Professional 10 4 User s Guide 261 Chapter 10 Multiprocessor Jobs 10 6 11 2 Examples Example 1 Runa single executable MVAPICH1 job with 6 ranks spread out across the PBS allocated hosts listed in PBS_NODEFILE PBS_NODEFILE pbs hostl pbs hostl pbs host2 pbs host2 pbs host3 pbs host3 Job script mpirun runs 6 processes mapped to each host listed in PBS NODEFILE mpirun np 6 path myprog Run job script qsub l select 3 ncpus 2 mpiprocs 2 job script lt job id gt 10 6 12 PBS Jobs with MVAPICH2 s mpiexec PBS provides an interface to MVAPICH2 s mpiexec MVAPICH2 allows the use of InfiniBa
172. icense using old syntax 1 ncpus 1 mem 5mb host sunny opti 1 arch solaris is converted to the equivalent 1 select 1 ncpus 1 mem 5gb host sunny arch solaris l place pack 1 opti 1 3 Request one host with a certain property using old syntax lnodes 1 property is converted to the equivalent l select 1 ncpus 1 property True l place scatter 4 Request 2 CPUs on each of four hosts with a given property using old syntax lnodes 4 property ncpus 2 is converted to the equivalent l select 4 ncpus 2 property True PBS Professional 10 4 User s Guide 63 Chapter 4 Submitting a PBS Job place scatter 5 Request 1 CPU on each of 14 hosts asking for certain software licenses and a job limit amount of memory using old syntax lnodes 14 mpi fluent ncpus 1 lfluent 1 fluent all 1 fluent par 13 1 mem 280mb is converted to the equivalent 1l select 14 ncpus 1 mem 20mb mpi_fluent True l place scatter 1 fluent 1 fluent all 1 fluent par 13 6 Requesting licenses using old syntax lnodes 3 dyna mpi Linux ncpus 2 ldyna 6 mem 100mb software dyna is converted to the equivalent 1 select 3 ncpus 2 mem 33mb dyna mpi Linux True l place scatter 1l software dyna 1 dyna 6 7 Requesting licenses using old syntax l ncpus 2 app lic 6 mem 200mb 1 software app is converted to the equivalent l select 1 ncpus 2 mem 200mb l place pack l software app l app lic 6 8 Additional example using old syntax
173. il qstat B f selected server_host s Hosts Panel submit qsub options selected Server s Hosts Panel terminate qterm selected server_host s Queues Panel detail qstat Q f selected queue s Queues Panel stop qstop selected queue s Queues Panel start qstart selected queue s Queues Panel enable qenable selected queue s Queues Panel disable qdisable selected queue s Jobs Panel detail qstat f selected job s Jobs Panel modify qalter selected job s Jobs Panel delete qdel selected job s Jobs Panel hold qhold selected job s Jobs Panel release qrls selected job s Jobs Panel run qrun selected job s Jobs Panel rerun qrerun Selected job s Jobs Panel signal qsig selected job s Jobs Panel msg qmsg selected job s 106 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 Table 5 4 xpbs Buttons and PBS Commands Command Location Buton PBS Command Jobs Panel move qmove selected job s Jobs Panel order qorder selected job s Indicates command button is visible only if xpbs is started with the admin option 5 6 How to Submit a Job Using xpbs To submit a job using xpbs perform the following steps First select a host from the HOSTS listbox in the main xpbs display to which you wish to submit the job Next click on the Submit button located next to the HOSTS panel The Submit button bring
174. ill include ncpus X If no spec is ncpus X and no spec is cpp X The chunk will include ncpus P If spec is ppn P The chunk will include mpiprocs P If the nodespec is l1nodes N ppn P It is converted to lselect N ncpus P mpiprocs P Example lnodes 4 ppn 2 is converted into lselect 4 ncpus 2 mpiprocs 2 If Incpus Z is specified and no spec contains ncpus X and no spec is cpp X Every chunk will include ncpus W where W is Z divided by the total number of chunks Note W must be an integer Z must be evenly divisible by the number of chunks If property is a suffix All chunks will include property true If excl is a suffix The placement directive will be lplace scatter excl If shared is a suffix The placement directive will be lplace scatter shared If neither excl nor shared is a suffix The placement directive will be lplace scatter PBS Professional 10 4 User s Guide 67 Chapter 4 Submitting a PBS Job Example 1l nodes 3 green ncpus 2 ppn 2 2 red is converted to 1 select 3 green true ncpus 4 mpiprocs 2 2 red true ncpus 1 l place scatter Node specification syntax for requesting properties is deprecated The boolean resource syntax property true is only accepted in a selection directive It is erroneous to mix old and new syntax 4 8 2 Resource Specification Conversion The resource specification is converted to select and place statements after any defaults have been applied Reso
175. instruc tions and e invoicing directions are not binding on the parties In the event of a conflict between the terms of this Agreement and the Additional Terms the Agreement shall take precedence Entire Agreement This Agreement the Additional Terms and the Order Form s attached hereto constitute the entire understanding between the parties related to the sub ject matter hereto and supersedes all proposals or prior agreements whether written or oral and all other communications between the parties with respect to such subject matter This Agreement may be executed in 312 PBS Professional 10 4 User s Guide Chapter 14 one or more counterparts all of which together shall constitute one and the same instrument Execution Copies of this Agreement executed via origi nal signatures facsimile or email shall be deemed binding on the parties PBS Professional 10 4 User s Guide 313 Chapter 14 314 PBS Professional 10 4 User s Guide Index A Access Control 6 Accounting 6 job arrays 222 accounting 196 ACCT_TMPDIR 196 Administrator Guide xii 19 Advance reservation creation 180 advance reservation 178 Aerospace computing 3 AIX 232 Large Page Mode 199 Altair Engineering 4 5 Altair Grid Technologies 4 Altering job arrays 217 Ames Research Center ix PBS Professional 10 4 User s Guide AOE 289 using 291 API xii 6 12 177 application licenses floating 48 node locked per CPU 49 per host 49 per use 4
176. ional 10 4 User s Guide 97 Chapter 5 Using the xpbs GUI 5 3 2 xpbs Hosts Panel The Hosts panel is composed of a leading horizontal HOSTS bar a listbox and a set of command buttons The HOSTS bar contains a minimize maxi mize button identified by a dot or a rectangular image for displaying or iconizing the Hosts region The listbox displays information about favorite Server host s and each entry is meant to be selected via a single left click shift left click for contiguous selection or control left click for non con tiguous selection To the right of the Hosts Panel are buttons that represent actions that can be performed on selected host s Use of these buttons will be explained in detail below 98 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 detail Provides information about selected Server host s This functionality can also be achieved by double clicking on an entry in the Hosts listbox submit For submitting a job to any of the queues managed by the selected host s terminate For terminating shutting down PBS Servers on selected host s Visible via the admin option only IMPORTANT Note that some buttons are only visible if xpbs is started with the admin option which requires manager or operator privilege to function The middle portion of the Hosts Panel has abbreviated column names indi cating the information being displayed as the following table shows Ta
177. isory mode qsub export LDR_CNTRL LARGE PAGE DATA Y path to exe bigprog D 8 14 Checking License Availability You can check to see where licenses are available You can do either of the following e Display license information for the current host qstat Bf e Display resources available including licenses on all hosts qmgr Qmgr print node default When looking at the server s license_count attribute use the sum of the Avail_Global and Avail_Local values 200 PBS Professional 10 4 User s Guide Chapter 9 Job Arrays This chapter describes job arrays and their use A job array represents a collection of jobs which only differ by a single index parameter The pur pose of a job array is twofold It offers the user a mechanism for grouping related work making it possible to submit query modify and display the set as a single unit Second it offers a way to possibly improve perfor mance because the batch system can use certain known aspects of the col lection for speedup 9 1 Definitions Subjob Individual entity within a job array e g 1234 7 where 1234 is the job array itself and 7 is the index which has many properties of a job as well as additional semantics defined below PBS Professional 10 4 User s Guide 201 Chapter 9 Job Arrays Sequence_number The numeric part of a job or job array identifier e g 1234 Subjob index The unique index which differentiates one subjob from a
178. ist If either select or place is missing from the job s new resource list it will be automatically generated using any newly inherited default values Example Given the following set of queue and server default values Server resources default ncpus 1 Queue QA resources default ncpus 2 default_chunk mem 2gb Queue QB default _chunk mem 1lgb no default for ncpus The following illustrate the equivalent select specification for jobs submit ted into queue QA and then moved to or submitted directly to queue QB qsub l ncpus 1 lmem 4gb In QA select 1 ncpus 1 mem 4gb No defaults need be applied In QB select 1 ncpus 1 mem 4gb PBS Professional 10 4 User s Guide 51 Chapter 4 Submitting a PBS Job No defaults need be applied qsub 1 ncpus 1 In QA select 1 ncpus 1 mem 2gb Picks up 2gb from queue default chunk and ncpus from qsub In QB select 1 ncpus 1 mem lgb Picks up 1gb from queue default chunk and 1 ncpus from qsub qsub lmem 4gb In QA select 1 ncpus 2 mem 4gb Picks up 2 ncpus from queue level job wide resource default and 4gb mem from qsub In QB select 1 ncpus 1 mem 4gb Picks up 1 ncpus from server level job wide default and 4gb mem from qsub qsub 1 nodes 4 In QA select 4 ncpus 1 mem 2gb Picks up a queue level default memory chunk of 2gb This is not 4 ncpus 2 because in prior versions nodes x implied 1 CPU per node unless otherwise explicitly stated In QB select 4 ncpu
179. istbox shows a collection of command but tons whose labels describe criteria used for filtering the Jobs listbox con tents The list of jobs can be selected according to the owner of jobs Owners job state Job _ States name of the job Job_ Name type of hold placed on the job Hold_Types the account name associated with the job Account Name checkpoint attribute Checkpoint time the job is eligi ble for queueing execution Queue_Time resources requested by the job Resources priority attached to the job Priority and whether or not the job is rerunnable Rerunnable The selection criteria can be modified by clicking on any of the appropriate command buttons to bring up a selection box The criteria command but tons are accompanied by a Select Jobs button which when clicked will update the contents of the Jobs listbox based on the new selection criteria Note that only jobs that meet all the selected criteria will be displayed PBS Professional 10 4 User s Guide 101 Chapter 5 Using the xpbs GUI Finally to the right of the Jobs panel are the following command buttons for operating on selected job s detail modify delete hold release signal msg move order run rerun provides information about selected job s This functional ity can also be achieved by double clicking on a Jobs list box entry for modifying attributes of the selected job s for deleting the selected job s
180. ithout returning subjob identifiers Note qselect will not return any job arrays when the state selection s option restricts the set to R S T or U because a job array will never be in any of these states However qselect can be used to return a list of subjobs by using the t option Options to qselect can be combined For example to restrict the selection to subjobs use both the J and the T options To select only running sub jobs use J T sR Table 9 5 Options to qselect for Job Arrays Option Selects Result none jobs Shows job and job array identifiers job arrays J job arrays Shows only job array identifiers T jobs Shows job and subjob identifiers subjobs 220 PBS Professional 10 4 User s Guide Job Arrays Chapter 9 9 9 Job Arrays and xpbs xpbs does not support job arrays 9 10 More on Job Arrays 9 10 1 Job Array Run Limits Jobs and subjobs are treated the same way by job run limits For example if max_user_run is set to 5 a user can have a maximum of 5 subjobs and or jobs running 9 10 2 Starving A job array s starving status is based on the queued portion of the array This means that if there is a queued subjob which is starving the job array is starving A running subjob retains its starving status when it was started 9 10 3 Job Array Dependencies Job dependencies are supported between job arrays and job arrays between job arrays and
181. job limits and counts as follows Maximum Total Queued Running Held Waiting Transiting and Exit ing The last column gives the status of the Server itself active idle or scheduling qstat B Server Max Tot Que Run Hld Wat Trn Ext Status fast domain 0 14 13 1 0 0 0 0 Active PBS Professional 10 4 User s Guide 139 Chapter 7 Checking Job System Status When querying jobs Servers or queues you can add the option to qstat to change the display to the full or long display For example the Server status shown above would be expanded using as shown below qstat Bf Server fast mydomain com server state Active True total _jobs 14 scheduling state count Transit 0 Queued 13 Held 0 Waiting 0 Running 1 Exiting 0 managers userl fast mydomain com default _queue workg log_events 511 mail from adm query other jobs True resources available mem 64mb resources available ncpus 2 resources default ncpus 1 resources assigned ncpus 1 resources assigned nodect 1 scheduler iteration 600 pbs _ version PBSPro 10 4 41640 7 1 4 Checking Queue Status The Q option to qst at displays the status of all or any specified queues at the optionally specified PBS Server One line of output is gen erated for each queue queried The three letter abbreviations correspond to 140 PBS Professional 10 4 User s Guide Checking Job System Status Chapter 7 limits
182. k 4 on VnodeB OMP_NUM_THREADS 1 NCPUS 1 e For PBS task 5 on VnodeC OMP_NUM_THREADS 1 NCPUS 1 e For PBS task 6 on VnodeC OMP_NUM_THREADS 1 NCPUS 1 To run two threads on each of N chunks each running a process all on the same Altix qsub l select N ncpus 2 l place pack This starts N processes on a single host with two OpenMP threads per pro cess because OMP NUM_THREADS 2 230 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 MPI Jobs with PBS PBS creates one MPI process per chunk For most implementations of the Message Passing Interface MPI you would use the mpi run command to launch your application For example here is a sample PBS script for an MPI job PBS l select arch linux mpirun np 32 machinefile S PBS NODEFILE a out 10 6 1 MPICH Jobs With PBS For users of PBS with MPICH on Linux the mpirun command has been changed slightly The syntax and arguments are the same except for one option which should not be set by the user machinefile file PBS supplies the machinefile If the user tries to specify it PBS will print a warning that it is replacing the machinefile Example of using mpirun PBS 1 select arch linux mpirun a out Under Windows the localroot option to MPICH s mpirun command may be needed in order to allow the job s processes to run more efficiently 10 6 2 MPI Jobs Using LAM MPI The pbs_mpilam command follows the conventio
183. l Jobs can share the memory on a node board if they have a CPU from that nodeboard To fit as many small jobs as possible onto vnodes that already have shared jobs on them request sharing in the job resource requests PBS will try to put a job that will fit in a single nodeboard on just one node board However if the only CPUs available are on separate nodeboards and those vnodes are not allocated exclusively to existing jobs and the job can share a vnode then the job will be run on the separate nodeboards If a job is suspended its processes will be moved to the global cpuset When the job is restarted they are restored 10 8 PVM Jobs with PBS On a typical system to execute a Parallel Virtual Machine PVM program you can use the pvmexec command The pymexec command expects a hostfile argument for the list of hosts on which to spawn the parallel job 266 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 For example here is a sample PBS script for a PVM job PBS N pvmjob pvmexec a out inputfile data_in To start the PVM daemons on the hosts listed in PBS_NODEFILE start the PVM console on the first host in the list and print the hosts to the stan dard output file named jobname o lt PBS jobID gt use echo conf pvm PBS_NODEFILE To quit the PVM console but leave the PVM dae mons running use quit To stop the PVM daemons restart the PVM console and quit use echo hal
184. l user input is shown in this bold fixed width font manpage xX Following UNIX tradition manual page references include the corresponding section number in parentheses appended to the man page name terms xii PBS Professional 10 4 User s Guide Words or terms being defined as well as variable names are in italics PBS Professional 10 4 User s Guide xiii XIV PBS Professional 10 4 User s Guide Chapter 1 Introduction This book the User s Guide to PBS Professional is intended as your knowledgeable companion to the PBS Professional software The informa tion herein pertains to PBS in general with specific information for PBS Professional 10 4 1 1 Book Organization This book is organized into 10 chapters plus two appendices Depending on your intended use of PBS some chapters will be critical to you and oth ers may be safely skipped Chapter 1 gives an overview of this book PBS and the PBS team PBS Professional 10 4 User s Guide 1 Chapter 1 Introduction Chapter 2 discusses the various components of PBS and how they interact followed by definitions of terms used in PBS and in distributed workload management Chapter 3 introduces PBS describing both user interfaces and sug gested settings to the user s environment Chapter 4 describes the structure and components of a PBS job and explains how to create and submit a PBS job Chapter 5 introduces the xpbs graphical user interface a
185. ler uses walltime to predict when resources will become available Therefore it is useful to request a reasonable walltime for each job 4 5 9 Submitting Jobs with Resource Specification Old Syntax If neither a node specification nor a selection directive is specified then a selection directive will be created requesting 1 chunk with resources speci fied by the job and with those from the queue or server default resource list These are ncpus mem arch host and software as well as any other default resources specified by the administrator For example a job submitted with qsub l ncpus 4 mem 123mb arch linux will have the following selection directive created select 1 ncpus 4 mem 123mb arch linux Do not mix old style resource or node specification with the select and place statements Do not use one in a job script and the other on the com mand line This will result in an error 50 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 4 5 10 Moving Jobs From One Queue to Another If the job is moved from the current queue to a new queue any default resources in the job s resource list that were contributed by the current queue are removed This includes a select specification and place directive generated by the rules for conversion from the old syntax Ifa job s resource is unset undefined and there exists a default value at the new queue or server that default value is applied to the job s resource l
186. listed in the process group file procgrp 250 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 qsub 1 select 2 ncpus 1 echo pbs hostl 1 username x y a exe argl arg2 gt procgrp echo pbs host2 1 username x x b exe argl arg2 gt gt procgrp MPICH MX HOME bin mpirun mpd pg procgrp path myprog x 1200 rm f procgrp D lt job id gt mpirun mpd will print a warning message warning pg is allowed but it is up to user to make sure only PBS hosts are specified MPI processes spawned are not guaranteed to be under PBS control The warning is issued because if any of the hosts listed in procgrp are not under the control of PBS then the processes on those hosts will not be under the control of PBS 10 6 9 PBS Jobs with MPICH2 s mpirun PBS provides an interface to MPICH2 s mpirun If executed inside a PBS job this allows for PBS to track all MPICH2 processes so that PBS can perform accounting and have complete job control If executed outside of a PBS job it behaves exactly as if standard MPICH2 s mpirun had been used You use the same mpirun command as you would use outside of PBS PBS Professional 10 4 User s Guide 251 Chapter 10 Multiprocessor Jobs When submitting PBS jobs that invoke the pbsrun wrapper script for MPICH2 s mpirun be sure to explicitly specify the actual number of ranks or MPI tasks in the qsub select specification Otherwise jobs will fail to
187. llocate 2 CPUs and 50GB of memory on a host named zool and The value of place depends on the default which defaults to place free 1l select 1 ncpus 2 mem 50gb host zooland 9 This will allocate 1 CPU and 6GB of memory and one host locked swlicense from each of two hosts 1 select 2 ncpus 1 mem 6gb swlicense 1 lplace scatter 10 Request free placement of 10 CPUs across hosts 1l select 10 ncpus 1 l place free 11 Here is an odd sized job that will fit on a single Altix but not on any one node board We request an odd number of CPUs that are not shared so they must be rounded up 1 select 1 ncpus 3 mem 6gb l place pack excl 12 Here is an odd sized job that will fit on a single Altix but not on any one node board We are asking for small number of CPUs but a large amount of memory 1l select 1 ncpus 1 mem 25gb l place pack excl 13 Here is a job that may be run across multiple Altix systems packed into the fewest vnodes 1l select 2 ncpus 10 mem 12gb l place free 14 Submit a job that must be run across multiple Altix systems packed into the fewest vnodes 1 select 2 ncpus 10 mem 12gb PBS Professional 10 4 User s Guide 61 Chapter 4 Submitting a PBS Job l place scatter 15 Request free placement across nodeboards within a single host 1 select 1 ncpus 10 mem 10gb l place group host 16 Request free placement across vnodes on multiple Altixes 1l select 10 ncpus 1 mem lgb
188. lnodes 1 fserver 15 noserver is converted to the equivalent 1l select 1 ncpus 1 fserver True 15 ncpus 1 noserver True 64 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 l place scatter but could also be more easily specified with something like 1l select 1 ncpus 1 fserver True 15 ncpus 1 fserver False l place scatter 9 Allocate 4 vnodes each with 6 CPUs with 3 MPI processes per vnode with each vnode on a separate host The memory allocated would be one fourth of the memory specified by the queue or server default if one existed This results in a different placement of the job from ver sion 5 4 1 nodes 4 ppn 3 ncpus 2 is converted to 1 select 4 ncpus 6 mpiprocs 3 l place scatter 10 Allocate 4 vnodes from 4 separate hosts with the property blue The amount of memory allocated from each vnode is 2560MB 10GB 4 rather than 10GB from each vnode 1l nodes 4 blue ncpus 2 1l mem 10GB is converted to 1l select 4 blue True ncpus 2 mem 2560mb lplace scat ter 4 8 Backward Compatibility For backward compatibility a legal node specification or resource specifi cation will be converted into selection and placement directives Specify ing cpp is part of the old syntax and should be replaced with nepus Do not mix old style resource or node specification syntax with select and place statements Ifa job is submitted using l select on the command line and it contain
189. lt ret gt directives tasks ctrl D 4 5 Requesting Resources PBS provides built in resources and allows the administrator to define cus tom resources The administrator can specify which resources are avail able on a given vnode as well as at the queue or server level e g floating licenses See Built in Resources on page 336 of the PBS Professional Reference Guide for a listing of built in resources Resources defined at the queue or server level apply to an entire job If they are defined at the vnode level they apply only to the part of the job running on that vnode Jobs request resources which are allocated to the job along with any defaults specified by the administrator PBS Professional 10 4 User s Guide 43 Chapter 4 Submitting a PBS Job Custom resources are used for application licenses scratch space etc and are defined by the administrator See Chapter 5 Customizing PBS Resources on page 193 of the PBS Professional Administrator s Guide Custom resources are used the same way built in resources are used Jobs request resources in two ways They can use the select statement to define chunks and specify the quantity of each chunk A chunk is a set of resources that are to be allocated as a unit Jobs can also use a job wide resource request which uses resource value pairs outside of the select statement The qsub qalter and pbs_rsub commands are used to request resources However custom res
190. m 9gb In the above mem 2gb and ncpus 1 are inherited from default_chunk The job wide resource request is checked against queue resource defaults then against server resource defaults Ifa default resource is defined which is not specified in the resource request it is added to the resource request 4 5 6 Requesting Application Licenses Application licenses are set up as resources defined by the administrator PBS doesn t actually check out the licenses the application being run inside the job s session does that 4 5 6 1 Floating Licenses PBS queries the license server to find out how many floating licenses are available at the beginning of each scheduling cycle If you wish to request a site wide floating license it will typically have been set up as a server level job wide resource To request an application license called AppF use qsub l AppF lt number of licenses gt lt other qsub arguments gt 48 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 If only certain hosts can run the application they will typically have a host level boolean resource set to True To request the application license and the vnodes on which to run the application use qsub l AppF lt number of licenses gt lt other qsub arguments gt 1 select haveAppF True PBS doesn t actually check out the licenses the application being run inside the job s session does that 4 5 6 2 Node locked Licenses Per host nod
191. mer may acquire a copy of PBS Professional with the standard security and authentication module replaced with the KRB5 module This is not supported under Windows To use a forwardable renewable Kerberos V5 TGT specify the Ww cred krb5 option to qsub This will cause qsub to check the user s credential cache for a valid forwardable renewable TGT which it will send to the Server and then eventually to the execution MOM While it s at the Server and the MOM this TGT will be periodically refreshed until either the job finishes or the maximum refresh time on the TGT is exceeded whichever comes first If the maximum refresh time on the TGT is exceeded no KRBS services will be available to the job even though it will continue to run 8 13 Support for Large Page Mode on AIX A process running as part of a job can use large pages The memory reported in resources used mem may be larger with large page sizes You can set an environment variable to request large memory pages LDR_CNTRL LARGE PAGE DATA M LDR_CNTRL LARGE PAGE DATA Y For more information see the man page for setpcred This can be viewed with the command man setpcred on an AIX machine PBS Professional 10 4 User s Guide 199 Chapter 8 Advanced PBS Features You can run a job that requests large page memory in mandatory mode qsub export LDR_CNTRL LARGE PAGE DATA M path to exe bigprog D You can run a job that requests large page memory in adv
192. mitted with undefined resources specified either with 1 select or with l nodes will not be rejected at submission The job will be aborted upon being enqueued in an execution queue if the resources are still undefined This preserves backward compatibility 4 5 13 Limits on Resource Usage Each chunk s per chunk limits determine how much of any resource can be used in that chunk Per chunk resource usage limits are established by per chunk resources both from explicit requests and from defaults PBS Professional 10 4 User s Guide 53 Chapter 4 Submitting a PBS Job Job resource limits set a limit for per job resource usage Job resource lim its are established both by requesting job wide resources and by summing per chunk consumable resources Job resource limits from sums of all chunks including defaults override those from job wide defaults Limits include both explicitly requested resources and default resources If a job s job resource limit exceeds queue or server restrictions it will not be put in the queue or accepted by the server If while running a job exceeds its limit for a consumable or time based resource it will be termi nated See The PBS Professional Administrator s Guide Job limits are created from the directive for each consumable resource For example qsub lselect 2 ncpus 3 mem 4gb arch linux will have the following job limits set e ncpus 6 e mem 8gb 4 5 14 Resource Permissions Custom resou
193. mmand line that option and its argument if any will be taken from there 4 10 A Sample PBS Jobs The following is an example of a job script written in Python This script calculates the 10th Fibonacci number Q cat job py PBS 1 select 1 ncpus 3 mem lgb PBS N myjob def fibo n global fibo if n lt 2 return n else return fibo n 1 fibo n 2 print fibo 10 d fibo 10 Note that this script contains PBS directives PBS Professional 10 4 User s Guide 69 Chapter 4 Submitting a PBS Job Let s look at an example PBS job in detail UNIX Linux bin sh PBS 1 walltime 1 00 00 PBS 1 select mem 400mb PBS j oe date my_application date Windows PBS 1 walltime 1 00 00 PBS 1 select mem 400mb PBS j oe date t my application date t On line one in the example above Windows does not show a shell directive The default on Windows is the batch command language Also note that it is possible under both Windows and UNIX to specify to PBS the scripting language to use to interpret the job script see the S option to gsub in section 4 13 9 Specifying Scripting Language to Use on page 82 The Windows script will be a exe or bat file Lines 2 8 of both files are almost identical The primary differences will be in file and directory path specification such as the use of drive letters and slash vs backslash as the path separator Lines 2 4 are PBS directives PB
194. n t altered before the job uses it so PBS requeues it PBS requeues some jobs that are terminated before execution Two exam ples of this are multi host jobs where the job did not start on one or more execution hosts and provisioning jobs for which the provisioning script failed 4 13 9 Specifying Scripting Language to Use The S path_list option declares the path and name of the scripting language to be used in interpreting the job script The option argument path_list is in the form path host path host Only one path may be specified for any host named and only one path may be specified without the corresponding host name The path selected will be the one with the host name that matched the name of the execution host If no matching host is found then the path specified without a host will be selected if present If the S option is not specified the option argument is the null string or no entry from the path_list is selected then PBS will use the user s login shell on the execution host Example 1 Using bash via a directive PBS S bin bash mars usr bin bash jupiter Example 2 Running a Python script from the command line on UNIX Linux qsub S opt pbs default bin pbs python lt script name gt Example 3 Running a Python script from the command line on Windows qsub S C Program Files PBS Pro exec bin pbs python exe lt script name gt 82 PBS Professional 10 4 User s Guide Submitti
195. n page 84 I Interactive batch Jobs on page 91 J X Y Z Job Array on page 201 j join Merging Output and Error Files on page 89 k keep Retaining Output and Error Files on Execution Host on page 90 l resource list section 4 3 1 Rules for Submitting Jobs on page 35 M user list Setting Email Recipient List on page 80 m MailOptions Specifying Email Notification on page 80 N name Specifying a Job Name on page 81 o path Redirecting Output and Error Files on page 78 p priority Setting a Job s Priority on page 83 q destination Specifying Queue and or Server on page 77 r value Marking a Job as Rerunnable or Not on page 81 S path list Specifying Scripting Language to Use on page 82 u user list Specifying Job User ID on page 86 V Exporting Environment Variables on page 79 76 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 Table 4 2 Options to the qsub Command Option Function and Page Reference v variable list Expanding Environment Variables on page 79 W depend list Specifying Job Dependencies on page 159 W group list list Specifying Job Group ID on page 88 W stagein list Input Output File Staging on page 163 W stageout list Input Output File Staging on page 163 W cred dce Running PBS
196. n Fault Tolerance PBS attempts to reconfirm reservations for which associated vnodes have become unavailable See section 8 8 8 1 1 Reservation Fault Tolerance on page 192 PBS Professional 10 4 User s Guide 15 Chapter 3 Getting Started With PBS 3 4 New Features in Recent Releases 3 4 1 Path to Binaries 10 0 The path to the PBS binaries may have changed for your system If the old path was not one of opt pbs usr pbs or usr local pbs you may need to add opt pbs default to your PATH environment variable 3 4 2 Using job_sort_ key 10 0 The sort_priority option to job_sort_key is replaced with the job_ priority option 3 4 3 Job Specific Staging and Execution Directories 9 2 PBS can now provide a staging and execution directory for each job Jobs have new attributes sandbox and jobdir the MOM has a new option jobdir root and there is a new environment variable called PBS _JOBDIR If the job s sandbox attribute is set to PRIVATE PBS creates a job specific staging and execution directory If the job s sand box attribute is unset or is set to HOME PBS uses the user s home direc tory for staging and execution which is how previous versions of PBS behaved See section 8 6 Input Output File Staging on page 163 3 4 4 Standing Reservations 9 2 PBS now provides a facility for making standing reservations A standing reservation is a series of advance reservations The pbs_rsub command is used
197. n of LAM s mpirun The nodes here are LAM nodes LAM s mpirun has two syntax forms pbs_mpilam mpirun global_options lt where gt lt program gt args pbs_mpilam mpirun global_options lt schema file gt PBS Professional 10 4 User s Guide 231 Chapter 10 Multiprocessor Jobs Where lt where gt is a set of node and or CPU identifiers indicating where to start lt program gt Nodes n lt list gt e g n0 3 5 CPUS e lt list gt 3 c0 3 5 Extras h local node o origin node N all nodes C all CPUs lt schema file gt is an ASCII file containing a description of the pro grams which constitute an application The first form is fully supported by PBS all user MPI processes are tracked The second form is supported but user MPI processes are not tracked CAUTION Keep in mind that if the lt where gt argument and global option np or c are not specified in the command line then pbs_mpilam will expect an ASCII schema file as argument 10 6 3 MPI Jobs Using AIX POE PBS users of AIX machines running IBM s Parallel Operating Environ ment or POE can run jobs on the HPS using either IP or US mode PBS will manage the HPS PBS can track the resources for MPI LAPI pro grams or a mix of MPI and LAPI programs LoadLeveler is not required in order to use InfiniBand switches in User Space mode PBS works with a standard InfiniBand configuration Any job that can run under IBM poe can run under PBS
198. name Sess NDS TSK Mem Time S Time 63 south bob workq airfoil 1 0 13 Q 54 south bob workq twinkie 1 0 20 Q To change the order of two jobs using xpbs select the two jobs and then click the order button The gorder command can only be used with job array objects not on sub jobs or ranges This will change the queue order of the job array in associ ation with other jobs or job arrays in the queue 6 7 Moving Jobs Between Queues PBS provides the qmove command to move jobs between different queues even queues on different Servers To move a job is to remove the job from the queue in which it resides and instantiate the job in another queue IMPORTANT A job in the running state cannot be moved The usage syntax of the qmove command is qmove destination job_identifier s Job array job identifiers must be enclosed in double quotes 128 PBS Professional 10 4 User s Guide Working With PBS Jobs Chapter 6 The first operand is the new destination for queue server queue server If the dest ination operand describes only a queue then qmove will move jobs into the queue of the specified name at the job s current Server If the dest ination operand describes only a Server then qmove will move jobs into the default queue at that Server If the destination operand describes both a queue and a Server then qmove will move the jobs into the specified queue at the specified Server All following opera
199. nd If executed inside a PBS job this allows for PBS to track all MVAPICH2 processes so that PBS can perform accounting and have complete job control If executed outside of a PBS job it behaves exactly as if standard MVAPICH2 s mpiexec had been used You use the same mpiexec command as you would use outside of PBS The maximum number of ranks that can be launched is the number of entries in PBS_NODEFILE 262 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 12 1 Options If executed inside a PBS job script all of the options to the PBS interface are the same as MVAPICH2 s mpiexec except for the following host The host option is ignored machinefile lt file gt The file option is ignored mpdboot If mpdboot is not called before mpiexec it is called auto matically before mpiexec runs so that an MPD daemon is started on each host assigned by PBS 10 6 12 2 MPD Startup and Shutdown The interface ensures that the MPD daemons are started on each of the hosts listed in the S PBS_NODEFILE It also ensures that the MPD dae mons are shut down at the end of MPI job execution 10 6 12 3 Examples Example 1 Run a single executable MVAPICH2 job with 6 ranks on hosts listed in S PBS_NODEFILE PBS NODEFILE pbs host1l pbs host2 pbs host3 Job script mpiexec np 6 path mpiprog PBS Professional 10 4 User s Guide 263 Chapter 10 Multiprocessor Jobs Run job script qsub l select 3 nc
200. nd shows how to submit a PBS job using xpbs Chapter 6 describes how to check status of a job and request status of queues vnodes systems or PBS Servers Chapter 7 discusses commonly used commands and features of PBS and explains how to use each one Chapter 8 describes and explains how to use the more advanced fea tures of PBS Chapter 9 describes and explains the job array features in PBS Chapter 10 explains how PBS interacts with multi vnode and parallel applications and illustrates how to run such applications under PBS Appendix A provides a quick reference summary of PBS environment variables Appendix B includes information for converting from NQS NQE to PBS 2 PBS Professional 10 4 User s Guide Introduction Chapter 1 1 2 Supported Platforms For a list of supported platforms see the Release Notes 1 3 What is PBS Professional PBS Professional is the professional version of the Portable Batch System PBS a flexible workload management system originally developed to manage aerospace computing resources at NASA PBS has since become the leader in supercomputer workload management and the de facto stan dard on Linux clusters Today growing enterprises often support hundreds of users running thou sands of jobs across different types of machines in different geographical locations In this distributed heterogeneous environment it can be extremely difficult for administrators to collect detaile
201. nds are job_identifiers which specify the jobs to be moved to the new destination To move jobs between queues or between Servers using xpbs select the job s of interest and then click the move button Doing so will launch the Move Job dialog box from which you can select the queue and or Server to which you want the job s moved The qmove command can only be used with job array objects not with subjobs or ranges Job arrays can only be moved from one server to another if they are in the Q H or W states and only if there are no running subjobs The state of the job array object is preserved in the move The job array will run to completion on the new server As with jobs a qstat on the server from which the job array was moved will not show the job array A qstat on the job array object will be redirected to the new server Note The subjob accounting records will be split between the two servers PBS Professional 10 4 User s Guide 129 Chapter 6 Working With PBS Jobs 6 8 Converting a Job into a Reservation Job The pbs_rsub command can be used to convert a normal job into a reser vation job that will run as soon as possible PBS creates a reservation queue and a reservation and moves the job into the queue Other jobs can also be moved into that queue via qmove 1B or submitted to that queue via qsub 1B The reservation is called an ASAP reservation The format for converting a normal job into a reserv
202. ng a PBS Job Chapter 4 4 13 9 1 Windows Caveats Using this option under Windows is more complicated because if you change from the default shell of cmd then a valid PATH is not automati cally set Thus if you use the S option under Windows you must explicitly set a valid PATH as the first line of your job script 4 13 10 Setting a Job s Priority The p priority option defines the priority of the job The priority argument must be an integer between 1024 lowest priority and 1023 highest priority inclusive The default is no priority which is equivalent to a priority of zero This option allows the user to specify a priority for their jobs However this option is dependant upon the local scheduling policy By default the sort jobs by job priority feature is disabled If your local PBS administra tor has enabled it then all queued jobs will be sorted based on the user specified priority If you need an absolute ordering of your own jobs see Specifying Job Dependencies on page 159 qsub p 120 my_job PBS p 300 4 13 11 Deferring Execution The a date_time option declares the time after which the job is eli gible for execution The date_time argument is in the form CC YY MM DD hhmm SS where CC is the first two digits of the year the century YY is the second two digits of the year MM is the two digits for the month DD is the day of the month hh is the hour mm is the min
203. ng requirements Automatic Load Leveling provides numerous ways to distribute the work load across a cluster of machines based on hardware configuration resource availability keyboard activity and local scheduling policy 6 PBS Professional 10 4 User s Guide Introduction Chapter 1 Distributed Clustering allows customers to utilize physically distributed systems and clusters even across wide area networks Common User Environment offers users a common view of the job submis sion job querying system status and job tracking over all systems Cross System Scheduling ensures that jobs do not have to be targeted to a specific computer system Users may submit their job and have it run on the first available system that meets their resource requirements Job Priority allows users the ability to specify the priority of their jobs defaults can be provided at both the queue and system level Username Mapping provides support for mapping user account names on one system to the appropriate name on remote server systems This allows PBS to fully function in environments where users do not have a consistent username across all hosts Fully Configurable PBS was designed to be easily tailored to meet the needs of different sites Much of this flexibility is due to the unique design of the scheduler module which permits significant customization Broad Platform Availability is achieved through support of Windows and every major version
204. ng reserva tion as a whole not to a specific occurrence You can only specify when the job is eligible to run See the qsub 1B man page e when an advance reservation ends it and all of its jobs running or queued are deleted but when an occurrence ends only its running jobs are deleted Each occurrence of a standing reservation has reserved resources which satisfy the resource request but each occurrence may have its resources drawn from a different source A query for the resources assigned to a standing res ervation will return the resources assigned to the soonest occurrence shown in the resv_nodes attribute reported by pbs_rstat Soonest occurrence of a standing reservation The occurrence which is currently active or if none is active then it is the next occurrence Degraded reservation An advance reservation for which one or more associated vnodes are unavailable A standing reservation for which one or more vnodes asso ciated with any occurrence are unavailable PBS Professional 10 4 User s Guide 179 Chapter 8 Advanced PBS Features 8 8 2 Introduction to Creating and Using Reservations The user creates both advance and standing reservations using the pbs_rsub command PBS either confirms that the reservation can be made or rejects the request Once the reservation is confirmed PBS cre ates a queue for the reservation s jobs Jobs are then submitted to this queue When a reservation is confirmed it me
205. nment By accepting delivery of the Products except as pro vided below the government or the party procuring the Products under government funding hereby agrees that the Products qualify as ccommer PBS Professional 10 4 User s Guide 309 cial computer software as that term is used in the acquisition regulations applicable to this procurement and that the government s use and disclosure of the Products is controlled by the terms and conditions of this Agreement to the maximum extent possible This Agreement supersedes any contrary terms or conditions in any statement of work contract or other document that are not required by statute or regulation If any provision of this Agreement is unacceptable to the government Vendor may be contacted at Altair Engineering Inc 1820 E Big Beaver Road Troy MI 48083 2031 telephone 248 614 2400 If any provision of this Agreement violates applicable federal law or does not meet the government s actual minimum needs the government agrees to return the Products for a full refund For procurements governed by DFARS Part 227 72 OCT 1998 the Software except as described below is provided with only those rights specified in this Agreement in accordance with the Rights in Commercial Computer Software or Commercial Computer Software Documentation policy at DFARS 227 7202 3 a OCT 1998 For procurements other than for the Department of Defense use reproduction or disclosure of the Soft
206. nother This must be a non negative integer Job array identifier The identifier returned upon success when submitting a job array The format is sequence_number or sequence number server domain com Job array range A set of subjobs within a job array When specifying a range indices used must be valid members of the job array s indices 9 1 1 Description A job array is a compact representation of one or more jobs called subjobs when part of a Job array which have the same job script and have the same values for all attributes and resources with the following exceptions e each subjob has a unique index e Job Identifiers of subjobs only differ by their indices e the state of subjobs can differ All subjobs within a job array have the same scheduling priority A job array is submitted through a single command which returns on suc cess a job array identifier with a server unique sequence number Sub job indices are specified at submission time These can be e acontiguous range e g 1 through 100 e arange with a stepping factor e g every second entry in 1 through 100 1 3 5 99 202 PBS Professional 10 4 User s Guide Job Arrays Chapter 9 A job array identifier can be used e by itself to represent the set of all subjobs of the job array e with a single index a job array identifier to represent a single subjob e with a range a job array range to represent the subjobs designat
207. o default value is provided by the PBS interface It is up to the local mpirun to decide what the reasonable default value should PBS Professional 10 4 User s Guide 257 Chapter 10 Multiprocessor Jobs be which is usually 1 The maximum number of ranks that can be launched is the number of entries in PBS_NODEFILE 10 6 10 2 MPD Startup and Shutdown Intel MPI s mpirun takes care of starting stopping the MPD daemons The PBS interface to Intel MPI s mpirun always passes the arguments totalnum lt number of mpds to start gt and file lt mpd_hosts_ file gt to the actual mpirun taking its input from unique entries in PBS_NODEFILE 10 6 10 3 Examples Example 1 Runa single executable Intel MPI job with 6 processes spread out across the PBS allocated hosts listed in PBS_NODEFILE PBS NODEFILE pbs hostl pbs host2 pbs host3 pbs hostl pbs host2 pbs host3 Job script mpirun takes care of starting the MPD daemons on unique hosts listed in SPBS NODEFILE and also runs 6 processes mapped to each host listed in SPBS NODEFILE mpirun takes care of shutting down MPDs mpirun path myprog x 1200 258 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 Run job script qsub l select 3 ncpus 2 job script lt job id gt Example 2 Run an Intel MPI job with multiple executables on multiple hosts using PBS_NODEFILE and mpiexec arguments to mpirun PBS_NODEFILE hostA hostA hostB hostB h
208. o not write the job identifier assigned to the job to the command s standard output qsub z my_job PBS z 4 13 20 Specifying Staging and Execution Directory The W sandbox lt value gt option allows you to specify where PBS will stage files and execute the job script See section 8 6 Input Output File Staging on page 163 4 13 21 Interactive batch Jobs PBS provides a special kind of batch job called interactive batch An inter active batch job is treated just like a regular batch job in that it is queued up and has to wait for resources to become available before it can run Once it is started however the user s terminal input and output are con nected to the job in a matter similar to a Login session It appears that the user is logged into one of the available execution machines and the resources requested by the job are reserved for that job Many users find this useful for debugging their applications or for computational steering The I option declares that the job is an interactive batch job If the I option is specified on the command line the job is an interactive job If a script is given it will be processed for directives but any execut able commands will be discarded When the job begins execution all input to the job is from the terminal session in which qsub is running The I option is ignored in a script directive When an interactive job is submitted the gsub command will not termi
209. ob s start time If the job is still run ning you can do a qstat f and look for the stime attribute If the job has finished you look in the accounting log for the S record for the job For an array job only the S record is available 146 PBS Professional 10 4 User s Guide Checking Job System Status Chapter 7 7 1 16 Viewing Job Status in Wide Format The w qstat option displays job status in wide format The total width of the display is extended from 80 characters to 120 characters The Job ID column can be up to 30 characters wide while the Username Queue and Jobname column can be up to 15 characters wide The SessID column can be up to eight characters wide and the NDS column can be up to four char acters wide Note You can use this option only with the a n or s qstat options 7 1 17 Show state of Job Arrays The J option to qstat will show only the state of job arrays In combina tion with t qstat will show only the state of subjobs 7 1 18 Print Job Array Percentage Completed The p option to qstat prints the default display with a column for Per centage Completed For a job array this is the number of subjobs com pleted and deleted divided by the total number of subjobs 7 1 19 Getting Information on Jobs Moved to Another Server If your job is running at another server you can use the qstat command to see its status If your site is using peer scheduling your job may be
210. ob arrays that are queued running finished or moved use qstat xJ PBS Professional 10 4 User s Guide 149 Chapter 7 Checking Job System Status When information for a moved job is displayed the destination queue and server are shown as lt queue gt lt server gt Example 2 qstat x output for moved job destination queue is destq and destination server is server2 Job id Name User Time Use S Queue 101 sequoia STDIN userl 00 00 00 F workq 102 sequoia STDIN userl 00 00 00 M _ destq server2 103 sequoia STDIN userl 00 00 00 R workq Example 3 Viewing moved job There are three servers with hostnames ServerA ServerB and ServerC Userl submits job 123 to ServerA After some time Userl moves the job to ServerB After more time the administrator moves the job to QueueC at ServerC This means The qstat command will show QueueC ServerC for job 123 7 1 21 2 Job History In Alternate Format You can use the H option to the qstat command to see job history for finished or moved jobs in alternate format Usage qstat H Displays information for finished or moved jobs in alternate format qstat H job identifier Displays information for that job in alternate format whether or not it is finished or moved qstat H destination Displays information for finished or moved jobs at that destination 150 PBS Professional 10 4 User s Guide Checking Job System Status Chapter 7 Example 1 Job history
211. ocesses spread out across the PBS allocated hosts listed in PBS_NODEFILE PBS NODEFILE pbs host1l pbs host2 pbs host3 qsub 1 select 3 ncpus 1 mpirun ch_mx np 64 path myprog x 1200 D lt job id gt Example 2 Run an MPICH Mx job with multiple executables on multiple hosts listed in the process group file procgrp qsub 1 select 2 ncpus 1 echo pbs hostl 1 username x y a exe argl arg2 gt procgrp echo pbs host2 1 username x x b exe argl arg2 gt gt procgrp mpirun ch_ mx pg procgrp path myprog x rm f procgrp D lt job id gt mpirun ch_mx will give the warning message warning pg is allowed but it is up to user to make sure only PBS hosts are specified MPI processes spawned are not guaranteed to be under PBS control The warning is issued because if any of the hosts listed in procgrp are not under the control of PBS then the processes on those hosts will not be under the control of PBS PBS Professional 10 4 User s Guide 245 Chapter 10 Multiprocessor Jobs 10 6 7 PBS Jobs with MPICH GM s mpirun Using MPD mpirun mpd PBS provides an interface to MPICH GM s mpirun using MPD If exe cuted inside a PBS job this allows for PBS to track all MPICH GM pro cesses started by the MPD daemons so that PBS can perform accounting have and complete job control If executed outside of a PBS job it behaves exactly as if standard mpirun mpd with MPD had been used You use the same comman
212. of the user under which the job will be run Under Windows the primary group is the first group found for the user by PBS when querying the accounts database qsub W group list grpA grpB jupiter my_job 4 13 16 Specifying a Local Account The A account_string option defines the account string associ ated with the job The account_string is an opaque string of characters and is not interpreted by the Server which executes the job This value is often used by sites to track usage by locally defined account names IMPORTANT Under Unicos if the Account string is specified it must be a valid account as defined in the system User Data Base UDB qsub A Math312 my_job PBS A accountNumber 4 13 17 Merging Output and Error Files The 4 join option declares if the standard error stream of the job will be merged with the standard output stream of the job A join argument value of oe directs that the two streams will be merged intermixed as standard output A join argument value of eo directs that the two streams will be merged intermixed as standard error If the join argument is n or the option is not specified the two streams will be two separate files qsub j oe my _job PBS j eo PBS Professional 10 4 User s Guide 89 Chapter 4 Submitting a PBS Job 4 13 18 Retaining Output and Error Files on Execution Host The k keep option defines which if either of standard output STD OUT or standard e
213. ollowing table lists how the path to the executable can be specified Table 11 1 Executable Path Specification Path Specification Location of Executable You can specify an absolute path Anywhere available to the HPCBP to the executable Server You can specify a path relative to A path relative to your home direc your home directory on the HPC tory on the HPC Server Server You can specify just the name of The executable is in your PATH or the executable in your default working directory 11 4 Submitting HPC Basic Profile Jobs As with PBS jobs you do not need to specify destination specific parame ters 11 4 1 Restrictions on Submitting Jobs for Execution at HPCBP Server 11 4 1 1 Specifying Executable for Job The job must specify exactly one executable and its arguments This must be done on the gsub command line 272 PBS Professional 10 4 User s Guide HPC Basic Profile Jobs Chapter 11 11 4 1 2 HPCBP Jobs Run on One HPCBP Server The job must not be split across more than one HPCBP Server e It cannot be split across two or more HPCBP Servers e It cannot be split across an HPCBP Server and another node 11 4 1 3 Number of CPUs and mpiprocs For each chunk the aggregate number of requested ncpus must match the aggregate number of requested mpiprocs The default value per chunk for both ncpus and mpiprocs is 1 If you request 1 CPU per chunk you do not have to specify the mpiprocs
214. om Italy 39 0832 315573 support altairengineering it 39 800 905595 Japan 81 3 5396 2881 pbs altairjp co jp Korea 82 31 728 8600 support altair co kr Scandinavia 46 0 46 286 2050 support altair se UK 44 0 2476 323 600 support uk altair com This document is proprietary information of Altair Engineering Inc Table of Contents Acknowledgements ix Preface xi 1 Introduction l Ll Book Organization 0a 5454 ecient eas 1 1 2 Supported PISAUOEMNS 600 ca es etn sade owe See 3 1 3 What is PBS Professional cs i0 s 40 2444600 3 lot History of PBS sssccicoreitsu ngr nidd idaan 4 LS About the PBS Teas ocd dcacnudede ces ewe os 5 1 6 About Altair Engineering 05 5 Mey Why Use PDS ays 2 ones Se ons aoe Kees ee Koes 5 2 Concepts and Components 9 2h PBS Components 2 icck ied aes Ree a ee oe Ks 10 3 Getting Started With PBS 13 3 1 New Features in PBS Professional 10 4 14 3 2 New Features in PBS Professional 10 2 14 PBS Professional 10 4 User s Guide iii iv Table of Contents 3 3 New Features in Version 10 1 15 3 4 New Features in Recent Releases 16 3 5 SDEPRCCANONS cx oy kiahe koa eee Ve oe bo 35 6 dee 17 3 6 Backward Compatibility 0 18 3 7 Mss PBS wees ced aeons ooo eee ee ee eee eek 18 3 8 PBS Interfaces wi oie sev ee ee aweadwahedheoewa 19 3 9 Users PBS BNvitOnimenls lt i2340204e5 082a 0 21 3 10 Usernames
215. ommands and Finished Jobs The commands listed above cannot be used with finished jobs whether they finished at the local server or a remote server These jobs are no longer running PBS is storing their information and this information can not be deleted altered etc Trying to use one of the above commands with a finished job results in the following error message lt command name gt Job lt jobid gt has finished 6 9 5 Viewing Information for Finished and Moved Jobs You can view information for finished and moved jobs in the same way as for queued and running jobs as long as the job history is still being pre served The x option to the gstat command allows you to see information for all jobs whether they are running queued finished or moved This informa tion is presented in standard format The H option to the qstat com mand allows you to see alternate format information for finished or moved jobs only See section 7 1 21 Viewing Job History on page 149 6 9 5 1 UNIX Linux qstat fx qselect x s MF 6 9 5 2 Windows for F usebackq j in Program Files PBSPro exec bin qselect x s MF do Program Files PBS Pro exec bin qstat fx 33 PBS Professional 10 4 User s Guide 133 Chapter 6 Working With PBS Jobs 6 9 6 Listing Job Identifiers of Finished and Moved Jobs You can list identifiers of finished and moved jobs in the same way as for queued and running jobs as long as the jo
216. on 114 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 iconizeQueues View A boolean value true or false indicating whether or not to iconize the QUEUES region jconizeJobs View A boolean value true or false indicating whether or not to iconize the JOBS region iconizeInfo View A boolean value true or false indicating whether or not to iconize the INFO region jobResourceList A curly braced list of resource names as according to archi tecture known to xpbs The format is as follows lt arch typel gt resnamel resname2 resnameN lt arch type2 gt resnamel resname2 resnameN lt arch typeN gt resnamel resname2 resnameN PBS Professional 10 4 User s Guide 115 Chapter 5 Using the xpbs GUI 116 PBS Professional 10 4 User s Guide Chapter 6 Working With PBS Jobs This chapter introduces the reader to various commands useful in working with PBS jobs Covered topics include modifying job attributes holding and releasing jobs sending messages to jobs changing order of jobs within a queue sending signals to jobs and deleting jobs In each section below the command line method for accomplishing a particular task is presented first followed by the xpbs method 6 1 Modifying Job Attributes Most attributes can be changed by the owner of the job or a manager or operator while the job is still queued However once a job begins execu tion the only resource
217. on on converting from NQS to PBS see Appendix B The PBS story continued when MRJ Veridian the R amp D contractor that developed PBS for NASA released the Portable Batch System Profes sional Edition PBS Pro a commercial enterprise ready workload man agement solution Three years later the MRJ Veridian PBS Products business unit was acquired by Altair Engineering Inc Altair set up the PBS Products unit as a subsidiary company named Altair Grid Technolo gies focused on PBS Professional and related Grid software This unit then became part of Altair Engineering 4 PBS Professional 10 4 User s Guide Introduction Chapter 1 1 5 About the PBS Team The PBS Professional product is developed by the same team that origi nally designed PBS for NASA In addition to the core engineering team Altair Engineering includes individuals who have supported PBS on com puters around the world including some of the largest supercomputers in existence The staff includes internationally recognized experts in resource management and job scheduling supercomputer optimization message passing programming parallel computation and distributed high performance computing In addition the PBS team includes co architects of the NASA Metacenter the first full production geographically distrib uted meta computing grid co architects of the Department of Defense MetaQueueing prototype Grid Project co architects of the NASA Infor mation Power Grid
218. onal 10 4 User s Guide 243 Chapter 10 Multiprocessor Jobs not be under the control of PBS 10 6 6 PBS Jobs with MPICH MX s mpirun Using rsh ssh mpirun ch_mx PBS provides an interface to MPICH MX s_ mpirun using rsh ssh If executed inside a PBS job this allows for PBS to track all MPICH Mx pro cesses started by rsh ssh so that PBS can perform accounting and has com plete job control If executed outside of a PBS job it behaves exactly as if standard mpirun ch_mx had been used You use the same command as you would use outside of PBS either mpirun ch_mx or mpirun 10 6 6 1 Options Inside a PBS job script all of the options to the PBS interface are the same as mpirun ch_mx except for the following machinefile lt file gt The file argument contents are ignored and replaced by the contents of the PBS_NODEFILE Np If not specified the number of entries found in the PBS_NODEFILE is used The maximum number of ranks that can be launched is the number of entries in PBS_NODEFILE pg The use of the pg option for having multiple executables on multiple hosts is allowed but it is up to user to make sure only PBS hosts are specified in the process group file MPI processes spawned on non PBS hosts are not guaranteed to be under the control of PBS 244 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 6 2 Examples Example 1 Run a single executable MPICH MX job with 64 pr
219. onfirmed reservation R2 and an occurrence of R1 and an occurrence of R2 overlap in time and share a vnode for which different AOEs are requested by the two occurrences 12 4 Using Provisioning 12 4 1 Requesting Provisioning You request a reservation with an AOE in order to reserve the resources and AOE required to run a job You request an AOE for a job if that job requires that AOE You request provisioning for a job or reservation using the same syntax You can request an AOE for the entire job reservation l aoe lt AOE gt Example l aoe suse The lt AOE gt form cannot be used with select You can request an AOE for a single chunk job reservation l select lt chunk request gt aoe lt AOE gt Example ls select 1 ncpus 2 aoe rhel 294 PBS Professional 10 4 User s Guide Using Provisioning Chapter 12 You can request the same AOE for each chunk of a job reservation l select lt chunk request gt aoe lt AOE gt lt chunk request gt aoe lt AOE gt Example l select 1 ncpus 1 aoe suse 2 ncpus 2 aoe suse 12 4 2 Commands and Provisioning If you try to use PBS commands on a job that is in the provisioning sub state the commands behave differently The provisioning of vnodes is not affected by the commands if provisioning has already started it will con tinue The following table lists the affected commands Table 12 2 Effect of Commands on Jobs in Provisioning Substate Command Behavi
220. options If you don t have a existing script file to load into xpbs you can start typ ing the executable lines of the job in the file text box Next review the Destination listbox This box shows the queues found in the host that you selected A special entry called host refers to the default queue at the indicated host Select appropriately the destination queue for the job Next define any required resources in the Resource List subwindow The resources specified in the Resource List section will be job wide resources only In order to specify chunks or job placement use a script To run an array job use a script You will not be able to query individual subjobs or the whole job array using xpbs Type the script into the File entry box Do not click the Load button Instead use the Submit but ton PBS Professional 10 4 User s Guide 109 Chapter 5 Using the xpbs GUI Finally review the optional settings to see if any should apply to this job For example e Use the one of the buttons in the Output region to merge output and error files e Use Stdout File Name to define standard output file and to redirect output e Use the Environment Variables to Export subwindow to have current environment variables exported to the job e Use the Job Name field in the OPTIONS subwindow to give the job a name e Use the Notify email address and one of the buttons in
221. or While in Provisioning Substate qdel Without force Job is not deleted With force Job is deleted qsig s suspend Job is not suspended ghold Job is not held qrerun Job is not requeued qmove Cannot be used on a job that is provisioning qalter Cannot be used on a job that is provisioning qrun Cannot be used on a job that is provisioning 12 4 3 How Provisioning Affects Jobs A job that has requested an AOE will not preempt another job Therefore no job will be terminated in order to run a job with a requested AOE A job that has requested an AOE will not be backfilled around PBS Professional 10 4 User s Guide 295 Chapter 12 Using Provisioning 12 5 Caveats and Errors 12 5 1 Requested Job AOE and Reservation AOE Should Match Do not submit jobs that request an AOE to a reservation that does not request the same AOE Reserved vnodes may not supply that AOE your job will not run 12 5 2 Allow Enough Time in Reservations If a job is submitted to a reservation with a duration close to the walltime of the job provisioning could cause the job to be terminated before it finishes running or to be prevented from starting If a reservation is designed to take jobs requesting an AOE leave enough extra time in the reservation for provisioning 12 5 3 Requesting Multiple AOEs For a Job or Reservation Do not request more than one AOE per job or reservation The job will not run or the r
222. or a forwardable renewable Ker beros V5 TGT Specify the W cred dce option to qsub if a forwardable renewable Kerberos V5 TGT ticket granting ticket with the user as the listed princi pal is what is to be sent with the job If the user has an established creden tials cache and a non expired forwardable renewable TGT is in the cache that information is used 6c The other choice W cred dce pass causes the qsub command to interact with the user to generate a DES encryption of the user s password This encrypted password is sent to the PBS Server and MOM processes where it is placed in a job specific file for later use by pbs_mom in acquir ing a DCE login context for the job The information is destroyed if the job terminates is deleted or aborts IMPORTANT The W pwd option to qsub has been superseded by the above two options and therefore should no longer be used Any acquired login contexts and accompanying DCE credential caches established for the job get removed on job termination or deletion qsub Wcred dce lt other qsub options gt job script IMPORTANT The W cred option to qsub is not available under Win dows 198 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 8 12 Running PBS in a UNIX Kerberos Environment PBS Professional includes optional support for Kerberos only i e no DCE environment By optional we mean that the custo
223. or any reason Altair s Suppliers and ISVs as third party bene ficiaries shall be entitled to enforce the terms of this Agreement directly against Licensee as necessary to protect Supplier s intellectual property or other rights Altair and its resellers providing support and training to licensed end users of the Products shall keep confidential all Licensee information provided to Altair in order that Altair may provide Support and training to Licensee Licensee information shall be used only for the purpose of assisting Lic ensee in its use of the licensed Products Altair agrees to take the same pre cautions necessary to protect and maintain the confidentiality of the Licensee information as it does to protect its own information of a confi dential nature but in any event no less than a reasonable degree of care and shall not disclose or make them available to any person or entity except as expressly provided in this Agreement 5 MAINTENANCE AND SUPPORT Maintenance Altair will provide Licensee at no additional charge for annual licenses and for a maintenance fee for paid up licenses with Maintenance Releases and Updates of the Products that are generally released by Altair during the term ofthe licenses granted under this Agreement except that this shall not apply to any Term or Renewal Term for which full payment has not been received Altair does not promise that there will be a certain number of Updates or any Updates during a pa
224. or b as follows a PBS l select 1 ncpus 1 host hostA 2 ncpus 1 host hostB b PBS l select 1 ncpus 1 host hostA 1 ncpus 2 host hostB mpiprocs 2 256 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 which will cause PBS_NODEFILE to list hostA hostB hostB and an mpirun np 3 will then be consistent 10 6 10 1 Options If executed inside a PBS job script all of the options to the PBS interface are the same as for Intel MPI s mpirun except for the following host ghost For specifying the execution host to run on Ignored machinefile lt file gt The file argument contents are ignored and replaced by the contents of the PBS_ NODEFILE mpdboot option totalnum Ignored and replaced by the number of unique entries in SPBS_ NODEFILE mpdboot option file np Ignored and replaced by the name of PBS_NODEFILE The argument to this option is replaced by SPBS_ NODEFILE Argument to mpdboot option f lt mpd_hosts_ file gt replaced by PBS_ NODEFILE If the PBS interface to Intel MPI s mpirun is called inside a PBS job Intel MPI smpirun s argument to mpdboot is not supported as this closely matches the mpirun option s lt spec gt The user can simply run a separate mpd boot s before calling mpirun A warning message is issued by the PBS interface upon encountering a s option telling users of the supported form If the user does not specify a np option then n
225. or deny access to PBS systems on the basis of username group host and or net work domain Job Accounting offers detailed logs of system activities for charge back or usage analysis per user per group per project and per compute host Automatic File Staging provides users with the ability to specify any files that need to be copied onto the execution host before the job runs and any that need to be copied off after the job completes The job will be sched uled to run only after the required files have been successfully transferred Parallel Job Support works with parallel programming libraries such as MPI PVM and HPF Applications can be scheduled to run within a single multi processor computer or across multiple systems System Monitoring includes a graphical user interface for system monitor ing Displays vnode status job placement and resource utilization infor mation for both stand alone systems and clusters Job Interdependency enables the user to define a wide range of inter dependencies between jobs Such dependencies include execution order and execution conditioned on the success or failure of another specific job or set of jobs Computational Grid Support provides an enabling technology for meta computing and computational grids Comprehensive API includes a complete Application Programming Inter face API for sites who desire to integrate PBS with other applications or who wish to support unique job scheduli
226. ostC hostC Job script mpirun runs MPD daemons on hosts listed in PBS NODEFILE mpirun runs 2 instances of mpitestl on hostA 2 instances of mpitest2 on hostB 2 instances of mpitest3 on hostc mpirun takes care of shutting down the MPDs at the end of MPI job run mpirun np 2 tmp mpitestl np 2 tmp mpitest2 np 2 tmp mpitest3 Run job script qsub l select 3 ncpus 2 job script lt job id gt PBS Professional 10 4 User s Guide 259 Chapter 10 Multiprocessor Jobs Example 3 Run an Intel MPI job with multiple executables on multiple hosts via the configfile option and PBS_NODEFILE PBS_NODEFILE hostA hostA hostB hostB hostc hostc Job script echo np 2 tmp mpitest1 gt gt my_config file echo np 2 tmp mpitest2 gt gt my_config file echo np 2 tmp mpitest3 gt gt my_config file mpirun takes care of starting the MPD daemons config file says run 2 instances of mpitestl on hostA 2 instances of mpitest2 on hostB 2 instances of mpitest3 on hostc mpirun takes care of shutting down the MPD S SH HH daemons mpirun configfile my config file cleanup rm f my_ config file Run job script qsub l select 3 ncpus 2 job script lt job id gt 260 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 11 PBS Jobs with MVAPICH1 s mpirun PBS provides an interface to MVAPICH1 s mpirun MVAPICH1 allows use of Infin
227. ot be put in the queue or accepted by the server If while running a job exceeds its limit for a con sumable or time based resource it will be terminated A consumable resource is one that is reduced by being used for exam ple nepus licenses or mem A non consumable resource is not reduced through use for example walltime or a boolean resource Resources are tracked in server queue vnode and job attributes Servers queues and vnodes have two attributes resources_available RESOURCE and resources_assigned RESOURCE The resources available RESOURCE attribute tracks the total amount of the resource available at that server queue or vnode without regard to how much is in use The resources_assigned RESOURCE attribute tracks how much of that resource has been assigned to jobs at that server queue or vnode Jobs have an attribute called resources_used RESOURCE which tracks the amount of that resource used by that job The administrator can set server and queue defaults for resources used in chunks See the PBS Professional Administrator s Guide and the pbs_server_attributes 7B and pbs_queue_attributes 7B manual pages 4 2 0 1 Unset Resources When job resource requests are being matched with available resources a numerical resource that is unset on a host is treated as if it were zero and an unset string cannot be matched An unset Boolean resource is treated as if it is set to False An unset resource at th
228. ould have been Incpus 10 mem 20gb 2 Request four chunks each with 1 CPU and 4GB of memory taken from anywhere 1l select 4 ncpus 1 mem 4GB l place free 3 Allocate 4 chunks each with 1 CPU and 2GB of memory from between PBS Professional 10 4 User s Guide 59 Chapter 4 Submitting a PBS Job one and four vnodes which have an arch of linux 1 select 4 ncpus 1 mem 2GB arch linux 1l place free 4 Allocate four chunks on 1 to 4 vnodes where each vnode must have 1 CPU 3GB of memory and 1 node locked dyna license available for each chunk l select 4 dyna 1 ncpus 1 mem 3GB l place free 5 Allocate four chunks on 1 to 4 vnodes and 4 floating dyna licenses This assumes dyna is specified as a server dynamic resource 1l dyna 4 1 select 4 ncpus 1 mem 3GB l place free 6 This selects exactly 4 vnodes where the arch is linux and each vnode will be on a separate host Each vnode will have 1 CPU and 2GB of memory allocated to the job lselect 4 mem 2GB ncpus 1 arch linux lplace scatter 7 This will allocate 3 chunks each with 1 CPU and 10GB of memory This will also reserve 100mb of scratch space if scratch is to be accounted Scratch is assumed to be on a file system common to all 60 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 hosts The value of place depends on the default which is place free 1l scratch 100mb 1 select 3 ncpus 1 mem 10GB 8 This will a
229. ources which were created to be invisible or unre questable cannot be requested See section 4 5 14 Resource Permissions on page 54 The 1 nodes form is deprecated and if it is used it will be converted into a request for chunks and job wide resources Most jobs submitted with Inodes will continue to work as expected These jobs will be auto matically converted to the new syntax However job tasks may execute in an unexpected order because vnodes may be assigned in a different order Jobs submitted with old syntax that ran successfully on versions of PBS Professional prior to 8 0 can fail because a limit that was per chunk is now job wide This is an example of a job submitted using nodes X Imem M that would fail because the mem limit is now job wide If the following conditions are true a PBS Professional 9 0 or later using standard MPICH b The job is submitted with qsub Inodes 5 lmem 10GB c The master process of this job tries to use more than 2GB The job will be killed where in lt 7 0 the master process could use 10GB before being killed 10GB is now a job wide limit divided up into a 2GB limit per chunk For more information see the qsub 1B qalter 1B pbs_rsub 1B and pbs _resources 7B manual pages Do not use an old style resource or node specification Inodes with Iselect or lplace This will produce an error 44 PBS Professional 10 4 User s Guide Submitting a PBS Job Ch
230. page 22 discusses PBS environment variables A queue complex in NQS was a grouping of queues within a batch Server The purpose of a complex was to provide additional control over resource usage The advanced scheduling features of PBS eliminate the requirement for queue complexes PBS Professional 10 4 User s Guide 299 13 1 Converting Date Specifications Converting NQS date specifications to the PBS form may result in a warn ing message and an incomplete converted date PBS does not support date specifications of today tomorrow or the name of the days of the week such as Monday If any of these are encountered in a script the PBS specification will contain only the time portion of the NQS specification i e PBS a hhmm ss It is suggested that you specify the execu tion time on the gsub command line rather than in the script All times are taken as local time If any unrecognizable NQS directives are encountered an error message is displayed The new PBS script will be deleted if any errors occur 300 PBS Professional 10 4 User s Guide Appendix B License Agreement CAUTION PRIOR TO INSTALLATION OR USE OF THE SOFTWARE YOU MUST CONSENT TO THE FOLLOWING SOFTWARE LICENSE TERMS AND CONDITIONS BY CLICKING THE I ACCEPT BUT TON BELOW YOUR ACCEPTANCE CREATES A BINDING LEGAL AGREEMENT BETWEEN YOU AND ALTAIR IF YOU DO NOT HAVE THE AUTHORITY TO BIND YOUR ORGANIZATION TO THESE TERMS AND CONDITIONS YOU
231. pe poe fileset there are three directories containing sample code that may be of interest from How installing the POE fileset alters your system e usr lpp ppe poe samples swtbl Directory containing sample code for running User Space POE jobs without LoadLeveler e usr lpp ppe poe samples ntbl Directory containing sample code for running User Space jobs without LoadLeveler using the network table API e usr lpp ppe poe samples nrt Directory that contains the sample code for running User Space jobs on InfiniBand interconnects without LoadLeveler using the network resource table API 10 6 3 7 Notes Since PBS is tracking tasks started by poe these tasks are counted towards a user s run limits Running multiple poe jobs in the background will not work Instead run poe jobs one after the other or submit separate jobs Otherwise HPS windows will be used by more than one task The tracejob command will show any of various error messages For more information on using IBM s Parallel Operating Environment see IBM Parallel Environment for AIX 5L Hitchhiker s Guide 238 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 6 3 8 Examples Using poe Example 1 Using IP mode run a single executable poe job with 4 ranks on hosts spread across the PBS allocated nodes listed in PBS_NODEFILE cat PBS_NODEFILE hostl host2 host3 host4 cat job script poe path mpiprog euilib ip
232. pecific Staging and Execution Directories In this example you want the file jay fem to be delivered to the job spe cific staging and execution directory given in PBS_JOBDIR by being cop ied from the host submithost The job script is executed in PBS_JOBDIR and jay out is staged out from PBS_JOBDIR to your home directory on the submittal host i e hostname qsub Wsandbox PRIVATE Wstagein jay fem submit host jay fem Wstageout jay out submithost jay out 8 6 6 Summary of the Job s Lifecycle This is a summary of the steps performed by PBS The steps are not neces sarily performed in this order e On each execution host if specified PBS creates a job specific staging and execution directory e PBS sets PBS JOBDIR and the job s jobdir attribute to the path of the job s staging and execution directory e On each execution host allocated to the job PBS creates a job specific temporary directory e PBS sets the TMPDIR environment variable to the pathname of the temporary directory e If any errors occur during directory creation or the setting of variables the job is requeued e PBS stages in any files or directories e The prologue is run on the primary execution host with its current working directory set to PBS_ HOME mom_ priv and with PBS _JOBDIR and TMPDIR set in its environment e The job is run as the user on the primary execution host e The job s associated tasks are run as
233. pus 2 job script lt job id gt Example 2 Launch an MVAPICH2 MPI job with multiple executables on multiple hosts listed in the default file mpd hosts Here run executa bles prog and prog2 with 2 ranks of prog on host1 2 ranks of prog2 on host2 and 2 ranks of prog2 on host3 all specified on the command line PBS_NODEFILE pbs hostl pbs host2 pbs host3 Job script mpiexec n 2 progl n 2 prog2 n 2 prog2 Run job script qsub l select 3 ncpus 2 job script lt job id gt Example 3 Launch an MVAPICH2 MPI job with multiple executables on multiple hosts listed in the default file mpd hosts Run executables progl and prog2 with 2 ranks of prog on host1 2 ranks of prog2 on host2 and 2 ranks of prog2 on host3 all specified using the configfile option PBS_NODEFILE pbs hostl pbs host2 pbs host3 264 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 Job script echo n 2 host hostl progl gt tmp jobconf echo n 2 host host2 prog2 gt gt tmp jobconf echo n 2 host host3 prog2 gt gt tmp jobconf mpiexec configfile tmp jobconf rm tmp jobconf Run job script qsub l select 3 ncpus 2 job script lt job id gt 10 6 13 PBS Jobs with HP MPI In order to override the default rsh set PBS RSHCOMMAND in your job script export PBS _RSHCOMMAND lt rsh_cmd gt 10 7 MPI Jobs on the Altix 10 7 1 Jobs on an Altix Running ProPack 4 5 PBS has its own mpiexec for the Altix runnin
234. put file of the specified job s If neither option is specified the message will be written to the error file of the job The first operand message_st ring is the message to be written If the string contains blanks the string must be quoted If the final character of the string is not a newline a newline character will be added when written to the job s file All remaining operands are job_identifiers which specify the jobs to receive the message string For example qmsg E hello to my error e file 55 qmsg O hello to my output o file 55 qmsg this too will go to my error e file 55 PBS Professional 10 4 User s Guide 125 Chapter 6 Working With PBS Jobs To send a message to a job using xpbs first select the job s of interest then click the msg button Doing so will launch the Send Message to Job dialog box From this window you may enter the message you wish to send and indicate whether it should be written to the standard output or the standard error file of the job Click the Send Message button to complete the process 6 5 Sending Signals to Jobs The qsig command requests that a signal be sent to executing PBS jobs The signal is sent to the session leader of the job Usage syntax of the qsig command is qsig s signal job_identifier Job array job identifiers must be enclosed in double quotes If the s option is not specified SIGTERM is sent If the s option is spec ified it
235. qstat will output information about each destination The various options to qstat take as an operand either a job identifier or a destination If the operand is a job identifier it must be in the following form sequence_number server_name server where sequence_number server_name is the job identifier assigned at submittal time see qsub If the server_name is omitted the name of the default Server will be used If server is supplied the request will be for the job identifier currently at that Server If the operand is a destination identifier it takes one of the following three forms queue server queue server If queue is specified the request is for status of all jobs in that queue at the default Server If the server form is given the request is for status of all jobs at that Server If a full destination identifier queue server is given the request is for status of all jobs in the named queue at the named server IMPORTANT If a PBS Server is not specified on the qst at command line the default Server will be used See discussion of PBS_DEFAULT in section 3 13 Environment Variables on page 28 138 PBS Professional 10 4 User s Guide Checking Job System Status Chapter 7 7 1 3 Checking Server Status The B option to gst at displays the status of the specified PBS Batch Server One line of output is generated for each Server queried The three letter abbreviations correspond to various
236. queue states and job counts as follows Maximum Total Enabled Status Started Status Queued Running Held Waiting Transiting and Exiting The last column gives the type of the queue routing or execution qstat Q Queue Max Tot Ena Str Que Run Hld Wat Trn Ext Type workq 0 10 yes yes 7 1 1 1 0 0 Execution The full display for a queue provides additional information qstat Qf Queue workq queue_type Execution total_jobs 10 state_count Transit 0 Queued 7 Held 1 Waiting 1 Running 1 Exiting 0 resources_assigned ncpus 1 hasnodes False enabled True started True 7 1 5 Viewing Job Information We saw above that the f option could be used to display full or long information for queues and Servers The same applies to jobs By specify ing the option and a job identifier PBS will print all information known about the job e g resources requested resource limits owner PBS Professional 10 4 User s Guide 141 Chapter 7 Checking Job System Status source destination queue etc as shown in the following example See Job Attributes on page 404 of the PBS Professional Reference Guide for a description of attributes qstat f 89 Job Id 89 south 142 Job Name tns3d Job Owner userl south example com resources used cput 00 00 00 resources _used mem 2700kb resources _used ncpus 1 resources _used vmem 5500kb resources used walltime 00 00 00 job_state R queu
237. r Form b execute the Prod ucts on Licensed Workstations in accordance with the LMS for use solely by Licensee s employees or its onsite Contractors who have agreed to be bound by the terms of this Agreement for Licensee s internal business use on Licensed Workstations within the Global Zone s as identified on the Order Form and for the term identified on the Order Form c make backup copies of the Products provided that Altair s and its Suppliers and ISV s Proprietary Rights Notices are reproduced on each such backup copy d freely modify and use Templates and create interfaces to Licensee s propri etary software for internal use only using APIs provided that such modifi cations shall not be subject to Altair s warranties indemnities support or other Altair obligations under this Agreement and e copy and distribute Documentation inside Licensee s organization exclusively for use by Lic ensee s employees and its onsite Contractors who have agreed to be bound by the terms of this Agreement A copy of the License Log File shall be made available to Altair automatically on no less than a monthly basis In 304 PBS Professional 10 4 User s Guide the event that Licensee uses a third party vendor for information technol ogy IT support the IT company shall be permitted to access the Software only upon its agreement to abide by the terms of this Agreement Licensee shall indemnify defend and hold harmless Altair for the actions of its IT
238. r or by a PBS Operator or Manager For example to delete S304 south pbs_rdel S304 south or pbs_rdel S304 190 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 8 8 8 Submitting a Job to a Reservation Jobs can be submitted to the queue associated with a reservation or they can be moved from another queue into the reservation queue You submit a job to a reservation by using the q lt queue gt option to the qsub command to specify the reservation queue For example to submit a job to the soon est occurrence of a standing reservation named 123 south submit to its queue 123 qsub q S123 lt script gt You move a job into a reservation queue by using the qmove command For more information see the qsub 1B and qmove 1B man pages For example to qmove job 22 myhost from workq to S123 the queue for the reservation named 123 south qmove S123 22 myhost or qmove S123 22 A job submitted to a standing reservation without a restriction on when it can run will be run if possible during the soonest occurrence In order to submit a job to a specific occurrence use the a lt start time gt option to the qsub command setting the start time to the time of the occurrence that you want You can also use a cron job to submit a job at a specific time See the qsub 1B and cron 8 man pages 8 8 8 1 Running Jobs in a Reservation A confirmed reservation will accept jobs into its queue at any time Jobs
239. rces can be created so that they are invisible or cannot be requested or altered Ifa resource is invisible it also cannot be requested or altered The function of some PBS commands depends upon whether a resource can be viewed requested or altered These commands are those which view or request resources or modify resource requests pbsnodes Users cannot view restricted host level custom resources pbs_rstat Users cannot view restricted reservation resources pbs_rsub Users cannot request restricted custom resources for reser vations qalter Users cannot alter a restricted resource 54 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 qmer Users cannot print or list a restricted resource qselect Users cannot specify restricted resources via l resource list qsub Users cannot request a restricted resource qstat Users cannot view a restricted resource 4 6 Placing Jobs on Vnodes The place statement controls how the job is placed on the vnodes from which resources may be allocated for the job The place statement can be specified in order of precedence via 1 Explicit placement request in qalter 2 Explicit placement request in qsub 3 Explicit placement request in PBS job script directives 4 Default qsub place statement 5 Queue default placement rules 6 Server default placement rules 7 Built in default conversion and placement rules The place statement may be not be used without the s
240. reated to be invisible or unrequestable whether this resource is on a queue the server or is a job attribute See section 4 5 14 Resource Permissions on page 54 148 PBS Professional 10 4 User s Guide Checking Job System Status Chapter 7 7 1 21 Viewing Job History You can view information for jobs that have finished or were moved as long as that information is still being stored by PBS See section 6 9 Using Job History Information on page 131 You can view the same attribute information regardless of whether the job is queued running finished or moved as long as job history information is being preserved 7 1 21 1 Job History In Standard Format You can use the xX option to the qstat command to see information for finished moved queued and running jobs in standard format Usage qstat x Displays information for queued running finished and moved jobs in standard format qstat x lt job ID gt Displays information for a job regardless of its state in standard for mat Example 1 Showing finished and moved jobs with queued and running jobs qstat x Job id Name User Time Use S Queue 101 serverl STDIN userl 00 00 00 F workq 102 server1 STDIN userl 00 00 00 M destq server2 103 serverl STDIN userl 00 00 00 R workq 104 serverl STDIN userl 00 00 00 Q workq To see status for jobs job arrays and subjobs that are queued running fin ished and moved useqstat xt To see status for j
241. rence Guide and the qstat 1B man page For example if a job named MyJob has been submitted to the soonest occurrence of the standing reservation named 304 south it is listed under 304 the name of the queue qstat Job id Name User Time Use S Queue 139 south MyJob userl 0 Q 5304 8 8 9 Reservation Caveats and Errors 8 8 9 1 Setting the Submission Host s Time Zone The environment variable PBS_TZID must be set at the submission host The time for which a reservation is requested is the time defined at the sub mission host The format for PBS_TZID is a timezone location rather than a timezone POSIX abbreviation Examples of values for PBS_TZID are America Los_Angeles America Detroit Europe Berlin Asia Calcutta PBS Professional 10 4 User s Guide 193 Chapter 8 8 8 9 2 Reservation Errors Advanced PBS Features The following table describes the error messages that apply to reservations Table 8 5 Reservation Errors Server ae Log Description of Error Error Message Error Code Invalid syntax when specifying a 15133 pbs_rsub error Unde standing reservation fined iCalendar syntax Recurrence rule has both a COUNT 15134 pbs_rsub error Unde and an UNTIL parameter fined iCalendar syntax COUNT or UNTIL is required Recurrence rule missing valid 15134 pbs _rsub error Unde COUNT or UNTIL parameter fined iCalendar syntax A valid COUNT or UNTIL is required
242. reserve the job s exit status in your logout file by saving it at the top then doing an explicit exit at the end as shown below set EXITVAL Sstatus previous contents of logout here exit SEXITVAL Likewise if the user s login shell is csh the following message may appear in the standard output of a job PBS Professional 10 4 User s Guide 23 Chapter 3 Getting Started With PBS Warning no access to tty thus no job control in this shell This message is produced by many csh versions when the shell deter mines that its input is not a terminal Short of modifying csh there is no way to eliminate the message Fortunately it is just an informative mes sage and has no effect on the job An interactive job comes complete with a pseudotty suitable for running those commands that set terminal characteristics But more importantly it does not caution the user that starting something in the background that would persist after the user has exited from the interactive environment might cause trouble for some moms They could believe that once the inter active session terminates all the user s processes are gone with it For example applications like ssh agent background themselves into a new session and would prevent a CPU set enabled mom from deleting the CPU set for the job This in turn might cause subsequent failed attempts to run new jobs resulting in them being placed in a held state 3 11 3 Setting MANPATH on SGI Systems
243. rofessional 10 4 User s Guide 235 Chapter 10 Multiprocessor Jobs If N lt Mand US mode is being used US mode is turned off and a warning is printed pbsrun poe Warning user mode disabled due to MP PROCS setting 10 6 3 5 Caveats 10 6 3 5 1 Multi host Jobs If you wish to run a multi host job it must not run on a mix of InfiniBand and non InfiniBand hosts It can run entirely on hosts that are non Infini Band or on hosts that are all using InfiniBand but not both 10 6 3 5 2 Job Submission Format Do not submit InfiniBand jobs in which the select statement specifies only a number for example export PBS_GET_IBWINS 1 qsub koe mn l1 select 1 V jobname Instead use the equivalent request which specifies a resource export PBS_GET_IBWINS 1 qsub koe mn l select 1 ncpus 1 V jobname 10 6 3 5 3 Environment Variable Do not set the PBS_O_HOST environment variable Using the qsub command with the V option will fail 10 6 3 5 4 If Your Complex Contains Machines Not on the HPS If your complex contains machines that are not on the HPS and you wish to run on the HPS you must specify machines on the HPS Your adminis trator will define a resource on each host on the HPS To specify machines on the HPS you must request the hps resource in your select statement For this example the resource is hps 236 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 Using place scatter Wh
244. rofile Jobs Chapter 11 Table 11 4 Warning and Error Messages Logged by HPCBP MOM Error Condition Message Job request contains a HPCBP can t submit job to HPC Basic script MOM Profile Server HPCBP MOM doesn t accept job script JSDL script file problem HPCBP unable to create JSDL document MOM gSOAP related problems cannot create SSL based HPCBP unable to create ssl based chan channel MOM nel to connect to the Web Service endpoint Username token problem HPCBP unable to add username pass MOM word to soap message Cannot initialize g OAP HPCBP unable to initialize gsoap runtime runtime environment MOM environment Problems encountered during job submission Cannot add SOAP HPCBP unable to add soap header to the Header MOM create activity request message Bad JSDL script file HPCBP unable to open JSDL document MOM Problem with JSDL HPCBP error in reading contents of the attribute MOM JSDL document Problem with HPCBP HPCBP unable to submit job to the hpcbp Server connection MOM web service endpoint Problem with user s pass HPCBP unable to submit job with user s word MOM credential amp PBS Server Problem reading SOAP HPCBP unable to read HPCBP job iden response MOM tifier from create activity response PBS Professional 10 4 User s Guide 281 Chapter 11 HPC Basic Profile Jobs Table 11 4 Warning and Error Messages Logged by HPCBP
245. rror STDERR of the job will be retained in the job s staging and execution directory on the primary execution host If set this option overrides the path name for the corresponding file If not set neither file is retained on the execution host The argument is either the single let ter e or o or the letters e and o combined in either order Or the argument is the letter n If k is not specified neither file is retained e The standard error file is to be retained in the job s staging and execution directory on the primary execution host The job s name will be the default file name given by job_name esequence where job_name is the name specified for the job and sequence is the sequence num ber component of the job identifier The standard output file is to be retained in the job s staging and execution directory on the primary execution host The file name will be the default file name given by job_name osequence where job_name is the name specified for the job and sequence is the sequence num ber component of the job identifier eo oe Both standard output and standard streams are retained on the primary execution host in the job s staging and execution directory Neither file is retained qsub k oe my_job PBS k eo 90 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 4 13 19 Suppressing Job Identifier The z option directs the gsub command t
246. rticular year If there is any question or dispute as to whether a particular release is a Maintenance Release an Update or a new product the categorization of the release as determined by Altair shall be final Licensee agrees to install Maintenance Releases and Updates promptly after receipt from Altair Maintenance Releases and Updates are subject to this Agreement Altair shall only be obligated to pro vide support and maintenance for the most current release of the Software and the most recent prior release Support Altair will provide support via telephone and email to Licensee at the fees if any as listed on the Order Form If Support has not been procured for any period of time for paid up licenses a reinstatement fee shall apply Support consists of responses to questions from Licensee s personnel related to the use of the then current and most recent prior release version of the Software Licensee agrees to provide Altair with sufficient information to resolve technical issues as may be reasonably requested by Altair Licensee agrees to the best of its 306 PBS Professional 10 4 User s Guide abilities to read comprehend and follow operating instructions and proce dures as specified in but not limited to Altair s Documentation and other correspondence related to the Software and to follow procedures and rec ommendations provided by Altair in an effort to correct problems Lic ensee also agrees to notify Altair of a programmin
247. s 1 mem 1gb In prior versions nodes x implied 1 CPU per node unless otherwise explicitly stated so the ncpus 1 is not inherited from the server default qsub 1l mem 1l6gb l nodes 4 In QA select 4 ncpus 1 mem 4gb This is not 4 ncpus 2 because in prior versions nodes x implied 1 CPU per node unless 52 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 otherwise explicitly stated In QB select 4 ncpus 1 mem 4gb In prior versions nodes x implied 1 CPU per node unless otherwise explicitly stated so the ncpus 1 is not inherited from the server default 4 5 11 Resource Request Conversion Dependent on Where Resources are Defined A job s resource request is converted from old style to new according to various rules one of which is that the conversion is dependent upon where resources are defined For example The boolean resource Red is defined on the server and the boolean resource Blue is defined at the host level A job requests qsub 1 Blue True This looks like an old style resource request and PBS checks to see where Blue is defined Since Blue is defined at the host level the request is converted into 1 select 1 Blue True However if a job requests qsub 1 Red True while this looks like an old style resource request PBS does not convert it to a chunk request because Red is defined at the server 4 5 12 Jobs Submitted with Undefined Resources Any job sub
248. s Agreement and in the event of such breach 310 PBS Professional 10 4 User s Guide the aggrieved party shall be entitled to seek specific performance and or injunctive relief without posting a bond or other security in order to enforce or prevent any violation of this Agreement 11 RESERVED 12 Notice All notices given by one party to the other under the Agree ment or these Additional Terms shall be sent by certified mail return receipt requested or by overnight courier to the respective addresses set forth in this Agreement or to such other address either party has specified in writing to the other All notices shall be deemed given upon actual receipt Written notice shall be made to Altair Licensee Name amp Address Altair Engineering Inc 1820 E Big Beaver Rd Troy MI 48083 Attn Tom M PerringAttn 13 TERM For annual licenses or Support provided for perpetual licenses renewal shall be automatic for each successive year Renewal Term upon mutual written execution of a new Order Form All charges and fees for each Renewal Term shall be set forth in the Order Form exe cuted for each Renewal Term All Software licenses procured by Licensee may be made coterminous at the written request of Licensee and the con sent of Altair 14 TERMINATION Either party may terminate this Agreement upon thirty 30 days prior written notice upon the occurrence of a default or material breach by the oth
249. s an old style specification in the job script that will result in an error PBS Professional 10 4 User s Guide 65 Chapter 4 Submitting a PBS Job When a nodespec is converted into a select statement the job will have the environment variables NCPUS and OMP_ NUM_THREADS set to the value of ncpus in the first piece of the nodespec This may produce incom patibilities with prior versions when a complex node specification using different values of ncpus and ppn in different pieces is converted 4 8 1 Node Specification Conversion Node specification format Inodes N spec_list spec_list N spec_list spec_list suffix Incpus Z where spec_list has syntax spec spec spec is any of hostname property ncpus X cpp X ppn P suffix is any of property excl shared N and P are positive integers X and Z are non negative integers The node specification is converted into selection and placement directives as follows Each spec_list is converted into one chunk so that N spec_list is converted into N chunks If spec is hostname The chunk will include host hostname If spec matches any vnode s resources_available host value The chunk will include host hostname If spec is property The chunk will include property true Property must be a site defined vnode level boolean resource 66 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 If spec is ncpus X or cpp X The chunk w
250. s inside of quote marks are expanded e Wildcards cannot be used to match UNIX files that begin with period or Windows files that have the SYSTEM or HIDDEN attributes e When using the qsub command line on UNIX you must prevent the shell from expanding wildcards For some shells you can enclose the pathnames in double quotes For some shells you can use a back space before the wildcard e Wildcards can only be used in the source side of a staging specifica tion This means they can be used in the remote_path specification for stagein and in the local_path specification for stageout e When staging using wildcards the destination must be a directory If the destination is not a directory the result is undefined So for PBS Professional 10 4 User s Guide 169 Chapter 8 Advanced PBS Features example when staging out all out files you must specify a directory for remote_path e Wildcards can only be used in the final path component i e the base name e When wildcards are used during stagein PBS will not automatically delete staged files at job end Note that if PBS created the staging and execution directory that directory and all its contents are deleted at job end Examples 1 Stage out all files from the execution directory to a specific directory e UNIX W stageout myworkstation user projectl1 casel e Windows W stageout mypc E projectl1l casel 2 Stage out specific types of
251. s is sent to the job owner 162 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 For delivery of output files on the local host PBS uses the cp command UNIX or the xcopy command Windows XP or the robocopy com mand Windows Vista Local and remote delivery of output may fail for the following additional reasons e A directory in the specified destination path does not exist e A directory in the specified destination path is not searchable by the user e The target directory is not writable by the user 8 6 Input Output File Staging File staging is a way to specify which files should be copied onto the exe cution host before the job starts and which should be copied off the execu tion host when it finishes 8 6 1 Staging and Execution Directory User s Home vs Job specific The job s staging and execution directory is the directory to which files are copied before the job runs and from which output files are copied after the job has finished This directory is either your home directory or a job spe cific directory created by PBS just for this job If you use job specific stag ing and execution directories you don t need to have a home directory on each execution host as long as those hosts are configured properly In addition each job gets its own staging and execution directory so you can more easily avoid filename collisions PBS Professional 10 4 User s Guide 163 Chapter 8 Thi
252. s table lists the differences between using your home directory for stag Advanced PBS Features ing and execution and using a job specific staging and execution directory created by PBS Table 8 1 Differences Between User s Home and Job specific Directory for Staging and Execution Question Regarding Action User s Home Job specific Requirement or Setting Directory Directory Does PBS create a job specific No Yes staging and execution directory User s home directory must exist Yes No on execution host s Standard out and standard error No Yes automatically deleted when qsub k option is used When are staged out files are Successfully Only after all deleted staged out files are success are deleted oth fully staged ers go to unde out livered Staging and execution directory No Yes deleted after job finishes How is job s sandbox attribute set HOME or not set PRIVATE 8 6 2 Using Job specific Staging and Execution Directories 8 6 2 1 Setting the Job s Staging and Execution Directory The job s sandbox attribute controls whether PBS creates a unique job specific staging and execution directory for this job Ifthe job s sandbox attribute is set to PRIVATE PBS creates a unique staging and execution 164 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 directory for the job If sandbox is unset or is set to HOME PBS uses the user s home
253. s that can be modified are cputime and wall time These can only be reduced PBS Professional 10 4 User s Guide 117 Chapter 6 Working With PBS Jobs When the qalter 1 option is used to alter the resource list of a queued job it is important to understand the interactions between altering the select directive and job limits If the job was submitted with an explicit 1 select then vnode level resources must be qaltered using the 1 select form In this case a vnode level resource RES cannot be qaltered with the RES form For example Submit the job qsub l select 1 ncpus 2 mem 512mb jobscript Job s ID is 230 qalter the job using 1 RES form qalter 1 ncpus 4 230 Error reported by qalter galter Resource must only appear in select specification when select is used ncpus 230 qalter the job using the 1 select form qalter 1 select 1 ncpus 4 mem 512mb 230 No error reported by qalter 6 1 1 Changing the Selection Directive If the selection directive is altered the job limits for any consumable resource in the directive are also modified 118 PBS Professional 10 4 User s Guide Working With PBS Jobs Chapter 6 For example if a job is queued with the following resource list select 2 ncpus 1 mem 5gb ncpus 2 mem 10gb and the selection directive is altered to request select 3 ncpus 2 mem 6gb then the job limits are reset to ncpus 6 and mem 18gb 6 1 2 Changing the Job wide
254. s the following pbsrun poe Warning hostfile value replaced by PBS or pbsrun poe Warning hfile value replaced by PBS 234 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 If this environment variable is set when a poe job is sub mitted PBS prints the following error message pbsrun poe Warning MP _HOSTFILE value replaced by PBS instances MP_INSTANCES The option and the environment variable are treated differ ently instances If the option is set PBS prints a warning pbsrun poe Warning instances cmd line option removed by PBS MP_INSTANCES If the environment variable is set PBS uses it to calcu late the number of network windows for the job The maximum value allowed can be requested by using the string max for the environment variable If the environment variable is set to a value greater than the maximum allowed value it is replaced with the max imum allowed value The default maximum value is 4 procs MP_PROCS This option or environment variable should be set to the total number of mpiprocs requested by the job when using US mode If neither this option nor the MP_PROCS environment variable is set PBS uses the number of entries in SPBS_ NODEFILE If this option is set to N and the job is submitted with a total of M mpiprocs If N gt M The value N is passed to IBM poe If N lt Mand US mode is not being used The value N is passed to poe PBS P
255. s up the Submit Job Dialog box see below which is composed of four distinct regions The Job Script File region is at the upper left The OPTIONS region containing various widgets for setting job PBS Professional 10 4 User s Guide 107 Chapter 5 Using the xpbs GUI attributes is scattered all over the dialog box The OTHER OPTIONS is located just below the Job Script file region and COMMAND BUTTONS region is at the bottom The job script region is composed of a header box the text box FILE entry box and two buttons labeled load and save If you have a script file con taining PBS options and executable lines then type the name of the file on the FILE entry box and then click on the oad button Alternatively you may click on the FILE button which will display a File Selection browse window from which you may point and click to select the file you wish to open The File Selection Dialog window is shown below Clicking on the Select File button will load the file into xpbs just as does the oad button described above 108 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 The various fields in the Submit window will get loaded with values found in the script file The script file text box will only be loaded with execut able lines non PBS found in the script The job script header box has a Prefix entry box that can be modified to specify the PBS directive to look for when parsing a script file for PBS
256. s whose execution failed because of system or net work failure e Jobs which were deleted before they could start execution 6 9 3 Job History Information PBS can keep all job attribute information including the following e Submission parameters e Whether the job started execution e Whether execution succeeded e Whether staging out of results succeeded e Which resources were used PBS Professional 10 4 User s Guide 13 Chapter 6 Working With PBS Jobs PBS keeps job history for the following jobs e Jobs that have finished execution e Jobs that were deleted e Jobs that were moved to another server The job history for finished and moved jobs is preserved and available for the specified duration After the duration has expired PBS deletes the job history information and it is no longer available The state of a finished job is F and the state of a moved job is M See Job States on page 435 of the PBS Professional Reference Guide Subjobs are not considered finished jobs until the parent array job is fin ished which happens when all of its subjobs have terminated execution 6 9 4 Working With Finished and Moved Jobs 6 9 4 1 Working With Moved Jobs You can use the following commands with moved jobs They will function as they do with normal jobs qdel gqalter ghold qmove qmsg qorder qrerun qrls qrun qsig 132 PBS Professional 10 4 User s Guide Working With PBS Jobs Chapter 6 6 9 4 2 PBS C
257. servation for vnodes that have a specific AOE avail able This way jobs needing that AOE can be submitted to that reserva tion This means that jobs needing that AOE are guaranteed to be running on vnodes that have that AOE available Each reservation can have at most one AOE specified for it Any jobs that run in that reservation must not request a different AOE from the one requested for the reservation 12 2 3 Job Substates and Provisioning When a job is in the process of provisioning its substate is provisioning This is the description of the substate provisioning The job is waiting for vnode s to be provisioned with its requested AOE Integer value is 77 See Job Substates on page 436 of the PBS Professional Reference Guide for a list of job substates PBS Professional 10 4 User s Guide 291 Chapter 12 Using Provisioning The following table shows how provisioning events affect job states and substates Table 12 1 Provisioning Events and Job States Substates Event Initial Job State Substate Resulting Job State Substate Job submitted Queued and ready for selection Provisioning starts Queued Queued Running Provision ing Provisioning fails to start Queued Queued Held Held Provisioning fails Running Provision ing Queued Queued Provisioning suc ceeds and job runs Running Provision ing Running Running Internal error occurs
258. servations for the reservations and jobs in them to clean up A job consumes resources even while it is in the E or exiting state This can take longer when large files are being staged If the job is still running when the reservation ends it may take up to two min utes to be cleaned up The reservation itself cannot finish cleaning up until its jobs are cleaned up This will delay the start time of jobs in the next res ervation unless there is enough time between the reservations for cleanup 8 8 10 Reservation Information in the Accounting Log The PBS Server writes an accounting record for each reservation in the job accounting file The accounting record for a reservation is similar to that for a job The accounting record for any job belonging to a reservation will include the reservation ID See Accounting Log on page 445 of the PBS Professional Reference Guide PBS Professional 10 4 User s Guide 195 Chapter 8 Advanced PBS Features 8 9 Dedicated Time Dedicated time is one or more specific time periods defined by the admin istrator These are not repeating time periods Each one is individually defined During dedicated time the only jobs PBS starts are those in special dedi cated time queues PBS schedules non dedicated jobs so that they will not run over into dedicated time Jobs in dedicated time queues are also sched uled so that they will not run over into non dedicated time PBS will attempt to backfill around th
259. sides controls the minimum inter val at which a job can be checkpointed The interval is specified in CPU minutes or walltime minutes The same value is used for both so for example if the minimum interval is specified as 12 then a job using the queue s interval for CPU time will be checkpointed every 12 minutes of CPU time and a job using the queue s interval for walltime will be check pointed every 12 minutes of walltime 84 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 4 13 13 3 Checkpoint Interval The c checkpoint spec option defines the interval in CPU min utes or in walltime minutes at which the job will be checkpointed The checkpoint spec argument is specified as c Checkpointing is to be performed according to the interval measured in CPU time set on the execution queue in which the job resides c lt minutes of CPU time gt Checkpointing is to be performed at intervals of the speci fied number of minutes of CPU time used by the job This value must be greater than zero Ifthe interval specified is less than that set on the execution queue in which the job resides the queue s interval is used Format Integer Checkpointing is to be performed according to the interval measured in walltime set on the execution queue in which the job resides w lt minutes of walltime gt Checkpointing is to be performed at intervals of the speci fied number of minutes of walltime used
260. south userl workq my_job 1 0 10 Q qstat u userl barry 51 south barry workq airfoil 930 1 0 13 R 0 01 52 south userl workq my_job 1 0 10 Q 54 south barry workq airfoil 1 0 13 Q PBS Professional 10 4 User s Guide 143 Chapter 7 Checking Job System Status 7 1 7 List Running Jobs The r option to qstat displays the status of all running jobs at the optionally specified PBS Server Running jobs include those that are run ning and suspended One line of output is generated for each job reported and the information is presented in the alternative display 7 1 8 List Non Running Jobs The i option to qst at displays the status of all non running jobs at the optionally specified PBS Server Non running jobs include those that are queued held and waiting One line of output is generated for each job reported and the information is presented in the alternative display see description above 7 1 9 Display Size in Gigabytes The option to qstat displays all jobs at the requested or default Server using the alternative display showing all size information in gigabytes GB rather than the default of smallest displayable units Note that if the size specified is less than 1 GB then the amount if rounded up to 1 GB 7 1 10 Display Size in Megawords The M option to qstat displays all jobs at the requested or default Server using the alternative displa
261. sponding area on the lower half of the File Staging window Now you need to provide location information For stagein type in the path and filename where you want the named file placed For stageout specify the hostname and pathname where you want the named file delivered You may repeat this process for as many files as you need to stage When you are done selecting files click the OK button 8 6 10 Stagein and Stageout Failure When stagein fails the job is placed in a 30 minute wait to allow the user time to fix the problem Typically this is a missing file or a network out age Email is sent to the job owner when the problem is detected Once the problem has been resolved the job owner or the Operator may remove the wait by resetting the time after which the job is eligible to be run via the a option to qalter The server will update the job s comment with infor mation about why the job was put in the wait state When the job is eligible to run it may run on different vnodes When stageout encounters an error there are three retries PBS waits sec ond and tries again then waits 11 seconds and tries a third time then finally waits another 21 seconds 176 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 and tries a fourth time Email is sent to the job owner if all attempts fail Files that cannot be staged out are saved in PBS_ HOME undelivered See section 12 5 6 Non delivery of Output on page
262. ssional 10 4 User s Guide Getting Started With PBS Chapter 3 3 11 2 Preventing Problems A user s job may not run if the user s start up files i e cshrc login or profile contain commands which attempt to set terminal character istics Any such command sequence within these files should be skipped by testing for the environment variable PBS_ENVIRONMENT This can be done as shown in the following sample login setenv MANPATH usr man usr local man MANPATH if PBS ENVIRONMENT then do terminal settings here endif You should also be aware that commands in your startup files should not generate output when run under PBS As in the previous example com mands that write to stdout should not be run for a PBS job This can be done as shown in the following sample login setenv MANPATH usr man usr local man MANPATH if PBS ENVIRONMENT then do terminal settings here run command with output here endif When a PBS job runs the exit status of the last command executed in the job is reported by the job s shell to PBS as the exit status of the job We will see later that this is important for job dependencies and job chaining However the last command executed might not be the last command in your job This can happen if your job s shell is csh on the execution host and you have a logout there In that case the last command executed is from the Logout and not your job To prevent this you need to p
263. structions on how to accomplish the same task using the xpbs graphical interface 7 1 The gqstat Command The qstat command is used to the request the status of jobs queues and the PBS Server The requested status is written to standard output stream usually the user s terminal When requesting job status any jobs for which the user does not have view privilege are not displayed For detailed usage information see the qstat 1B man page or the PBS Professional External Reference Specification PBS Professional 10 4 User s Guide 135 Chapter 7 Checking Job System Status Usage qstat J p t x job identifier destina tion J sss qstat f J p t x job_identifier desti Nation J qstat a w H i r G M J n 1 w s 1 w t T w u user job_id destination qstat Q f destination qstat q G M destination qstat B f server_name qstat version 7 1 1 Checking Job Status Executing the qstat command without any options displays job informa tion in the default format An alternative display format is also provided and is discussed below The default display includes the following infor mation e The job identifier assigned by PBS e The job name given by the submitter e The job owner e The CPU time used e The job state e The queue in which the job resides See Job States on page 435 of
264. t pvm To submit a PVM job to PBS use qsub your _pvm_job Here is an example script for your_pvm_job PBS N pvmjob PBS V cd PBS_O WORKDIR echo conf pvm PBS_NODEFILE echo quit pvm my_pvm program echo halt pvm 10 9 Checkpointing SGI MPI Jobs 10 9 1 Jobs on an Altix Jobs are suspended on the Altix using the PBS suspend feature Jobs are checkpointed on the Altix using application level checkpointing There is no OS level checkpoint Suspended or checkpointed jobs will resume on the original nodeboards PBS Professional 10 4 User s Guide 267 Chapter 10 Multiprocessor Jobs 268 PBS Professional 10 4 User s Guide Chapter 11 HPC Basic Profile Jobs PBS Professional can schedule and manage jobs on one or more HPC Basic Profile compliant servers using the Grid Forum OGSA HPC Basic Profile web services standard You can submit a generic job to PBS so that PBS can run it on an HPC Basic Profile Server This chapter describes how to use PBS for HPC Basic Profile jobs 11 1 Definitions HPC Basic Profile HPCBP Proposed standard web services specification for basic job execution capabilities defined by the OGSA High Perfor mance Computing Profile Working Group HPC Basic Profile Server Service that executes jobs from any HPC Basic Profile compliant client PBS Professional 10 4 User s Guide 269 Chapter 11 HPC Basic Profile Jobs HPCBP MOM MOM that sends jobs for execution to an HPC Basic Pro
265. t by the user o operator require operator privilege to unset PBS Professional 10 4 User s Guide 121 Chapter 6 Working With PBS Jobs Table 6 1 Hold Types Letter Meaning s system requires manager privilege to unset If no h option is given the user hold will be applied to the jobs described by the job_identifier operand list If the job identified by job_identifier isin the queued held or waiting states then all that occurs is that the hold type is added to the job The job is then placed into held state if it resides in an execution queue If the job is running then the following additional action is taken to inter rupt the execution of the job If the job is checkpointable requesting a hold on a running job will cause 1 the job to be checkpointed 2 the resources assigned to the job to be released and 3 the job to be placed in the held state in the execution queue If the job is not checkpointable ghold will only set the requested hold attribute This will have no effect unless the job is requeued with the grerun command See section 4 13 13 1 Check pointable Jobs on page 84 The qhold command can be used on job arrays but not on subjobs or ranges of subjobs On job arrays the qhold command can be applied only in the Q B or W states This will put the job array in the H held state If any subjobs are running they will run to completion Job arrays c
266. t specification cannot be used with a nodes specification e A select specification cannot be used with old style resource requests such as Incpus Imem lvmem larch lhost e The built in resource software is not a vnode level resource See Built in Resources on page 336 of the PBS Professional Reference Guide e A PBS job can be submitted at the command line or via xpbs e At the command line the user can create a job script and submit it During submission it is possible to override elements in the job script Alternatively PBS will read from input typed at the command line PBS Professional 10 4 User s Guide 35 Chapter 4 Submitting a PBS Job 4 3 2 Introduction to the PBS Job Script 4 3 2 1 Contents of a Job Script A PBS job script consists of e An optional shell specification e PBS directives e Tasks programs or commands 4 3 2 2 Types of Job Scripts PBS allows you to use various kinds of job scripts You can use any of the following e A Python script that can run under Windows or UNIX Linux e A UNIX shell script that runs under UNIX Linux e Windows command batch script under Windows 4 3 2 3 Submitting a Job Script Before submitting a job script using these instructions be sure to set your environment appropriately If you want the correct value for PBS EXEC to be used automatically see section 3 11 1 Setting PBS_EXEC on UNIX Linux on page 22 and section 3 12 1 Setting PBS EXEC
267. t from the keyboard 4 4 2 1 Submitting Jobs by Specifying Executables When you specify only the executable with any options and arguments PBS starts a shell for you To submit a job from the command line the for mat is the following qsub options executable arguments to executable lt return gt For example to run myprog with the arguments a and b qsub myprog a b lt return gt To run myprog with the arguments a and b naming the job JobA qsub N JobA myprog a b lt return gt 42 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 4 4 2 2 Submitting Jobs Using Keyboard Input It is possible to submit a job to PBS without first creating a job script file If you run the qsub command with the resource requests on the command line and then press enter without naming a job file PBS will read input from the keyboard This is often referred to as a here document You can direct qsub to stop reading input and submit the job by typing on a line by itself a cont rol d UNIX or cont rol z then enter Win dows Note that under UNIX if you enter a cont rol c while qsub is reading input qsub will terminate the process and the job will not be submitted Under Windows however often the cont rol c sequence will depend ing on the command prompt used cause qsub to submit the job to PBS In such case a cont rol break sequence will usually terminate the qsub command qsub
268. tA hostB hostB hostc hostc Job script PBS 1 select 3 ncpus 2 echo np 2 tmp mpitest1 gt my config file echo np 2 tmp mpitest2 gt gt my_config file echo np 2 tmp mpitest3 gt gt my_config file mpirun configfile my config file rm f my_ config file Run job qsub job script PBS Professional 10 4 User s Guide 255 Chapter 10 Multiprocessor Jobs 10 6 10 PBS Jobs with Intel MPI s mpirun PBS provides an interface to Intel MPI s mpirun If executed inside a PBS job this allows for PBS to track all Intel MPI processes so that PBS can perform accounting and have complete job control If executed outside of a PBS job it behaves exactly as if standard Intel MPI s mpirun was used You use the same mpirun command as you would use outside of PBS When submitting PBS jobs that invoke the pbsrun wrapper script for Intel MPI be sure to explicitly specify the actual number of ranks or MPI tasks in the qsub select specification Otherwise jobs will fail to run with too few entries in the machinefile For instance specification of the following in 7 1 PBS 1 select 1 ncpus 1 host hostA 1 ncpus 2 host hostB mpirun np 3 tmp mytask would result in a 7 1 PBS_NODEFILE listing hostA hostB hostB but in 8 0 or later it would be hostA hostB which would conflict with the np 3 specification in mpirun as only 2 MPD daemons will be started The correct way now is to specify either a
269. te your PBS job you can reuse it if you wish Or you can modify it for subsequent runs For example here is a simple PBS batch job UNIX bin sh PBS 1 walltime 1 00 00 PBS 1 mem 400mb ncpus 4 my_application Windows PBS 1 walltime 1 00 00 PBS 1 mem 400mb ncpus 4 my application Don t worry about the details just yet the next chapter will explain how to create a batch job of your own 3 8 PBS Interfaces PBS provides two user interfaces a command line interface CLI and a graphical user interface GUI The CLI lets you type commands at the sys tem prompt The GUI is a graphical point and click interface The user commands are discussed in this book the administrator commands are discussed in the PBS Professional Administrator s Guide The subse quent chapters of this book will explain how to use both the CLI and GUI versions of the user commands to create submit and manipulate PBS jobs Table 3 1 PBS Professional User and Manager Commands User Commands Administrator Commands Command Purpose Command Purpose PBS Professional 10 4 User s Guide 19 Chapter 3 Getting Started With PBS Table 3 1 PBS Professional User and Manager Commands User Commands Administrator Commands nqs2pbs Convert from NQS pbs report Report job statis ti pbs_rdel Delete a Reservation ii pbs_rstat Sta
270. ted Failed or Finished the HPCBP MOM will no longer query for the status of that job A job whose status is Running can become Terminated Failed or Fin ished or Exiting 11 5 1 2 Deleting jobs running at HPC Basic Profile Server You can delete your jobs via the qdel command qdel lt job ID gt 278 PBS Professional 10 4 User s Guide HPC Basic Profile Jobs Chapter 11 11 6 Errors Logging and Troubleshooting 11 6 1 Job Submission Password Problems If you specify the wrong password or the password is different from the one at the HPC Basic Profile Server e The HPCBP MOM rejects the job and the PBS Server sets the job s comment e The PBS Server logs a message in the server log e The PBS Server changes the state of the job to Hold and the substate to waiting on dependency and keeps it in the queue 11 6 2 Job Format Problems If you submit only a job script without any executable and argument list and PBS attempts to run the job on the HPCBP Server the HPCBP MOM will log a message and return an error If you submit a job requesting non HPCBP vnodes and HPCBP nodes or requesting nodes from two different HPCBP Servers e The job is rejected e The HPCBP MOM logs an error message 11 6 3 Password related Job Deletion Issues If any problem such as bad user credentials occurs during an attempt to delete a job e The qdel command displays an error message e The PBS server writes the error message to the s
271. that you can dis able job tracking at any time by clicking the Track Job button on the main xpbs display and then clicking the stop tracking button 156 PBS Professional 10 4 User s Guide Chapter 8 Advanced PBS Features This chapter covers the less commonly used commands and more complex topics which will add substantial functionality to your use of PBS The reader is advised to read chapters 5 7 of this manual first 8 1 UNIX Job Exit Status On UNIX systems the exit status of a job is normally the exit status of the shell executing the job script If a user is using csh and has a logout file in the home directory the exit status of csh becomes the exit status of the last command in logout This may impact the use of job dependen cies which depend on the job s exit status To preserve the job s exit status the user may either remove logout or edit it as shown in this example set EXITVAL Sstatus logout s original content PBS Professional 10 4 User s Guide 137 Chapter 8 Advanced PBS Features Doing so will ensure that the exit status of the job persists across the invo cation of the logout file The exit status of a job array is determined by the status of each of the com pleted subjobs It is only available when all valid subjobs have completed The individual exit status of a completed subjob is passed to the epilogue and is available in the E accounting log record of that subjob See
272. the PBS Professional Reference Guide 136 PBS Professional 10 4 User s Guide Checking Job System Status Chapter 7 The following example illustrates the default display of qstat 0 H workq 0 W workgq 00 21 03 R workq 21 09 12 R workq 0 Q workq 0 Q workq 0 Q workq qstat Job id Name User 16 south aims14 userl 18 south aims14 user1 26 south airfoil barry 27 south airfoil barry 28 south myjob user1 29 south tns3d susan 30 south airfoil barry 31 south seq 35 3 donald 0 Q workq An alternative display accessed via the a option is also provided that includes extra information about jobs including the following additional fields Session ID Number of vnodes requested Number of parallel tasks or CPUs Requested amount of memory Requested amount of wall clock time Walltime or CPU time whichever submitter specified if job is running Queue Req d Elap Jobname Ses NDS TSK Mem Time S Time aims14 1 aimsl4 1 airfoil 930 1 myjob 1 tns3d 1 airfoil 1 PBS Professional 10 4 User s Guide 0 01 W 0 13 R 0 01 0 10 Q 0 20Q 0 13 Q 137 Chapter 7 Checking Job System Status 55 south donald workq seq _35_ 1 2 009Q Other options which utilize the alternative display are discussed in subse quent sections of this chapter 7 1 2 Viewing Specific Information When requesting queue or Server status
273. the selected queue s admin only enable for enabling the selected queue s admin only The middle portion of the Queues Panel has abbreviated column names indicating the information being displayed as the following table shows Table 5 2 xpbs Queue Column Headings Heading Meaning Max Maximum number of jobs permitted Tot Count of jobs currently enqueued in any state Ena Is queue enabled yes or no Str Is queue started yes or no Que Count of jobs in the Queued state 100 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 Table 5 2 xpbs Queue Column Headings Heading Meaning Run Count of jobs in the Running state Hld Count of jobs in the Held state Wat Count of jobs in the Waiting state Tm Count of jobs in the Transiting state Ext Count of jobs in the Exiting state Type Type of queue execution or route Server Name of Server on which queue exists 5 3 4 xpbs Jobs Panel The Jobs panel is composed of a leading horizontal JOBS bar a listbox and a set of command buttons The JOBS bar lists the queues that are con sulted when listing jobs the bar also contains a minimize maximize button for displaying or iconizing the Jobs region The listbox displays informa tion about jobs that are found in the queue s selected from the Queues list box each listbox entry can be selected as described above for the Hosts panel The region just above the Jobs l
274. this process if needed An alternative but less secure method is to direct your X Windows ses sion to permit the xpbs client to connect to your local X server Do this by running the xhost command with the name of the host from which you will be running xpbs as shown in the example below xhost server mydomain com Next on the system from which you will be running xpbs set your X Windows DISPLAY variable to your local workstation For example if using the C shell setenv DISPLAY my Workstation 0 0 94 PBS Professional 10 4 User s Guide Using the xpbs GUI Chapter 5 However if you are using the Bourne or Korn shell type the following export DISPLAY my Workstation 0 0 5 2 Using xpbs Definitions of Terms The various panels boxes and regions collectively called widgets of xpbs and how they are manipulated are described in the following sec tions A listbox can be multi selectable a number of entries can be selected highlighted using a mouse click or single selectable one entry can be highlighted at a time For a multi selectable listbox the following operations are allowed e left click to select highlight an entry e shift left click to contiguously select more than one entry e control left click to select multiple non contiguous entries e click the Select All Deselect All button to select all entries or deselect all entries at once e double clicking an entry usually activates some action
275. ticular naming scheme for the new direc tories that PBS creates for staging and execution 8 6 7 2 2 User s Home Directory as Staging and Execution Directory The user must have a home directory on each execution host The absence of the user s home directory is an error and causes the job to be aborted PBS Professional 10 4 User s Guide 173 Chapter 8 Advanced PBS Features 8 6 7 3 Setting Environment Variables and Attributes PBS sets PBS_JOBDIR and the job s jobdir attribute to the pathname of the staging and execution directory The TMPDIR environment vari able is set to the pathname of the job specific temporary scratch directory 8 6 7 4 Staging Files Into Staging and Execution Directories PBS evaluates local_path and remote_path relative to the staging and execution directory given in PBS_JOBDIR whether this directory is the user s home directory or a job specific directory created by PBS PBS copies the specified files and or directories to the job s staging and execu tion directory 8 6 7 5 Running the Prologue The MOM s prologue is run on the primary host as root with the current working directory set to PBS_HOME mom_priv and with PBS _ JOBDIR and TMPDIR set in its environment 8 6 7 6 Job Execution PBS runs the job script on the primary host as the user PBS also runs any tasks created by the job as the user The job script and tasks are executed with their current working directory set to the job s st
276. time amp duration 183 soonest occurrence 179 standing 178 instance 179 soonest occurrence 179 standing reservation 181 Submitting jobs 191 reservations time for provisioning 296 Resource Specification Conversion 68 Resource specification format 68 resource list 76 resources 39 restrictions AOE 293 provisioning hosts 293 resv_nodes 179 rhosts 26 run limits job arrays 221 Running a Job Array 219 S scatter 56 Scheduler 12 Scheduling 10 job Arrays 223 scp 21 321 Index Selection of Job Arrays 220 selection statement 45 Sequence number 202 Server 11 setting job attributes 39 SGI MPI 267 share 56 sharing 55 shell 36 SIGKILL 126 SIGNULL 126 SIGTERM 126 single signon 73 Single Signon Password Method 73 SMP jobs 226 soonest occurrence 179 spec 66 spec_list 66 stageout 77 staging Windows job arrays 211 Standing Reservation 178 standing reservation 178 181 Starving job arrays 221 States job array 206 states 113 154 Status job arrays 214 stepping factor 204 Subjob 201 Subjob index 202 submission options 75 Submitting a job array 204 Submitting a PBS Job 31 suffix 66 Suppressing job identifier 91 syntax identifier 203 322 System integration 7 monitoring 6 T Task Manager 177 TCL 93 TGT 199 time between reservations 195 TK 93 tm 3 177 TMPDIR 29 tracejob 21 job arrays 219 tracejob on Job Arrays 219 tracking 155 U umask 158 Unset Resources 34 until spec 182 US mode
277. try_time attribute lists the next time when PBS will try to recon firm the reservation If PBS is able to reconfirm a degraded reservation the reservation s state becomes CO or RESV_CONFIRMED and the reservation s resv_nodes attribute shows the new vnodes 8 8 8 2 Access to Reservations By default the reservation accepts jobs only from the user who created the reservation and accepts jobs submitted from any group or host You can specify a list of users and groups whose jobs will and will not be accepted by the reservation by setting the reservation s Authorized_Users and Authorized Groups attributes using the U auth_user_ list and G auth_group_list options to pbs_rsub You can specify the hosts from which jobs can and cannot be submitted by setting the reservation s Authorized_Hosts attribute using the H auth_host_list option to pbs_rsub The administrator can also specify which users and groups can and cannot submit jobs to a reservation and the list of hosts from which jobs can and cannot be submitted For more information see the pbs_rsub 1B and pbs_resv_attributes 7B man pages 192 PBS Professional 10 4 User s Guide Advanced PBS Features Chapter 8 8 8 8 3 Viewing Status of a Job Submitted to a Reservation You can view the status of a job that has been submitted to a reservation or to an occurrence of a standing reservation by using the qstat command See qstat on page 217 of the PBS Professional Refe
278. ts are not guaranteed to be under the control of PBS 10 6 8 2 MPD Startup and Shutdown The PBS mpirun interface starts MPD daemons on each of the unique hosts listed in PBS_NODEFILE using either the rsh or ssh method based on value of environment variable RSHCOMMAND The default is rsh The interface also takes care of shutting down the MPD daemons at the end of a run If the MPD daemons are not running the PBS interface to mpirun mpd will start MX s MPD daemons as this user on the allocated PBS hosts The MPD daemons may already have been started by the administrator or by the user MPD daemons are not started inside a PBS prologue script since it won t have the path of mpirun mpd that the user executed GM or MX which would determine the path to the MPD binary PBS Professional 10 4 User s Guide 249 Chapter 10 Multiprocessor Jobs 10 6 8 3 Examples Example 1 Run a single executable MPICH MX job with 64 processes spread out across the PBS allocated hosts listed in PBS_NODEFILE PBS NODEFILE pbs hostl pbs host2 pbs host3 qsub 1 select 3 ncpus 1 MPICH MX HOME bin mpirun mpd np 64 path myprog x 1200 D lt job id gt If the MPD daemons are not running the PBS interface to mpirun mpd will start GM s MPD daemons as this user on the allo cated PBS hosts The MPD daemons may be already started by the administrator or by the user Example 2 Run an MPICH Mx job with multiple executables on multiple hosts
279. tus a Reservation pbs_hostn Report host name s pbs_ Update per user per pbs migrate Migrate per user password server password _users per server pass words pbs_rsub Submit a Reserva pbs_probe PBS diagnostic tion tool pbsdsh PBS distributed shell qalter Alter job pbs_tclsh TCL with PBS API qdel Delete job pbsfs Show fairshare usage qhold Hold a job pbsnodes Vnode manipula tion qmove Move job printjob Report job details qmsg Send message to job qdisable Disable a queue qorder Reorder jobs qenable Enable a queue qrls Release hold on job qmer Manager inter face qselect Select jobs by crite qrerun Requeue running ria job qsig Send signal to job qrun Manually start a job qstat Status job queue qstart Start a queue Server qsub Submit a job qstop Stop a queue 20 PBS Professional 10 4 User s Guide Getting Started With PBS Chapter 3 Table 3 1 PBS Professional User and Manager Commands User Commands Administrator Commands tracejob Report job history qterm Shutdown PBS xpbs Graphical User xpbsmon GUI monitoring Interface tool Notes 1 Available on Windows only 3 9 User s PBS Environment In order to have your system environment interact seamlessly with PBS there are several items that need to be checked In many cases your system administrator will have already set up your environment to work with PBS In order to use PBS to run your work the following are needed e
280. tworked compute servers smaller general systems and workstations led to the requirement of a networked batch scheduling capability The first such UNIX based system was the Network Queueing System NQS funded by NASA Ames Research Center in 1986 NQS quickly became the de facto standard for batch queueing Over time distributed parallel systems began to emerge and NQS was inadequate to handle the complex scheduling requirements presented by such systems In addition computer system managers wanted greater con trol over their compute resources and users wanted a single interface to the systems In the early 1990 s NASA needed a solution to this problem but found nothing on the market that adequately addressed their needs So NASA led an international effort to gather requirements for a next genera tion resource management system The requirements and functional speci fication were later adopted as an IEEE POSIX standard 1003 2d Next NASA funded the development of a new resource management system compliant with the standard Thus the Portable Batch System PBS was born PBS was quickly adopted on distributed parallel systems and replaced NQS on traditional supercomputers and server systems Eventually the entire industry evolved toward distributed parallel systems taking the form of both special purpose and commodity clusters Managers of such systems found that the capabilities of PBS mapped well onto cluster systems For informati
281. u UserA Serv erA UserB ServerB UserC for the job The job owner will always be Users Usernames are limited to 15 characters 86 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 4 13 14 1 qsub u User ID with UNIX The server s flatuid attribute determines whether it assumes that identical usernames mean identical users If true it assumes that if UserS exists on both the submission host and the server host then UserS can run jobs on that server If not true the server calls ruserok which uses etc hosts equiv and rhosts to authorize UserS to run as UserS Table 4 3 UNIX User ID and flatuid lue of os Va eS Submission host username server host username flatuid Same UserS UserS Different UserS UserA True Server assumes user has Server checks whether UserS permission to run job can run job as UserA Not true Server checks whether Server checks whether UserS UserS can run job as UserS can run job as UserA Note that if different names are listed via the u option then they are checked regardless of the value of flatuid 4 13 14 2 qsub u User ID with Windows Under Windows if a user has a non admin account the server s hosts equiv file is used to determine whether that user can run a job ona given server For an admin account PROFILE PATH rhosts is used and the server s acl_roots attribute must be set to allow job submissions PBS Professional 10 4 User s Gui
282. u cannot use the qalter command or any other command to alter a custom resource which has been created to be invisible or unrequestable See section 4 5 14 Resource Permissions on page 54 For more information see the qalter 1B manual page 6 2 Holding and Releasing Jobs PBS provides a pair of commands to hold and release jobs To hold a job is to mark it as ineligible to run until the hold on the job is released The qhold command requests that a Server place one or more holds on a job A job that has a hold is not eligible for execution There are three types of holds user operator and system A user may place a user hold upon any job the user owns An operator who is a user with operator privilege may place either an user or an operator hold on any job The PBS Manager may place any hold on any job The usage syntax of the ghold command is ghold h hold list job_identifier Note that for a job array the job_ identifier must be enclosed in dou ble quotes The hold_list defines the type of holds to be placed on the job The hold_list argument is a string consisting of one or more of the letters u P o or s in any combination or the letter n The hold type associated with each letter is Table 6 1 Hold Types Letter Meaning n none no hold type specified u user the user may set and release this hold type p password set if job fails due to a bad password can be unse
283. ub u Bob Jones my job 3 12 4 Windows rhosts File The Windows rhosts file is located in the user s PROF ILE_PATH for example Documents and Settings username rhosts with the format hostname username 26 PBS Professional 10 4 User s Guide Getting Started With PBS Chapter 3 IMPORTANT Be sure the rhosts file is owned by user or an adminis trator type group and has write access granted only to the owning user or an administrator or group This file can also determine if a remote user is allowed to submit jobs to the local PBS Server if the mapped user is an Administrator account For example the following entry in user susan s rhosts file on the server would permit user susan to run jobs submitted from her workstation wks031 wks031 susan Furthermore in order for Susan s output files from her job to be returned to her automatically by PBS she would need to add an entry to her rhosts file on her workstation naming the execution host Host 1 Hostl susan If instead Susan has access to several execution hosts she would need to add all of them to her rhosts file Hostl susan Host2 susan Host3 susan Note that Domain Name Service DNS on Windows may return different permutations for a full hostname thus it is important to list all the names that a host may be known For instance if Host4 is known as Host4 Host4 lt subdomain gt or Host4 lt subdomain gt lt domain gt you sho
284. uery the state of system resources and with the Server for availability of jobs to execute The interface to the Server is through the same API as used by the client commands Note that the Scheduler interfaces with the Server with the same privilege as the PBS manager 12 PBS Professional 10 4 User s Guide Chapter 3 Getting Started With PBS This chapter introduces the user to PBS Professional It describes new user level features in this release explains the different user interfaces intro duces the concept of a PBS job and shows how to set up your environ ment for running batch jobs with PBS PBS Professional 10 4 User s Guide 13 Chapter 3 Getting Started With PBS 3 1 New Features in PBS Professional 10 4 3 1 1 Estimated Job Start Times PBS can estimate the start time and vnodes for jobs See section 7 1 22 Viewing Estimated Start Times For Jobs on page 151 3 1 2 Unified Job Submission PBS allows users to submit jobs using the same scripts whether the job is submitted on a Windows or UNIX Linux system See section 4 3 3 1 Python Job Scripts on page 37 3 2 New Features in PBS Professional 10 2 3 2 1 Provisioning PBS provides automatic provisioning of an OS or application on vnodes that are configured to be provisioned When a job requires an OS that is available but not running or an application that is not installed PBS provi sions the vnode with that OS or application See Chapter 1
285. uld list all three in the rhosts file Host4 susan Host4 subdomain susan Host4 subdomain domain susan As discussed in the previous section usernames with embedded white space must also be quoted if specified in any hosts equivor rhosts files as shown below Host5 subdomain domain Bob Jones PBS Professional 10 4 User s Guide 27 Chapter 3 Getting Started With PBS 3 12 5 Windows Mapped Drives and PBS In Windows XP when you map a drive it is mapped locally to your ses sion The mapped drive cannot be seen by other processes outside of your session A drive mapped on one session cannot be un mapped in another session even if it s the same user This has implications for running jobs under PBS Specifically if you map a drive chdir to it and submit a job from that location the vnode that executes the job may not be able to deliver the files back to the same location from which you issued qsub The workaround is to use the o or e options to qsub and specify a local non mapped directory location for the job output and error files For details see section 4 13 2 Redirecting Output and Error Files on page 78 3 13 Environment Variables There are a number of environment variables provided to the PBS job Some are taken from the user s environment and carried with the job Oth ers are created by PBS Still others can be explicitly created by the user for exclusive use by PBS jobs All PBS provid
286. urce specification format lresource value resource value The resource specification is converted to lselect 1 resource value Iplace pack with one instance of resource value for each of the following vnode level resources in the resource request built in resources ncpus mem vmem arch host site defined vnode level resources 4 9 How PBS Parses a Job Script The qsub command scans the lines of the script file for directives Scan ning will continue until the first executable line that is a line that is not blank not a directive line nor a line whose first non white space character is If directives occur on subsequent lines they will be ignored 68 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 A line in the script file will be processed as a directive to qsub if and only if the string of characters starting with the first non white space character on the line and of the same length as the directive prefix matches the direc tive prefix i e PBS The remainder of the directive line consists of the options to qsub in the same syntax as they appear on the command line The option character is to be preceded with the character If an option is present in both a directive and on the command line that option and its argument if any will be ignored in the directive The com mand line takes precedence If an option is present in a directive and not on the co
287. urrences i e for three weeks pbs_rsub R 1500 E 1700 r FREQ WEEKLY BYDAY MO WE FR COUNT 9 8 8 5 The pbs rsub Command The pbs_rsub command returns a reservation ID string and the current status of the reservation For the options to the pbs_rsub command see pbs_rsub on page 84 of the PBS Professional Reference Guide 8 8 5 1 Getting Confirmation of a Reservation By default the pbs_rsub command does not immediately notify you whether the reservation is confirmed or denied Instead you receive email with this information You can specify that the pbs_rsub command should wait for confirmation by using the I lt block_time gt option The pbs_rsub command will wait up to lt block_time gt seconds for the reser PBS Professional 10 4 User s Guide 185 Chapter 8 Advanced PBS Features vation to be confirmed or denied and then notify you of the outcome If block_time is negative and the reservation is not confirmed in that time the reservation is automatically deleted To find out whether the reservation has been confirmed use the pbs_rstat command It will display the state of the reservation CO and RESV_CONFIRMED indicate that it is confirmed If the reservation does not appear in the output from pbs_rstat that means that the reser vation was denied To ensure that you receive mail about your reservations set the reserva tion s Mail_Users attribute via the M lt email address gt option to pbs_rsub By
288. ute and the optional SS is the seconds If the month MM is not speci fied it will default to the current month if the specified day DD is in the future Otherwise the month will be set to next month Likewise if the day D is not specified it will default to today if the time hhmm is in the PBS Professional 10 4 User s Guide 83 Chapter 4 Submitting a PBS Job future Otherwise the day will be set to tomorrow For example if you sub mit a job at 11 15am with a time of 1110 the job will be eligible to run at 11 10am tomorrow Other examples include qsub a 0700 my_job PBS a 10220700 4 13 12 Holding a Job Delaying Execution The h option specifies that a user hold be applied to the job at submis sion time The job will be submitted then placed in a hold state The job will remain ineligible to run until the hold is released For details on releasing a held job see Holding and Releasing Jobs on page 121 qsub h my_job PBS h 4 13 13 Specifying Job Checkpoint Interval 4 13 13 1 Checkpointable Jobs A job is checkpointable if any of the following is true e Its application supports checkpointing and there are checkpoint scripts e There is a third party checkpointing application available e The OS supports checkpointing Checkpoint scripts are set up by the local system administrator 4 13 13 2 Queue Checkpoint Intervals The execution queue in which the job re
289. ute second section by a capital T Default 3 years from time of reservation creation 8 8 4 1 Setting Reservation Start Time and Duration In a standing reservation the arguments to the R and E options to pbs_rsub can provide more information than they do in an advance res ervation In an advance reservation they provide the start and end time of the reservation In a standing reservation they can provide the start and end time but they can also be used to compute the duration and the offset from the interval start The difference between the values of the arguments for R and E is the duration of the reservation For example if you specify R 0930 E 1145 the duration of your reservation will be two hours and fifteen minutes If you specify R 150800 E 170830 the duration of your reservation will be two days plus 30 minutes The interval spec can be used to specify the day or the hour at which the interval starts If you specify R 0915 E 0945 BYHOUR 9 10 the duration is 30 minutes and the offset is 15 minutes from the start of the interval The interval start is at 9 and again at 10 Your reservation will run from 9 15 to 9 45 and again at 10 15 and 10 45 Similarly if you specify R 0800 E 1000 BYDAY WE TH the duration is two hours and the offset is 8 hours from the start of the inter val Your reservation will run Wednesday from 8 to 10 and again on Thursday from 8 to 10 Elements specified in t
290. verride directives 4 3 3 5 The User s Tasks These can be programs or commands This is where the user specifies an application to be run 4 3 3 6 Setting Job Attributes Job attributes can be set by either of the following methods e Using PBS directives in the job script e Giving options to the qsub command at the command line These two methods have the same functionality Options to the qsub com mand will override PBS directives which override defaults Some job attributes have default values preset in PBS Some job attributes default values are set at the user s site After the job is submitted you can use the gqalter command to change the job s characteristics Job attributes are case insensitive 4 3 3 7 Debugging Job Scripts You can run Python interactively outside of PBS to debug a Python job script You use the Python interpreter to test parts of your script Under UNIX Linux use the i option to the pbs_ python command for example opt pbs default bin pbs python i lt return gt PBS Professional 10 4 User s Guide 39 Chapter 4 Submitting a PBS Job Under Windows the i option is not necessary but can be used For exam ple either of the following will work C Program Files PBS Pro exec bin pbs_python exe lt return gt C Program Files PBS Pro exec bin pbs python exe i lt return gt When the Python interpreter runs it presents you with its own prompt For example opt pbs def
291. w you may click on any of the common signals or you may enter the signal number or signal name you wish to send to the job Click the Signal button to complete the process 6 6 Changing Order of Jobs PBS provides the qorder command to change the order of two jobs within or across queues To order two jobs is to exchange the jobs posi tions in the queue or queues in which the jobs reside Ifjob1 is at position 3 in queue A and job2 is at position 4 in queue B qordering them will result in job being in position 4 in queue B and job2 being in position 3 in queue A The two jobs must be located at the same Server and both jobs must be owned by the user No attribute of the job such as priority is changed The impact of changing the order within the queue s is dependent on local job scheduling policy contact your systems administrator for details IMPORTANT A job in the running state cannot be reordered Usage of the gorder command is gorder job_identifierl job_identifier2 Job array identifiers must be enclosed in double quotes Both operands are job_identifiers which specify the jobs to be exchanged qstat u bob Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time PBS Professional 10 4 User s Guide 127 Chapter 6 Working With PBS Jobs 54 south bob workq twinkie 1 0 20 Q 63 south bob workq airfoil 1 0 13 Q gorder 54 63 qstat u bob Req d Elap Job ID User Queue Job
292. ware is subject to the restrictions set forth in this Agreement and in the Commercial Computer Software Restricted Rights FAR clause 52 227 19 June 1987 and any restrictions in successor regulations thereto Portions of Altair s PBS Professional Software and Documentation are pro vided with RESTRICTED RIGHTS Use duplication or disclosure by the Government is subject to restrictions as set forth in subdivision c 1 ii of the rights in the Technical Data and Computer Software clause in DFARS 252 227 7013 or in subdivision c 1 and 2 of the Commercial Com puter Software Restricted Rights clause at 48 CFR52 227 19 as applica ble 10 CHOICE OF LAW AND VENUE This Agreement shall be gov erned by and construed under the laws of the state of Michigan without regard to that state s conflict of laws principles except if the state of Michi gan adopts the Uniform Computer Information Transactions Act drafted by the National Conference of Commissioners of Uniform State Laws as revised or amended as of June 30 2002 UCITA which is specifically excluded This Agreement shall not be governed by the United Nations Convention on Contracts for the International Sale of Goods the applica tion of which is expressly excluded Each Party waives its right to a jury trial in the event of any dispute arising under or relating to this Agreement Each party agrees that money damages may not be an adequate remedy for breach of the provisions of thi
293. words require the W option to qsub 206 PBS Professional 10 4 User s Guide Job Arrays Chapter 9 9 6 1 Specifying Files To Be Staged In or Staged Out You can specify files to be staged in before the job runs and staged out after the job runs by using Ww stagein file list and W stage out file list You can use these as options to qsub or as directives in the job script The file_list takes the form local_path hostname remote_path for both stagein and stageout The name ocal_path is the name of the file in the job s staging and execu tion directory on the execution host The local_path can be relative to the job s staging and execution directory or it can be an absolute path The character separates the local specification from the remote specifi cation The name remote_path is the file name on the host specified by hostname For stagein this is the location where the input files come from For stage out this is where the output files end up when the job is done You must specify a hostname The name can be absolute or it can be relative to the user s home directory on the remote machine IMPORTANT It is advisable to use an absolute pathname for the remote_path Remember that the path to your home direc tory may be different on each machine and that when using sandbox PRIVATE you may or may not have a home directory on all execution machines For stagein the direction of travel is from remot
294. xample above it would contain 4 lines The vari ables NCPUS and OMP_NUM_THREADS will be set to one If you do not care where the four MPI processes are run you may submit qsub 1 select 4 ncpus 1 l place free and the job will run on 2 3 or 4 hosts depending on what is available For this example PBS_NODEFILE will contain 4 entries either four sepa rate hosts or 3 hosts one of which is repeated once or 2 hosts etc NCPUS and OMP_NUM_THREADS will be set 1 or 2 depending on the number of cpus allocated from the first listed host 226 PBS Professional 10 4 User s Guide Multiprocessor Jobs Chapter 10 10 3 1 The mpiprocs Resource The number of MPI processes for a job is controlled by the value of the resource mpiprocs The mpiprocs resource controls the contents of the PBS_NODEFILE on the host which executes the top PBS task for the PBS job the one executing the PBS job script See Built in Resources on page 336 of the PBS Professional Reference Guide The PBS_ NODEFILE contains one line per MPI process with the name of the host on which that process should execute The number of lines in PBS_NODEFILE is equal to the sum of the values of mpiprocs over all chunks requested by the job For each chunk with mpiprocs P where P gt 0 the host name the value of the allocated vnode s resources_available host is written to the PBS_NODEFILE exactly P times If a user wishes to run two MPI processes on each of 3 hosts and have
295. y showing all size information in mega words MW rather than the default of smallest displayable units A word is considered to be 8 bytes 7 1 11 List Hosts Assigned to Jobs The n option to qst at displays the hosts allocated to any running job at the optionally specified PBS Server in addition to the other informa tion presented in the alternative display The host information is printed 144 PBS Professional 10 4 User s Guide Checking Job System Status Chapter 7 immediately below the job see job 51 in the example below and includes the host name and number of virtual processors assigned to the job i e south 0 where south is the host name followed by the virtual processor s assigned A text string of is printed for non running jobs Notice the differences between the queued and running jobs in the example below qstat n Req d Elap Job ID User Queue Jobname Sess NDS TSK Mem Time S Time 16 south userl workq aims14 1 0 01 H 18 south userl workq aims14 1 0 01 W 51 south barry workq airfoil 930 1 0 13 R 0 01 south 0 52 south userl workgq my_job 1 0 10 Q 7 1 12 Display Job Comments The s option to qst at displays the job comments in addition to the other information presented in the alternative display The job comment is printed immediately below the job By default the job comment is updated by the Scheduler with the reason why a
296. y held due to an invalid password are released s server Allows user to specify server where password will be changed d Deletes the password user The password credential of user user is updated to the prompted password Ifuseris not the current user this action is only allowed if 1 The current user is root or admin 2 User user has given the current user explicit access via the ruserok mechanism a The hostname of the machine from which the cur rent user is logged in appears in the server s hosts equiv file or b The current user has an entry in user s HOMEDIR rhosts file Note that pbs_password encrypts the password obtained from the user before sending it to the PBS Server The pbs_password command does not change the user s password on the current host only the password that is cached in PBS The pbs password command is supported only on Windows and all supported Linux platforms on x86 and x86_64 74 PBS Professional 10 4 User s Guide Submitting a PBS Job Chapter 4 4 12 2 2 Per job Password Method If you are running in a password protected Windows environment but the single signon method has not been configured at your site then you will need to supply a password with the submission of each job You can do this via the qsub command with the Wpwd option and supply the password when prompted qsub Wpwd lt job script gt You will be prompted for the password which is passed on to

Download Pdf Manuals

Related Search

Related Contents

Samsung SGH-E210 Manuel de l'utilisateur  Hardware und Installationsanleitung  Mode d`emploi des Marchés Publics  systèmes experts - application a la modelisation en  ACU-250  Manual de instrucciones para el tutor del centro  共通取扱説明書 [PDF形式]  

Copyright © All rights reserved. Failed to retrieve file