Best practices for deploying VMware vSphere 5 with VMware High
Contents
1. quide pdf With many customer instances one might already have a Multi Site SAN cluster running VMware HA Incorporating VMware FT will be the next step for several customers It will be important to identify which virtual machines will be protected with VMware FT Enabling the VM for VMware FT is a one step task to complete To create a cluster in the vSphere Client right click on the Datacenter name and select New Cluster The popup window below shows creating a Cluster called FT_Site and selecting Turn On vSphere HA which is required in order to use Fault Tolerance Also select Turn on vSphere DRS in order to support the management monitoring placement and migration of virtual machines with Fault Tolerance turned on Figure 8 Example of creating a cluster and selecting HA and DRS vl New Cluster Wizard Miel Cluster Features What Features do you want to enable For this cluster Cluster Features vSphere DRS Power Management vSphere HA Virtual Machine Options yM Monitoring Cluster Features E Pi i Select the features you would like to use with this cluster YM Swapfile Location Ready to Complete M Turn On vSphere HA Sphere HA detects Failures and provides rapid recovery For the virtual machines running within a cluster Core Functionality includes host and virtual machine monitoring to minimize downtime when heartbeats cannot be detected vSphere HA must be turned on to use Fault Tolerance2. i Turn on vSphere DRS vSphere DRS enables yCenter Server to manage hosts as an aggregate pool of resources Cluster resources can be divided into smaller resource pools For users groups and virtual machines vSphere DRS also enables yCenter Server to manage the assignment of virtual machines to hosts automatically suggesting placement when virtual machines are powered on and migrating running virtual machines to balance load and enforce resource allocation policies vSphere DRS and VMware EVO should be enabled in the cluster in order to permit placing and migrating YMs with Fault Tolerance turned on during load balancing Help Back Cancel E 13 In preparation for deploying the infrastructure it will be necessary to work with the network administrator to properly plan for the network traffic VMware suggests isolating the network traffic for VMware FT logging from virtual machines iSCSI storage and vMotion Below is a screenshot of the test setup showing two separate vmnic devices for the main host network two for a vMotion network and two for the Fault Tolerant network This is done in order to provide sufficient network bandwidth for each network and to ensure minimum required response times for the vMotion and FT network requirements Figure 9 Example of three isolated networks the general VM network and Management the vMotion network and the FT logging network Mware ESXi 5 0 0 469512 Getting Started S
3. AM Enabled Disabled Unknown B Heartbeat_1 Normal LEFTHAND iSCSI Non SSD 1 75 GB 1 21 GB VMFSS 1 11 2012 11 19 10 AM Enabled Disabled Supported B Shared_Datastore Normal LEFTHAND iSCSI Non SSD 5 00 TB 4 99 TB VMFSS 1 11 2012 11 19 10 AM_ Enabled Disabled Supported X Storage Networking Storage Adapters Network Adapters Advanced Settings Power Management Software Licensed Features Time Configuration DNS and Routing Authentication Services Power Management virtual Machine Startup Shutdown Virtual Machine Swapfile Location gt Security Profile Datastore Details Properties Host Cache Configuration System Resource Allocation Agent M Settings Advanced Settings After creating a cluster and choosing the cluster options it is possible to drag and drop hosts into the cluster Remember that the host servers selected should be FT compliant with the same FT version or host build number Additionally ensure the servers have the same number of cores memory capacity and firmware levels so the virtual machines can run successfully on any of the hosts 15 Also be sure to select Put all of the host s virtual machines in the cluster s root resource pool This will ensure that the virtual machine information is always available Figure 11 Example of designating the location of the resource pool Mware ESXi 5 0 0 469512 Getting Started Summary Virtual Machine
4. Cache Configuration Name Crafted from i System Resource Allocation Agent M Settings Advanced Settings Help lt Back net gt Cancel 16 VMware FT can be enabled in the virtual machine menu Ensure that the secondary VMs created by fault tolerance are placed on an ESXi host in the secondary site This information can be seen from the Summary tab as show in red in Figure 12 Figure 12 FT virtual machine location APS40 28 ActiveDirectory DNS1 E Demo Datacenter dh FT_Site Getting Started Summary Resource Allocation Performance J aps41 187 J aps41 188 J aps41 189 Guest OS Microsoft Windows Server 2008 R2 64 J aps41 190 YM Version 8 CPU 1 vCPU OB Clienti Memory 4096 MB GB Exchange Memory Overhead 107 38 MB Mware Tools Running Current IP Addresses 10 136 40 27 DNS Name aps40 27 prod hp local EVC Mode Intel Nehalem Gen Xeon Core State Powered On Host aps41 187 Active Tasks vSphere HA Protection Protected E Shut Down Guest DI Suspend A Restart Guest E Edit Settings E Open Console p Migrate Tasks amp Events Alarms Console Resources Consumed Host CPU Consumed Host Memory Active Guest Memory Provisioned Storage Not shared Storage Used Storage Permissions Maps Storage Views 0 MHz 3219 00 MB 122 00 MB Refresh Storage Usage 40 05 GB 49 62 MB 40 05 GB Storage Status Drive Type S
5. G7 and 2 HP P4800 storage blades vSphere FT VM Secondary F z c7000 with 3 BL460cG7 and 2 HP P4800 storage blades HP LeftHand Multi Site Cluster vSphere 5 HP P4800 storage Cluster The result when the ESXi hosts on Site A went down due to a power outage was that the Primary VMs running on Blade 1 and Blade 2 were no longer available The Secondary VMs running on Blade 3 and Blade 4 now became the Primary VMs and new Secondary VMs were spawned on Blade 3 and Blade 4 on site B Since there was no disruption to the SAN the VMs continued running It is important to note that the VMs must be migrated back to the primary site once it becomes available again 20 The second test consisted of powering off storage nodes on site A HP manually powered off the storage nodes on site A The Primary VMs were running on Blade 1 and Blade 2 The Secondary VMs were running on Blade 3 and Blade 4 Both VMs had access to the same LUN or datastore which was protected by a Multi Site cluster Figure 16 Powering off the storage nodes on Site A vSphere Seay a oye Secondary rimary 7000 with 3 BL460c G7 _ 7000 with 3 BL460c G7 and 2 HP P4800 storage and 2 HP P4800 storage blades blades HP LeftHand Multi Site Cluster vSphere 5 HP P4800 HP P4800 storage storage Cluster The result demonstrates that even when power to the storage nodes was
6. Multi Site SANs even entire site wide failures are protected When the primary VM goes down due to an ESXi host failure or an entire site failure the secondary VM immediately takes over without interruption or data loss and starts communicating with the outside world The secondary VM now becomes the primary protected VMware FT VM while residing on a separate host in the same vSphere cluster It is important to emphasize that during a component or site failure there is no failover activity in the SAN A failover to the other SAN or mirrored volume is not necessary The data volumes continue to exist due to network RAID and no action is required on the part of the SAN administrator Note While Multi Site SANs do not require action after a site failure VMware FT does During a site wide failure VMware FT will create new secondary VMs on one of the remaining hosts After the original site comes back up a VMware administrator will need to migrate the newly created secondary FT VMs back to the original site non disruptively using vMotion This is required to prevent both the primary and secondary VMs from residing on the same site Hardware and software requirements There are numerous requirements to consider when configuring the solution discussed in this white paper The following sections cover the areas that need attention in planning and deployment VMware has written a document covering VMware FT and VMware HA configuration as well
7. Server managing the HA DRS cluster can connect to the ESXi hosts at both sites The storage network also must be configured with redundant links in vMSC configurations The diagram in Figure 18 provides a high level overview Figure 18 vSphere Metro Storage Cluster with HP LeftHand vSphere Metro Storage Cluster Architecture on HP LeftHand Storage Site 1 Site 2 Recommended Redundant Management 7 VM vMotion Networks Required Redundant Storage Network e a y WE o 5 P e e SHB ed A ili a a ES a Site 3 GHH Hiie ees E un Ge HP LeftHand Failover Manager resides on 3 Site For more information Best Practices for VMware vSphere HA Clusters http pubs vmware com vsphere 50 index jsp topic com vmware vsphere avail doc_50 GUID 39731BEC EBOC 48C9 813B CAF9DE884FD5 html Best Practices for Fault Tolerance http pubs vmware com vsphere 50 index jsp topic com vmware vsphere avail doc_50 GUID 2A87B967 5BBB 4394 B1E6 EE87D50A2A86 html HP P4000 Multi Site HA DR Solution Pack user guide http bizsupport2 austin hp com bc docs support SupportManual c03041871 c03041871 pdf HP P4000 SAN networking recommendations http h20195 www2 hp com V2 GetDocument aspx docname 4AA2 5615ENW HP P4000 LeftHand SAN Solutions with VMware vSphere Best Practices http h20195 www2 hp com V2 GetDocument aspx docname 4AA3 6918ENW VMware Fault Tolerance FAQ http kb vmware com kb 1013428 Processors and guest o
8. another ESXi host at the other site So as soon as they reboot the VMs will locate their data volumes on the Multi Site SAN and continue to operate The application and user will see minimal interruption Figure 2 shows a diagram of the hardware that was utilized for this characterization effort Figure 2 Hardware utilized in test configuration HHH E HP network switch vSphere ene FT VM Secondary Primary noe c7000 with 3 BL460c G7 and 2 HP P4800 storage blades c7000 with 3 BL460c G7 and 2 HP P4800 storage blades HP LeftHand Multi Site Cluster vSphere 5 HP P4800 HP P4800 storage storage Cluster Incorporating VMware FT VMware extends their high availability capability with the addition of VMware FT With VMware HA all VMs in the HA cluster are protected from server failure by being restarted on another healthy ESXi host within the same cluster Using VMware FT all the FT enabled VMs in the HA cluster experience no data loss and no service interruption in the event of a host failure Of course the other VMs in the HA cluster are still protected by VMware HA but will experience a disruption of service as they are restarted elsewhere in the cluster VMware FT provides continuous availability by running two identical VMs on two separate ESXi hosts Continuous availability of the VM is ensured when protecting VMs with VMware FT for hardware or component failures With
9. as best practices specific to these features The guide is located at http pubs vmware com vsphere 50 topic com vmware Cbase PDF vsphere esxi vcenter server 50 availability guide pdf HP suggests testing your proof of concept before deploying the solution HP hardware With respect to the hardware used for the solution this document will discuss HP c Class server blades LeftHand storage nodes and HP Networking In addition HP has also written documents describing the Multi Site SAN solution which explain the best practices for LeftHand storage nodes for VMware vSphere HP P4000 Multi Site HA DR Solution Pack User Guide http bizsupport2 austin hp com bc docs support SupportManual c03041871 c03041871 pdf HP P4000 LeftHand SAN Solutions with VMware vSphere Best Practices http h20195 www2 hp com V2 GetDocument aspx docname 4AA3 6918ENW HP c Class blades It is important to have identical server environments at the two sites For example if Datacenter A has a c7000 enclosure with four ProLiant BL460c G7 servers the secondary site for Datacenter B must utilize the same enclosure and server models This also means the firmware versions for HP Virtual Connect modules Onboard Administrator OA and server BIOS between sites must be the same It is also important to maintain configuration consistency between the two sites For example server profiles that are used in Datacenter A should be the same for Datacenter B This will
10. cut abruptly the virtual machines continue running In this case the Failover Manager kept quorum and noticed that storage on the primary site was no longer available Site B storage nodes were still in operation and processing I Os to the VMs on site A The CMC console reported the storage nodes being offline for site A Site B storage nodes performed in a degraded mode due to the lost storage nodes 21 The third test consisted of powering off the entire blade enclosure including the storage nodes on site A The Primary VMs were running on Blade 1 and Blade 2 The Secondary VMs were running on Blade 3 and Blade 4 HP abruptly powered off the entire site A to simulate a total loss Since the test configuration was running AD DNS services on Blade 1 and MS Exchange on Blade 2 in site A HP replicated a total site failure including the storage nodes The result when the entire site went down was that the primary VMs running on Blade 1 and Blade 2 were no longer available The Secondary VMs running on Blade 3 and Blade 4 now became the Primary VMs and new Secondary VMs were spawned on Blade 3 and Blade 4 on site B The applications experienced no downtime during the failure of an entire site Figure 17 Site A failure sAHHhHHa HP network switch vSphere eae A oe Secondary rimary Et c7000 with 3 BL460c G7 P4808 and 2 HP P4800 storage blades blades e P LeftHand Multi Site Cluster vSphere 5
11. data is persisted between each source snapshot Note that every write including a delete is a delta block persisted in the SmartClone delta space If long term space efficiency is required with SmartClone volumes minimizing writes to the SmartClone datastores include avoiding defragmentation within the guest VM Successful approaches have also included separation of User and Application data including file redirection Space reclamation can only be performed with a SmartClone volume by creating a new one retaining the original small source reference with minimal delta data By removing OS dependence in separation from the User and Applications periodic SmartClone deletion and re creation ensures that data delta growth is minimized Without these approaches SmartClone volumes may eventually occupy an entire SmartClone s volume space as delta change data SmartClone s initial value is in immediate cloning of golden image volumes Efficient space utilization objectives need to understand use and mitigation approaches to maximize success vSphere Metro Storage Cluster VMSC using HP LeftHand Multi Site vSphere Metro Storage Cluster vMSC vSphere Metro Storage Cluster vMSC is anew VMware certified configuration for stretched storage cluster configurations A vMSC configuration is designed to maintain data availability beyond a single physical or logical site A storage device configured in the MSC configuration is supported after successful vMSC ce
12. data recovery does not depend upon a paused or stopped VM state and flushed data to the SAN The configuration of the Application Aware Snapshot Manager is performed within the CMC Users configure the IP address of the Virtual Center server The Centralized Management Console communicates with the vSphere vCenter Server s during the snapshot process vCenter Server quiesces VMs and takes a vSphere snapshot VMware Tools within each guest VM quiesces applications Once quiesced the SAN performs a hardware based snapshot which contains vSphere VM snapshots when examined through snapshot manager and represented to vSphere hosts Snapshots of Raw Devices LeftHand SAN snapshots of vSphere 5 Raw Device Mappings RDMs are supported in exactly the same way as for physical servers either booting from the SAN or accessing LUNs on the SAN Detailed information about LeftHand snapshots and how they work for specific applications is available on the HP website at http www hp com go lefthand In vSphere 5 two compatibility modes for RDMs are available Virtual compatibility mode allows an RDM to act exactly like a virtual disk file including the use of vSphere based snapshots Physical compatibility mode allows direct access of the LUN device for applications that need lower level direct raw device control RDMs offer some benefits including e Dynamic Name Resolution the identification associates each RDM with the device regardless of physical server chan
13. i Total Mi ac Beech M Monitoring Disabled een Application Monitoring Disabled Total Storage 6 07 TB Advanced Runtime Info Number of Hosts 4 Cluster Status Total Processors 48 Configuration Issues Number of Datastore Clusters 0 vSphere DRS Total Datastores 6 Migration Automation Level Manual Virtual Machines and Templates 9 Power Management Automation Level Automatic Sokal Arat OFORO 2 DRS Recommendations 0 otal Migrations using vMotion DRS Faults 0 Commands Gt New Virtual Machine iP Add Host amp New Resource Pool GP New Datastore Cluster gfh Edit Settings Migration Threshold Target host load standard deviation Current host load standard deviation View Resource Distribution Chart view DRS Troubleshooting Guide Apply priority 1 priority 2 and priority 3 recommendations lt 0 141 0 006 Load balanced Solution tests Test hardware software configuration Storage Storage resources z Status Drive Type Capacity 3 datastorel Normal Non SSD 274 25 GE datastorel 1 Normal Non SSD 274 25 GE 3 datastorel 2 Normal Non S5D 274 25 GE amp datastorel 3 Normal Non SSD 274 25 GE SS Heartbeat_1 Normal Non SSD 1 75 GE E Shared_Datastore Normal Non SSD 5 00 TE aq gt Storage Views Please see the Hardware and software requirements section above for minimum hardware configuration requirements Hardware The following hardwa
14. make it easier to configure and install VMware vSphere VMware FT and VMware HA HP Networking Networking is a critical component of the overall solution Bandwidth and latency are significant considerations in the deployment of the solution described in this white paper Both VMware FT and HP LeftHand Multi Site SAN solutions have minimum latency requirements HP LeftHand Multi Site SAN clusters require less than 2ms round trip VMware FT however generally requires 1ms round trip latencies As such the solution outlined requires less than 1ms round trip latencies VMware and HP recommend isolating the traffic for VMware FT logging It is also recommended to use separate VLANs for the network settings of ESXi hosts Dual redundant network paths are an ideal solution to achieve the redundancy This will require two NICs each for vMotion VMware FT logging iSCSI VM and the host Service Management Console VMware FT logging traffic can be shared with the vMotion network but it is not recommended When a network port becomes saturated excess frames can be dropped because the port cannot handle the amount of network traffic received Packets will need to be resent causing a decrease in performance Flow control eliminates this problem by controlling the speed at which data is sent to the port For the iSCSI network ports enabling Flow Control supported by HP networking and Virtual Connect at the physical switch is recommended by both HP and VMw
15. the SAN can be expanded updated configured and maintained without the need to disconnect the systems accessing the storage 10 In order to support a multi site SAN solution with VMware vSphere 5 FT at least two iSCSI volumes are needed for a VMware High Availability heartbeat For each ESXi host to access each volume you must create the required volumes define each host in the CMC and export the volumes to all ESXi hosts To define a new server right click the management group from the left navigation and select New Server Figure 6 shows the new server creation screen Figure 6 New Server creation Name Description ESXi5_Host Controlling Server IP Address 192 168 2 12 VVhat IP address do use iSCSI Security Allow access via iSCSI Initiator Node Name lant 990 07 com emulex 9c Be 99 23 92 24 How do find my initiator node name V Enable load balancing Information on compliant initiators Enabling load balancing on non compliant initiators can compromise volume availability To function correctly load balancing requires that the cluster virtual IP be configured Authentication CHAP not required CHAP required Enter the IP address of the VMware vCenter Server as the Controlling Server IP Address A single subnet with a single Virtual IP VIP is recommended for a LeftHand Multi Site and vSphere 5 environment Select the option Allow access via iSCSI and enter t
16. 0 availability guide pdf In most cases customers already have HP Multi Site clusters running in their environments which makes enabling Fault Tolerance a simple one step process Note that VMware DRS may be used to load balance VMware FT VMs in a DRS enabled cluster when Enhanced vMotion Compatibility EVC is enabled and DRS is set to fully automated mode If EVC is disabled VMware FT VMs will not be moved for load balancing purposes EVC can be enabled in the cluster settings within VMware vSphere For further information please see the Using Fault Tolerance with DRS section of the vSphere Availability document listed above Solution requirements For the purposes of this document HP implemented the best practices for both VMware FT and LeftHand Multi Site SAN Having a good understanding of how to configure LeftHand managers is a must A collection of one or more storage systems can be defined as a management group Each management group serves as a container for clustering storage systems and creating volumes for storage LeftHand managers govern the activity of all the storage systems within a management group If the network infrastructure encounters unplanned downtime a Multi Site SAN cluster implementation must protect the SAN in case of a split brain network scenario A Failover Manager FOM handles automated data failover It is designed to run as a virtual appliance in a vSphere environment The FOM participates i
17. 00sb7a f P4800sb7b f P4800sb8a P4800sb8b FI bA List of IP addresses to search for IP Address Status Hostname Model g 192 168 2 5 Found P4800sb7a P4460sb G2 E 192 168 2 6 Found P4800sb8a P4460sb G2 o virtualize your storage network and 192 168 2 8 Found P4800sb7b P4460sb G2 i 192 168 2 9 Found P4800sb8b P4460sb G2 tg 192 168 2 11 Found Failover Manager Hyp our network topography you need to u define those permissions Ta mem Remove Auto Discover by Broadcast Status Normal Fitter Alarms Severity Date Time Event Message Type Name Component Ch 4 T e Alarm Tasks The first step in creating a multi site SAN is to create a management group A management group is a container where storage systems can be clustered and volumes can be created To create a management group launch the Management Groups Clusters and Volumes Wizard from the Centralized Management Console Select all the systems to be managed as a group and click Next The next screen will allow you to create an administrative user if one does not already exist for the targeted systems Figure 4 Management Groups Clusters and Volumes Wizard fe HP P4000 Centralized Management Console Oh x Fie Find Tasks Helpy 4 Upgrades Available tai SAN Status fe Management Groups Clusters and olumes Wizard xi d Getting Star Ast Configure Create a Management Group Hi Available S
18. 5 and Site Recovery Manager 5 require LeftHand OS v9 5 Application Integration Solution Pack An option for SRA for SRM 5 must be selected for vSphere 5 and Site Recovery Manager 5 support The previous version supported SRM 4 1 ina single installer option LeftHand OS v9 5 supports SRM 1 x 4 x or SRM 5 A best practice is to ensure that LeftHand hardware and software are always up to date Note that automated failback is now supported as part of the SRM 5 features SmartClone volumes LeftHand SmartClone volumes may also be used in a vSphere 5 environment Using HP LeftHand SmartClone technology all the virtual machines stored on a single volume can be cloned instantly without replicating data SmartClone volumes consume space for data changes only from the time the SmartClone volume was created SmartClone volumes are the best way to deploy small quantities of cloned golden image VMs or virtual desktops Manageability and deployment of Virtual Desktop Infrastructure VDI environments are best to leverage VMware View implementations with Linked Clones Please refer to the HP VDI Reference Architectures at http h18013 www1 hp com products servers vdi SmartClone volumes can be used seamlessly with other LeftHand software features including snapshots or Remote Copy SmartClone volumes are also very useful for performing tests on virtual machines by quickly reproducing them without actually using space on the SAN to copy them Unlike snapshots delta
19. HP P4800 HP P4800 storage storage Cluster In all three cases described when HP powered on the c7000 enclosure or powered on the storage node the action brought an entire site back up Once power was restored everything continued to function with site B as the primary site It is important to note that once site A is powered back on one set of VMs either primary or secondary must be manually migrated back to the primary site to establish fault tolerance again Implementing a proof of concept As a matter of best practice for all deployments HP recommends implementing a proof of concept using a test environment that matches the planned production environment as closely as possible In this way appropriate configuration and solution deployment can be obtained For help with a proof of concept and to ensure your configuration is supported contact an HP Services representative or your HP partner http www hp com large contact enterprise index html 22 Summary In summary an HP LeftHand SAN solution offers simple management and cost effective high availability HP LeftHand SANs are built on a clustered storage architecture that eliminates any single point of failure and provides always online volumes for VMware HA and FT in the event of a variety of failure scenarios This solution is a highly available IT infrastructure that is simple to implement and manage Appendix A Additional features of HP LeftHand with vSphere 5 LeftHand
20. Name your management group and select systems if to identify your g n by selecting g ipara b your network s vw P P4800s Management Group Name iMwareHA iil P4800s This name cannot be changed after the management group is created onsole in order Create management group with systems selected in the table below IP Address RAID Configura Software Versi Disk Type P4800sb7a 192 168 2 5 P4460sb G2 RAIDS 9 5 00 1215 0 P4800sb7b 192 168 2 8 P4460sb G2 RAID 5 9 5 00 1215 0 etwork and P4800sb8a 192 168 2 6 P4460sb G2 RAID 5 9 5 00 1215 0 P4800sb8b 192 168 2 9 P4460sb G2 RAID 5 9 5 00 1215 0 FOM 192 168 2 11 Failover Manage RAID virtual 9 5 00 1215 0 you need to S To continue click Next ponent Ch Alarm Tasks Next you will be able to select the Network Time Protocol NTP and Domain Name Servers DNS for the management group Afterwards the wizard will prompt the creation of a storage cluster Select the option for a multi site cluster and click Next Select the discovered LeftHand nodes from all sites and the Failover Manager to be part of the cluster Now the wizard will prompt you to create multiple sites for a cluster Create one site for each physical site and assign the storage nodes to the sites where they physically reside The Make this site primary checkbox can be selected to determine which site will maintain quorum in the event that the link between sites is lost This is o
21. SANs along with VMware vSphere 5 provide an excellent solution to improve your hardware utilization and system availability LeftHand SANs optimize and protect your storage while vSphere optimizes and protects your computing resources Together LeftHand and vSphere clusters provide a complete multi site failover and failback solution LeftHand Multi Site SANs protect replicated data between sites e Clusters are geographically aware and use that to optimize the data path e Clusters can traverse multiple subnets e 4 way replication can keep data redundant even after a complete site failure e LeftHand Failover Manager fully automates failover operations vSphere can be used to protect virtual machines across sites e One management interface for server configuration and clustering e Makes vMotion DRS and HA possible between two sites e Failover and failback can be automated and controlled by policies Application Managed Snapshots in vSphere HP LeftHand operating system v9 5 expands Application Integration functionality to include vSphere integrated snapshots This new Application Aware Snapshot Manager enables vSphere volume SAN based snapshots thereby creating application consistent point in time copies for reliable recovery of VM states Without this integration option snapshots previously were a volume of the LeftHand SAN state In flight cached data may not have been fully quiesced creating an application recovery point Now full
22. Technical white paper Best practices for deploying VMware vSphere 5 with VMware High Availability and Fault Tolerance on HP LeftHand Multi Site SAN cluster Table of contents Executive summary 2 Deploying VMware High Availability and VMware Fault Tolerance on HP LeftHand Multi Site SAN clusters 2 Overview 2 Solution components 3 Hardware and software requirements 5 Deployment best practices 7 Deployment strategy 7 Configuring the HP LeftHand Multi Site SAN 8 Solution tests 18 Test hardware software configuration 18 Test storage layout 19 Implementing a proof of concept 22 Summary 23 Appendix A Additional features of HP LeftHand with vSphere 5 23 Application Managed Snapshots in vSphere 23 Snapshots of Raw Devices 23 Snapshots on VMEFS datastores 24 Remote Copy volumes and SRM 24 SmartClone volumes 24 vSphere Metro Storage Cluster vMSC using HP LeftHand Multi Site 24 For more information 25 Executive summary Virtually every business organization understands that successful operations depend on the continuous availability of its applications Most companies rely on internal applications ranging from enterprise resource planning to payroll systems to keep the wheels of their enterprise turning They also depend on external facing applications for everything from selling products to automating the supply chain with suppliers and partners The failure of any of these business critical applications could be catast
23. are As a best practice with LeftHand storage nodes a minimum of two 1GbE NICs or preferably 10GbE NICs are recommended Since the HP LeftHand storage nodes come with two built in network ports on the system board adding two more NICs to the PCI bus can help with networking redundancy HP best practices recommend implementing rapid spanning tree and deploying layer 2 switching For detailed information see the document HP P4000 SAN networking recommendations at http h20195 www2 hp com V2 GetDocument aspx docname 4AA2 5615ENW LeftHand storage nodes The LeftHand software provides the ability to create and manage multi site clusters The Centralized Management Console CMC is the mechanism to manage one or more HP LeftHand Multi Site SANs from a single console LeftHand storage nodes use virtual IP addresses An additional IP address can also be assigned to each network interface on the node In the case of Adaptive Load Balancing ALB each node has a unique IP address regardless of bonding type and an additional Virtual IP VIP address per cluster HP recommends using a single IP address with ALB and configuring one Virtual IP address per cluster The virtual IP address will be the target IP address that the ESXi host will use in order to mount the provisioned LUNs A virtual IP address is a highly available address assigned to a storage cluster In the event that a storage node in a cluster becomes unavailable servers can still a
24. ccess volumes through other storage nodes in the cluster using a virtual IP address The VMkernel network for iSCSI should be placed in a separate VLAN than the management and virtual networks used by virtual machines If enough networks are available vMotion and VMware FT should also use separate VLANs Separating networks by functionality for example iSCSI VMotion VMware FT virtual machines provides higher reliability and improved performance of the individual functions Infrastructure When working with VMware HA it is recommended to have at least three hosts to allow for business continuity in the event of a host failure It is also highly recommended that the cluster be architected so no more than four hosts ina given cluster are located in the same server rack or blade enclosure For storage the recommendation is to include at least three sites in the plan where a site may be a rack a room or a Separate building two sites for the storage nodes and a third site for the Failover Manager FOM The FOM is a virtual appliance that manages quorum of the storage nodes in the event of a loss of network connectivity between the two sites Note While it is possible to only use two sites with the built in Virtual Manager this solution is not recommended as it requires a manual process for recovering quorum The Failover Manager is required to achieve transparent failover and it is the recommended solution Microsoft Active Direct
25. e solution Since the storage array generally stores the virtual disks of the VMs even the smallest disruption can negatively affect an environment where availability is critical HP LeftHand Network RAID stripes and mirrors multiple copies of data across a cluster of storage nodes eliminating any single point of failure in the SAN Applications have continuous data availability in the event of a power network disk controller or entire storage node failure HP LeftHand HA storage clusters provide highly available shared storage for VMware vSphere environments In the best practices assessment conducted two P4800s were used Each HP P4800 consists of two storage nodes and one MDS600 disk enclosure The P4800s were spread across two separate racks each rack contained one node from each P4800 and one MDS600 disk enclosure Each storage node was configured with network interfaces NIC teamed RAID level support and also Network RAID for Multi Site SAN cluster configuration The storage nodes operate independently during normal operation hours Each storage node has the ability to simultaneously provide server iSCSI LUNs for VMware Virtual Machine File System VMFS datastores or Raw Device Mappings RDM When a controller fails on one of the nodes data services are transferred from the failing controller to the surviving controller This happens without a disruption in service A controller failure does not adversely affect an ESXi host or the VMs on t
26. ges adapter hardware path changes or device relocation e Distributed File Locking RDMs are available as shared raw LUNs without losing data across multiple vSphere hosts accessing the same LUN 23 e Snapshots VM snapshots are possible on RDM LUNs in virtual compatibility mode Combined with the LeftHand Application Managed snapshots raw devices may be quiesced in context VMs Note that this is NOT supported with physical compatibility mode RDMs Snapshots on VMEFS datastores LeftHand SAN snapshots are very useful in a vSphere 5 environment All virtual machines stored on a single volume can be snap shot and rolled back to at any time Moreover snapshots can be mounted to any vSphere 5 host without interrupting access to their source volume This may be used for data mining or testing real world data without affecting live corporate data Remote Copy volumes and SRM Remote Copy replicates LeftHand snapshots over WAN links to remote sites for disaster recovery or backup VMware vSphere 5 environments can be protected by Remote Copy volumes on a scheduled basis and automated by VMware vCenter Site Recovery Manager SRM for a simple and complete disaster recovery solution HP provides a Storage Replication Adapter SRA for Site Recovery Manager SRM to integrate Remote Copy volumes seamlessly with a vSphere 5 environment For more information on Remote Copy volumes review the Remote Copy User Manual installed with the CMC vSphere
27. hared_Datastore Normal Non SSD 5 OEE Network Type YM Network Standard port group Af Fault Tolerance Fault Tolerance Status Protected Secondary Location aps41 190 Total Secondary CPU 30 MHz Total Secondary Memory 1269 00 MB vLockstep Interval 0 015 seconds Log Bandwidth 33 KBps M Storage Profiles Profiles Compliance M Storage Profiles Refresh 17 When the setup of the cluster is finished there will be an HA fault tolerant cluster configured with the various hosts networks host datastores and shared datastores The example in Figure 13 shows four hosts in a cluster called FT_Site which includes a shared datastore a heartbeat datastore an individual non shared datastore for each of the four hosts along with the HA and DRS settings Figure 13 Example of a completed cluster setup 4P540 28 hy Demo Datacenter TEES E aps41 187 g aps41 188 g aps4 189 g aps41 190 GB ActiveDirectory DNS1 Permissions Maps Profile Compliance FT_Site Getting Started Virtual Machines Hosts DRS Resource Allocation Performance Tasks Events Alarms General Sphere HA vSphere DRS On Admission Control Enabled vSphere H On Current Failover Capacity 3 hosts VMware EVC Mode Intel Nehalem Gen Xeon Configured Failover Capacity 2 hosts Core i7 Host Monitoring Enabled Total CPU Resources 147 GHz s
28. he Initiator Node Name The Initiator Node Name can be obtained from vSphere on the Storage Adapters page in the host Configuration tab As a best practice only select Enable load balancing if the initiator is compliant For a list of compliant initiators please see http h20195 www2 hp com V2 GetDocument aspx docname 4AA2 5654ENW As a best practice assign all servers to sites in the Centralized Management Console This avoids high latency connections by connecting servers only to storage systems in the same site when available After the ESXi hosts have been defined they can be assigned to a site by right clicking Sites and selecting Edit Site 11 ESXi hosts can be grouped into server clusters so that volumes can be assigned to the server clusters not individually After defining servers a user may define and assign volumes using the new volume wizard by right clicking the cluster name and selecting New Volume To use a volume in a VMware HA FT cluster it must be assigned with Read Write permissions to all hosts in the cluster Clicking the Advanced tab allows you to select a Network RAID level As a best practice use either Network RAID 10 or Network RAID 10 2 for a 2 site SAN and RAID 10 1 for a 3 site SAN Do not use Network RAID 0 Network RAID 5 or Network RAID 6 as they will not tolerate site failures Network RAID 10 2 offers an even higher level of data protection that enables the Multi Site SAN to sustain a complete failure at one site a
29. he host Network RAID provides additional resiliency for the LeftHand SAN by adding synchronous replication on a volume by volume basis This protection is provided in addition to RAID at disk level thereby allowing for better overall protection Network RAID is synchronous replication or mirroring of data for a volume across all storage nodes in the cluster Network RAID 10 or higher is required to replicate data in a Multi Site SAN environment In order to prevent impacting disk 1 0 to the ESXi hosts the round trip latency between the sites must be no more than 2 milliseconds Multi Site SAN clusters can maintain a low latency across long distances without compromising data For high availability there must be protection against failures of components such as servers storage arrays or nodes and networks Protection against site wide failures such as electrical and cooling is also necessary VMware vSphere 5 does this by creating virtual machines that are set up to failover to other servers or another site in the event of a site wide failure The HP LeftHand Multi Site SAN cluster does the same for storage With the HP Centralized Management Console CMC create a multi site cluster of storage nodes spread out across two locations that act together as one logical SAN but can accommodate a component failure or site wide failure and keep the data volumes up and available If a site failure occurs the VMs containing the applications are restarted on
30. managers on each site For two site clusters run a Failover Manager on a third site which can be located on a physical or logical server that is not included in the cluster For more information on using LeftHand SANs with VMware vSphere 5 please see http h20195 www2 hp com V2 GetDocument aspx docname 4AA3 6918ENW LeftHand SAN allows for quick and easy configuration through the Centralized Management Console With the Getting Started Launchpad a user can quickly discover systems assign volumes to Servers and create items such as management groups clusters and volumes To create a multi site SAN at least two storage systems must be in the Available Systems pool before starting Figure 3 Centralized Management Console system discovery FHP P4000 Centralized Management Console Fie Find Tasks Help Check for Upgrades f at SAN Status Page E Getting Started gt t Configuration Summary EHe Available Systems 5 Welcome to the Getting Started Launch Pad The Getting Started Launch Pad provides tools for common tasks to get you started re You can administer your storage area network in three easy steps This page provides the tools for you to identify your a systems to set up your data stores and data strategy and to restrict or provide access to the data Begin by selecting 4 Find Sveteme helow Remember vau ean ahwave come hack here when new Components are added to your network f P48
31. med in January 2012 Deploying VMware High Availability and VMware Fault Tolerance on HP LeftHand Multi Site SAN clusters Overview The HP and VMware solution described in this white paper discusses the usage of HP LeftHand Multi Site SAN clusters with VMware Fault Tolerance FT and VMware High Availability HA The solution gives the ability to stretch a VMware vSphere cluster across two sites and to provide full fault tolerance and built in zero downtime between the two sites HP customers have been using the HP LeftHand Multi Site SAN solution for years with VMware but typically only with VMware HA The solution in this paper extends the high availability offered by VMware HA by using a feature called VMware Fault Tolerance which protects mission critical virtual machines VMs from downtime and data loss VMware HA provides the ability to migrate VMs to another host and restart the VMs VMware FT allows two virtual machines to run simultaneously in lockstep through the use of VMware vLockstep technology When the primary VM fails the secondary VM immediately takes over and becomes the primary VM picking up where the failed VM left off A new secondary VM will then spawn on the next available ESXi host Figure 1 shows an example of VMware FT transparent failover While VMware FT and HA provide tremendous resiliency for VMs if the underlying storage fails it can take down the environment HP LeftHand Multi Site SAN clusters extend the protecti
32. n the management groups as a real manager in the system for quorum operations only When using a FOM configure an equal number of LeftHand managers on both sides of the Multi Site SAN and adda FOM at the third location If an event disables any of the three locations a quorum will still exist with the managers running at the two remaining locations and automatic failover will occur allowing the volumes to stay online As a best practice in a cluster that spans multiple sites run the same number of managers in each site In a cluster that Spans two sites run a FOM in a third site either physical or logical For more information please refer to the HP P4000 Multi Site HA DR Solution Pack User Guide at http bizsupport2 austin hp com bc docs support SupportManual c03041871 c03041871 pdf Deployment best practices Deployment strategy VMware HA designates hosts for particular roles and responds to problems as they arise Ina VMware HA deployment a set of ESXi hosts is combined into a cluster with a shared pool of resources Hosts in the cluster are monitored and if one of the hosts fails virtual machines that were running on the failed host are restarted on a different host In addition VMs are monitored to detect failures and they are restarted on the same host if this occurs VMware FT is designed to handle host failures with no downtime for VMs and no loss of data or network connections It works by creating an identical copy of a
33. nd Windows are U S registered trademarks of Microsoft Corporation 4AA0 4385ENW Created March 2010 Updated May 2012 Rev 6
34. nd still maintain replicated data at the other site With Network RAID 10 2 there are four copies of data for the volume on the SAN Figure 7 Edit Volume Basic Advanced Volume Name Cluster Cluster 0 Description Data Protection Level Network RAID 10 2 Way Mirror Cluster Available Space 35 01394653 TB Type Primary Remote Reported Size Provisioning Full Thin Choose servers to assign to volume VM_Shared_Datastore_1 Assigned Permission Read rite Read rite Read rite Read vrite 12 VMware Fault Tolerance VMware has released a white paper on VMware FT and VMware HA best practices One of the best practices is to verify that the HP ProLiant servers are on the VMware hardware compatibility list HCL for VMware Fault Tolerance http www vmware com resources compatibility search php deviceCategory server Several HP ProLiant servers are on the list and HP will keep updating the list as more servers become certified At the time of writing this white paper HP has several HP ProLiant servers such as the DL and BL server models on the list VMware vSphere 5 0 requires at least two FT compliant servers using the same FT version or host build number For more information about configuring a VMware FT cluster please see the Fault Tolerance Checklist in the vSphere 5 Availability Guide available at http pubs vmware com vsphere 50 topic com vmware Cbase PDF vsphere esxi vcenter server 50 availability
35. nly used in an environment where no FOM is available These types of environments are not recommended Figure 5 shows the creation of a site in the environment Figure 5 LeftHand cluster Site A Fe HP P4000 Centralized Management Console O x Fie v Find Tasks Help 4 Upgrades Available fait SAN Status F Management Groups Clusters and Yolumes Wizard d Getting Star gt t configure Set up Sites to be Used in Your Multi Site Cluster FH Available S Select or create sites to be used in the cluster j to identify your n by selecting your network The systems in the table below mz ee ee ee ee ee eee ee ee ee ee ee ee es Se ae T m New Site Ea P4800sb7 U onsole in order P4800sb7b Site Name Site_A P4800sb8a Site D ipti P4800sb8b primary _site FOM Make this site primary etwork and Site Systems Name IP Address Model Manager S P4800sb7a 192168 2 5 P4460sb G2 Ag radia To create sites click New To mc ff p4800sb8a 192 168 2 6 P4460sb G2 l New To continue click Next ponent Ch stem 4 Back EE Co e Alarm Tasks The environment in this example has a total of three sites with the third site containing only the Failover Manager On the next screen you are prompted to create one or more Virtual IPs VIPs for the cluster All clients that will use the SAN will communicate to the cluster using the VIP configured By using a VIP
36. o data loss in the event of failures of servers and or storage Complementing VMware vSphere 5 with LeftHand SANs addresses the storage demands and cost pressure associated with server virtualization data growth and business continuity HP LeftHand SANs scale capacity and performance linearly without incurring downtime enabling it to support small customers to mid sized enterprises requiring mission critical applications Note HP LeftHand was previously referred to as HP P4000 LeftHand operating system was previously referred to as SAN iQ Target audience The white paper is intended for solution architects engineers and project managers involved with the deployment of HP LeftHand Multi Site SAN clusters with VMware vSphere 5 virtualization solutions It provides recommendations and should not be regarded as a stand alone reference The audience should be familiar with the HP LeftHand Multi Site SAN architecture and virtualized infrastructures It is expected that the reader will also have a technical background with VMware vSphere 5 especially in VMware High Availability HA and VMware Fault Tolerance FT In addition the reader should understand the architecture of VMware vSphere 5 and how this product is able to virtualize hardware resources as outlined in various in depth white papers available at http www vmware com support pubs vsphere esxi vcenter server pubs html This white paper describes characterization work perfor
37. of concept purposes but it is not recommended for production networks due to latency requirements 14 It should also be noted that the Multi Site SAN was attached using two additional vmhba ports to further isolate and guarantee sufficient bandwidth for the SAN s datastore and its required heartbeat datastore Below is a screenshot of the datastores Note that the datastore1 3 is on a local HP Serial Attached SCSI device of the ESXi host and is used as its own local datastore It is not used by any of the virtual machines and would not be available for a failover since it is not on a shared device Datastore Heartbeat_1 was created on the SAN and used by the VMware HA cluster as the heartbeat storage The third datastore Shared Datastore is used by the various virtual machines for virtual disks that are always to be available to the cluster Figure 10 Example of the datastores in use Mware ESXi 5 0 0 469512 Getting Started Summary Virtual Machines EEA Configuration Tasks amp Events Alarms Permissions Maps Storage Views Hardware Status Hardware iew Datastores Devices Processors Datastores Refresh Delete Add Storage Rescan All Memory Identification Status Device Drive Type Capacity Free Type Last Update Alarm Actions Storage I O Control Hardware Accelerati amp datastore1 3 Normal HP Serial Attached Non SSD 274 25GB 273 30 GB YMFSS 1 11 2012 11 19 10
38. on down to storage just as the two VMware features protect the hosts and the virtual machines By combining the three capabilities VMware FT VMware HA and HP LeftHand Multi Site SAN cluster an entire site in one location is protected by a secondary site and vice versa Figure 1 VMware FT Transparent Failover New Secondary VM created Primary VM Secondary VM Client wie goes down ES becomes mary HP developed a test plan to demonstrate VMware FT and VMware HA functioning in a multi site environment with the Multi Site SAN The objective is to demonstrate business continuity with zero downtime in the event of a failure that takes one site offline A multi site single cluster configuration can be split across 2 logical or physical sites One site can be a single server storage configuration in a rack a server storage configuration in a separate room or an entire physical site Solution components HP LeftHand Multi Site SAN cluster with VMware HA clusters VMware ESXi hosts can be placed into groups called clusters A cluster contains 2 to 32 hosts that work together to enable features such as VMware HA and VMware Distributed Resource Scheduler DRS All the VMs in an HA cluster are protected by VMware HA which detects failures of ESXi hosts When a host fails VMware HA moves virtual machines from a failed host and restarts them on another unaffected host within the cluster Shared storage is an important component of th
39. ory AD and Domain Name Server DNS were used in the project to validate the solution Licensing In order to deploy the solution the appropriate VMware vSphere licenses will need to be acquired At the time of this writing VMware FT VMware HA and vMotion are part of the VMware vSphere Advanced Enterprise and Enterprise Plus licensing For VMware HA only implementations this can be done with VMware Essentials Plus Standard Advanced Enterprise or Enterprise Plus since they all include VMware HA functionality However for the complete solution licenses for VMware FT VMware HA and vMotion are required VMware HA VMware HA is a feature that monitors the health of ESXi hosts in a VMware HA cluster In the event of a host failure due to either planned or unplanned downtime the virtual machines that were running on the host are moved to another host in the VMware HA cluster and restarted With vSphere 5 0 all dependencies on DNS for HA have been removed However it is always a best practice to ensure all hosts can be resolved properly through DNS VMware Fault Tolerance VMware FT also has requirements that must be addressed A few examples include CPU network latency and shared storage requirements For complete details please refer to the Best Practices for Fault Tolerance section of the vSphere Availability documentation located at http pubs vmware com vsphere 50 topic com vmware Cbase PDF vsphere esxi vcenter server 5
40. perating systems that support VMware Fault Tolerance http kb vmware com kb 1008027 25 VMware vSphere 5 guides http www vmware com support pubs vsphere esxi vcenter server pubs html VMware Fault Tolerance Configuration Recommendations http pubs vmware com vsphere 50 index jsp topic com vmware vsphere avail doc_50 GUID ED61649B B8A6 4758 AF34 6C7A1CDF147B html Video Demonstrate Zero Downtime with HP Storage and VMware Advantage http h18000 www1 hp com products servers vmware webinar html VMware Fault Tolerance migration transition states KB article 1010634 http kb vmware com kb 1010634 vSphere High Availability Deployment Best Practices http www vmware com files pdf techpaper vmw vsphere high availability pdf To help us improve our documents please provide feedback at http h71019 www7 hp com ActiveAnswers us en solutions technical_tools_feedback html Get connected hp com go getconnected Current HP driver support and security alerts delivered directly to your desktop Copyright 2010 2012 Hewlett Packard Development Company L P The information contained herein is subject to change without notice The only warranties for HP products and services are set forth in the express warranty statements accompanying such products and services Nothing herein should be construed as constituting an additional warranty HP shall not be liable for technical or editorial errors or omissions contained herein Microsoft a
41. re components were used for this project e Four HP ProLiant BL460c G7 blade servers vSphere cluster with VMware HA and VMware FT logging e Two HP ProLiant BL460c G7 blade servers as physical hosts for infrastructure e Four HP Virtual Connect Ethernet Modules for c Class BladeSystem 18 Two HP 8212 zl Switches with four 10GbE SR SFP transceivers Four HP LeftHand P4800 BladeSystem storage nodes or equivalent LeftHand Multi Site SAN bundled solution package Software The following software components were used in the project VMware ESXi 5 0 VMware vCenter Server 5 0 Primary location on Site B VMware vSphere Client 5 0 Microsoft Windows Server 2008 R2 Enterprise Edition HP LeftHand operating system v9 5 00 HP LeftHand Centralized Management Console CMC v9 5 00 Figure 14 Hardware utilized in test configuration HP network switch Sphere vSphere X FTVM FTVM Primary Secondary l c7000 with 3 BL460c G7 c7000 with 3 BL460c G7 and 2 HP P4800 storage and 2 HP P4800 storage blades blades HP LeftHand Multi Site Cluster vSphere 5 HP P4800 HP P4800 storage storage Cluster Test storage layout The Multi Site SAN products used in the solution were storage nodes with 10Gb Ethernet HP followed the configuration best practices for the storage nodes as well as VMware s best practices for deploying iSCSI shared storage The storage layout consists of having the storage nodes
42. rophic to a company The causes of downtime are more numerous than most people think Top of mind causes often include fire and natural disasters such as floods tornadoes and hurricanes There are also unexpected causes including power and cooling system failures Just as serious are the failures that don t count as a true disaster A fire in another part of the building can cause smoke and water damage in the most well protected datacenters Then there are human errors that affect the datacenter which can range from the wrong circuit breaker being thrown to an accidental discharge of fire suppression material In fact 70 of all unplanned downtime is caused by human error When a failure occurs its affect can be as narrow as asingle system or as wide as aroom floor building or entire site A comprehensive business continuity solution should include the capacity to recover operations after a major disaster as well as the capacity to remain online during minor disruptions VMware vSphere 5 has brought great relief to IT organizations searching for high availability HA fault tolerance FT and disaster recovery DR solutions by reducing implementation cost and complexity This best practices white paper describes a cost effective simple to deploy easily managed highly available fault tolerant solution that can be put in place by using an HP LeftHand Multi Site SAN cluster with VMware vSphere 5 This solution provides zero downtime and zer
43. rtification All supported storage devices are listed in the VMware Storage Compatibility Guide HP LeftHand Multi Site HP LeftHand Multi Site is a feature of the LeftHand operating system This technology allows for storage clusters to be stretched across sites to provide high availability beyond failure domains defined by the administrator Traditionally in Metro Storage Cluster configurations these failure domains are distinct geographic locations but the technology can be 24 used to protect against the failure of a logical site that may be a rack room or floor in the same building as well as buildings within a campus or datacenters that are separated by as much as 100km or more The Multi Site feature is part of the all inclusive LeftHand feature set and does not require any special licensing The HP LeftHand Multi Site solution uses LeftHand OS Network RAID technology to stripe two copies of data across a storage cluster When deployed in a Multi Site configuration LeftHand OS ensures that a full copy of the data resides on site or each side of the cluster In Multi Site vMSC configurations data remains available in the event of a site failure or loss of link between sites A VMware HA DRS cluster is created across the two sites using ESXi 5 0 hosts and managed by vCenter Server 5 0 The vSphere Management vMotion and virtual machine networks are connected using a redundant network between the two sites It is assumed that the vCenter
44. s Resource Allocation Performance eGift Tasks amp Events Alarms Permissions Maps Storage Views Hardware Status Hardware View Datastores Devices Processors Datastores Refresh Delete Add Storage Rescan All Memory Identification 2 Status Device Drive Type Capacity Free Type Last Update Alarm Actions Storage I O Contral Hardware Accelerati Storage fA datastoret Aa Normal HP Serial Attached MNan 55D 274 25GR 273 30 GR VMFSS 11 2012 10 42 41 AM Enabled Disabled Unknown Networking 11 2012 10 42 57 AM Enabled Disabled Supported Storage Adapters Choose the Destination Resource Pool Network Adapters Choose where to place this host s virtual machines in the resource pool hierarchy Advanced Settings Power Management Software Choose Resource Pool Virtual Machine Resources Ready to Complete Licensed Features What would you like to do with the virtual machines and resource pools For this Time Configuration host ONS and Routing Put all of this host s virtual machines in the cluster s root resource pool Authentication Services Resource pools currently present on the host will be deleted Power Management Virtual Machine Startup Shutdown Create a new resource pool For this host s virtual machines and resource pools This preserves the host s current resource pool hierarchy Virtual Machine Swapfile Location Properties Security Profile Host
45. stretched across two sites These units were kept within the same lab but configured as separate sites for ease of test management and execution Both storage nodes had their NICs teamed for ALB ALB was used because it is the only supported active active network configuration on the P4800 Network traffic was isolated onto a separate network by creating VLANs to the storage array nodes Both the CMC console and FOM were configured to have access to both VLANs This was required for HP engineers to manage the environment HP performed several tests to validate the configuration The main tests conducted for this white paper are the loss of ESXi hosts at the primary site the loss of the storage on either site and the loss of an entire site An Outlook client was installed to verify that email could be sent and received at the start and end of each test scenario 19 The first test consisted of manually powering off ESXi hosts on Blade 1 and Blade 2 on site A with a hard shutdown to simulate power loss The Primary VMs enabled with VMware FT were running on Blade 1 AD DNS and Blade 2 MS Exchange which were the targeted servers to fail The Secondary VMs were running on Blade 3 AD DNS and Blade 4 MS Exchange on site B Both the Primary and Secondary VMs had access to the same LUN which was protected by the Multi Site SAN cluster Figure 15 Failover tests ESXi host fails E HP P4800 storage c7000 with 3 BL460c
46. ummary Virtual Machines Performance Riemer Tasks amp Events Alarms Permissions Maps Storage Views Hardware Status Hardware iew vSphere Standard Switch vSphere Distributed Switch Processors Networking Refresh Add Networking Properties Memory Storage Standard Switch vSwitchO Remove Properties gt Networking Virtual Machine Port Group Physical Adapters A irtu P y P uariis A vM Network Q B vmnict 4000 Full 53 Network Adapters a f 2 virtual machines EB vmnico 4000 Full SA Advanced Settings Exchange1 rab re ew ActiveDirectory DNS1 fab Software Mkemel Port 5 Management Network Q Licensed Features vmk0 10 136 41 187 Time Configuration DNS and Routing Authentication Services Power Management Virtual Machine Startup Shutdown Virtual Machine Swapfile Location Security Profile Host Cache Configuration System Resource Allocation Agent M Settings Advanced Settings Standard Switch vSwitch1 VMkemel Port J vMotion vmk1 192 168 1 20 Standard Switch vSwitch2 VMkemel Port D FT Logging vmk2 192 168 3 20 Q Remove Properties Physical Adapters e R vmnic3 1000 Full L e RB vmnic2 1000 Full a a Remove Properties y ia Physical Adapters o B vmnic5 1000 Full L e R vmnic4 1000 Full vMotion and VMware FT logging can occur simultaneously on the same network for testing or proof
47. virtual machine One copy of the virtual machine called the Primary VM is in the active state receiving requests serving information and running applications Another copy called the Secondary VM receives the same input that is received by the Primary VM but does not perform any disk writes or network outputs Any action performed on the Primary VM is recorded and replayed on the Secondary VM ensuring the two virtual machines stay in lockstep synchronization In this deployment strategy VMware FT provides preservation of run time state and the elimination of application downtime There are other deployment considerations to keep in mind Certain infrastructure applications such as Active Directory and DNS are important and can impact deployment if not taken into consideration For example if there is a Microsoft Exchange Server VM deployed it needs to have access to the AD server for authentication reasons If the Exchange server were to fail its secondary VM will become the primary VM but it will still need to access the AD server to authenticate the users If the network is totally lost the solution will not work Configuring the HP LeftHand Multi Site SAN HP LeftHand Multi Site SAN The Multi Site SAN feature enables the synchronous mirroring of data between sites automatically by configuring equal numbers of storage nodes in each site For a cluster that spans multiple sites the best practice is to run the same number of
Download Pdf Manuals
Related Search