Moving an HPE 3PAR storage system to a new location

During the past days we started to use our new data center in a new building. One of the first things that needed to be moved was our HPE 3PAR storage system. This storage is used only with VMware vCenter using a vSphere Metro Storage Cluster (vMSC) with Peer Persistence.

With the help of HPE 3PAR engineers we were able to carry out a hot moving without any interruption of production for users and servers. This was made possible by the excellent Peer Persistence system made by HPE and the VMware vSphere Metro Storage Cluster certified configuration.

Before trying to move this storage system, I made some search on the Internet without finding information regarding this kind of operation. So I hope that my personal notes can help someone else in the same situation.

⚠️ Please note that’s informational only, you need to be sure about the operations and the method to recover during the process for ALL operations. Getting help from HPE engineers will be the only way to be safe and to ensure that operations will be successful.

Before starting

  1. Identify all cables that need to be moved, you can use stickers and photos to facilitate the migration
    1. Take care of the SAS cables and FC ports used for Peer Persistence
  2. Prepare the new location by cabling and configuring everything that can be performing before the move
  3. Check that you’re respecting all the known best practices, especially the ones about multi-pathing
    1. Implementing vSphere Metro Storage Cluster using HPE 3PAR Peer Persistence
  4. Check that you’re not affected by a known HPE Advisor
    1. Give a look especially here about ATS: Array Host Disconnects While Running VMware vSphere 5.5 Update 2 and Later
  5. Check your setup with compatibility matrix:
    1. HPE 3PAR Operating System Upgrade Planning Guide (Not exactly what’s expected, but contains useful information)
    2. VMware Compatibility Guide
  6. Check the multi-path status of your ESXi servers in vSphere
  7. Execute a Health Check on both HPE 3PAR and validate the result
  8. Check that your Peer Persistence volumes are synced
    1. SSMC -> Remote Copy Groups -> Select the copy group -> Volume Pair

Perform the move

  1. On the device that need to be moved execute a switchover command
    1. SSMC -> Remote Copy Groups -> Select the copy group -> Switchover
    2. Wait until the operation is done
  2. Check the stability of your installation with vSphere
    1. Performance monitoring is a good way to check the latency
    2. Rescanning the hosts for HBA a storage path is a good way to know if you’ve lost something
  3. Check that the volumes on the HPE 3PAR which must be moved have no IOPS
    1. SSMC -> Virtual Volumes -> Filter by the unit that must be moved -> Select the volumes one by one -> Performance – > 1 hour
  4. Check that the activity is now handled by the remaining HPE 3PAR
    1. SSMS -> Virtual Volumes -> Filter by the unit that will stay -> Select the replicated volumes one by one -> Performance – > 1 hour
  5. Momentary disable the Quorum Witness
    1. SSMC -> Remote Copy Configurations -> Select the configuration -> Configure Quorum Witness -> Set to off the Start option
  6. Stop the Remote Copy Groups
    1. SSMS -> Remote Copy Groups -> Select affected groups one by one -> Stop
  7. Shutdown the HPE 3PAR that need to be moved
    1. CLI -> shutdownsys halt -> Confirm
    2. Richard T Arnold has a more complete blog post about how to shutdown a 3PAR
  8. Physically shutdown the HPE 3PAR by pressing the power button on the power supply
    1. Expect a complete shutdown, battery cache can keep the unit powered one still a few minutes
  9. Check again that your production is not affected and everything is working fine on vSphere
  10. Physically move all equipments and cables
  11. Check that the SAN fabric is ready and the zoning is corresponding as what you need
    1. If you’ve moved the SAN fabric switches at the same time, check that the ISL link is up and trunk is allowing the correct SAN
  12. Power On the HPE 3PAR
  13. Execute a check health to be sure that the health status has not changed
    1. Remote Copy Groups error will be reported because it was disabled
  14. Check that the Remote Copy Group configuration is up and ready
    1. SSMC ->  Remote Copy Configurations -> Links
  15. Start again the Remote Copy Groups
    1. SSMC -> Remote Copy Groups -> Select affected groups one by one -> Start
  16. Enable the Quorum Witness
    1. SSMC -> Remote Copy Configurations -> Select the configuration -> Configure Quorum Witness -> Set to On the Start option
  17. Wait until all the data is in sync again
    1. SSMC -> Remote Copy Groups -> Select affected groups one by one -> Volume Pairs
    2. Check that all Virtual Volumes are synced
  18. On the device that was moved execute a switchover command again
    1. SSMC -> Remote Copy Groups -> Select the copy group -> Switchover
    2. Wait until the operation is done
  19. Check that the volumes that were moved have again some IOPS
    1. SSMC -> Virtual Volumes -> Filter by the unit that have been moved -> Select the volumes one by one -> Performance – > 1 hour
  20. Check on the vSphere side if everything is running fine

Leave a Reply

Your email address will not be published. Required fields are marked *

*