Dear Community Members,
I plan to installed an additional HBA card to the ESX node in the cluster. Can you please verify the steps below.
1. Take the ESXi in to maintenance mode,
2. Shut down the ESXi
At this stage it is assumed that the VMs will automatically vmotion to the other ESX node.
3. Install the HBA card.
4. Boot up the ESXi
5. VMs fail back to the ESXi node as it was earlier.
Any help would be greatly appreciated.
Regards
Rahul
Almost correct, however:
1. Take the ESXi in to maintenance mode,
At this stage wait for the VMs to be vMotion'd to the other ESX node.
2. Shut down the ESXi
3. Install the HBA card.
4. Boot up the ESXi
5. Verify that the new HBA is recognized properly and perform the storage tasks to present the LUNs to the newly installed HBA's WWPN if required.
6. Exit Maintenance mode
Whether any of the VMs will be migrated back to the host by DRS depends on the resource usage and existing DRS rules. DRS does not do failover/failback, but moves VMs only if required. If you want to have the same VMs back on the host, you may need to initiate the migration manually.
André
Almost correct, however:
1. Take the ESXi in to maintenance mode,
At this stage wait for the VMs to be vMotion'd to the other ESX node.
2. Shut down the ESXi
3. Install the HBA card.
4. Boot up the ESXi
5. Verify that the new HBA is recognized properly and perform the storage tasks to present the LUNs to the newly installed HBA's WWPN if required.
6. Exit Maintenance mode
Whether any of the VMs will be migrated back to the host by DRS depends on the resource usage and existing DRS rules. DRS does not do failover/failback, but moves VMs only if required. If you want to have the same VMs back on the host, you may need to initiate the migration manually.
André
Thanks a lot André
How best to determine the resource usage on the ESX node is optimum so that the VMs can automatically get vmotioned to the other node?
Also can you please point me how to check the DRS rules. I am afraid, I will have to do manual move of the VMs back.
-Rahul
Hi Rahul,
Below blog explain you in detail how DRS takes migration decision:
DRS Deepdive - Yellow Bricks : It is worth to read twice.
If the earlier VM is migrated to other host and if it is not coming back to original host, it is mean that as per DRS load balancing algorithm CPU/Memory load is balanced and you need not to worry.
However, if you want to keep that VM always on original host, you need to configure DRS affinity rule.
Refer:http://blog.pluralsight.com/vmware-storage-drs-rules
For above requirement, you need to lokk for VM-Host affinity rule.
Let me know if you have any other doubt.