Reboot ESX with MSCS Cluster nodes

[Gael]

Hi,

I've an ESX Cluster which need to be restarded. This cluster host a MSCS Cluster (without VMWARE DRS affinity recommendations...).

What's best practice ?

- Keep maintenance mode running vmotion as other VMs or,

- Define DRS affinity asap and shutdown MSCS Cluster node when required (only guest shutdown or others actions to do before?) or,

- Another ideas?

Thanks,

[Troy_Clavell]

assuming your are clustering across nodes, which would be physical compatibility then you won't be able to vMotion.  What I would do is failover your cluster, reboot the ESX host in question, then if needed failover the cluster again and reboot the other ESX Host.

[weinstein5 · Accepted solution]

You should not be able to vmotion a node of an  MSCS cluster so DRS wikll not work - strting with ESX host that has the standby node place it into Maintenance mode and manually sut node the MSCS node - restart the ESXi host bring it out of maintenance node - once it is back up and the MSCS lcuster is stable -  failover the primary MSCS node and shute down the VM - place that ESX host in Maintanance mode allowinf DRS to evacuate the running VMs - restart the ESX host bring the MSCS VM back up and move the workload back to this MSCS node -