4

Which do you do first to avoid service interruption? Or does it not matter?

Is it:

Suspend-ClusterNode -Drain
Get-ClusterNode $env:COMPUTERNAME | Get-ClusterGroup | Move-ClusterGroup

or:

Get-ClusterNode $env:COMPUTERNAME | Get-ClusterGroup | Move-ClusterGroup
Suspend-ClusterNode -Drain

The microsoft doc for Move-ClusterGroup says:

"Moving a resource group is a way of simulating failover. It is also an appropriate step to take in preparation for routine maintenance on a node."

Which makes me think you move before suspending. So which cmdlet do you issue first?

Colyn1337
  • 2,397
  • 2
  • 23
  • 40

1 Answers1

3

According to Microsoft the recommended sequence is to move then suspend https://technet.microsoft.com/en-us/library/ee461051.aspx

Pausing (suspending) a node is usually done when applying software updates to the node, where the recommended sequence is to move all services and applications off of the node, pause the node, then apply software updates to the node. If you need to perform extensive diagnosis or maintenance on a cluster node, it might be more workable to stop (not pause) the Cluster service on that node.

Mass Nerder
  • 1,007
  • 5
  • 6
  • 1
    The Suspend-ClusterNode, with the -Drain stwich, moves services and applications off the node. The Move action is moving cluster resources (a separate thing from the applications and services). Perhaps this is evidence that the moving of resources ahead of any action is academic. – Colyn1337 Oct 27 '15 at 20:34