Maintenance work on the server

Disconnect the Direct Storage Spaces server for maintenance work

This topic provides instructions for restarting or shutting down servers correctly with Direct Storage Spaces .

If you use Direct Storage Spaces to disable a server (turn it off), keep in mind that parts of the storage that have been shared on all servers in the cluster will also be disabled. For this, you must pause (suspend) the server you want to disconnect, move the roles to other servers in the cluster and verify that all data is available on the other servers in the cluster; Thanks to this, the data will remain secure and you can access it during the maintenance task.

Follow the steps below to successfully pause a server in a Direct Storage Spaces cluster before deactivating it.

Check that it is safe to disconnect the server

Before disconnecting the server to perform maintenance work, check that all volumes are in good condition server maintenance job.

To do this, open a PowerShell session with administrator permissions and run the following command to see the status of the volumes:Verify that the HealthStatus property of all volume (virtual disk) is Correct .

To do this in the Failover Cluster Administrator, go to Storage > Disks .

Check that the Status column of each volume (virtual disk) is displayed Online .

Pause and purge the server

Before restarting or shutting down the server, pause and purge (remove) any type of role, such as virtual machines running on the server. Thanks to this, Direct Storage Spaces will be able to empty and confirm the data correctly, in order to ensure that the shutdown is carried out without mishaps taking into account the applications running on that server.

In PowerShell, run the following cmdlet (as administrator) to pause and purge.

To do this in the Failover Cluster Administrator, go to Nodes , right-click on the node, and then select Pause > Purge Roles

Turn off the server

Once the server is purged, it will show Paused in PowerShell and in Failover Cluster Administrator.You can now restart it or shut it down safely, just as you normally would (for example, using the Restart-Computer or Stop-Computer PowerShell cmdlets).

The incomplete or degraded operating state is normal when the nodes are closing or starting or stopping the cluster service on a node and should not be a problem. All volumes remain online and are accessible.