Solved

Windows Failover File level backup issue

  • 20 April 2022
  • 9 replies
  • 581 views

Badge +6

Dear Team,

We have configured two node Windows Failover File level backup. There are 8 Cluster disks and we want file level backup in all the cluster disks.

We installed File system agent in both nodes and then the backup job was configured under the Failober cluster client.

Now we are facing the issue that the cluster disk is showing only Cluster Owner node only, if the cluster disk will move to another node that disk is not showing under the Windows cluster client. The backup job is showing failed due to disk path is not available.

Please advise.
 

icon

Best answer by Jos Meijer 21 April 2022, 20:27

View original

9 replies

Userlevel 7
Badge +16

Hi @Anand K 

Have you positioned the job results directory on a shared cluster drive which is attached as a resource to the virtual instance?

Badge +6

@Jos Meijer, Yes, the Job results directory is configured one of the cluster resource disk.

Userlevel 7
Badge +16

Just to be sure, the cluster resource disks are all online in Fail Over Cluster Manager after the fail over?

Badge +6

Hi,

 

Yes, the cluster disks are online. 

 

We observed all the cluster disks in one node then we will see the disks, but the disks are distributed in two nodes then only see the active node cluster disks.

Userlevel 7
Badge +16

The client might not be updated properly after the fail over.
Can you try this option and check again?

  1. From the CommCell Browser, go to the Client Computers node, right-click the Cluster_Group_Client and then click Properties.

    The Client Computer Properties dialog box appears.

  2. Click Advanced.

    The Advanced Client Properties dialog box appears.

  3. On the Cluster Group Configuration tab, click the Force Sync configuration on remote nodes check box and then click OK.

Userlevel 7
Badge +16

After further inspection the previous post might won’t work.
When looking at the requirements for server packages on windows fail over clusters I found this text:

Quote: “The owner node of the cluster disks must be the active physical node”

This indicates that the cluster disk resource must be active on the node running the virtual instance, not the passive node. This might be related to the cluster plugin running on the active node as a resource and therefore might have problems when resources reside active on another node.


Still try the previous post, it might be the nudge which is needed to update the cluster config.
Will keep looking, but currently only these two findings.

Let me know the outcome please, if not working I can setup a windows failover cluster in a lab to validate the function.

Userlevel 7
Badge +23

@Anand K , following up to see if you have had any success on your side in this issue.

Thanks!

Userlevel 7
Badge +23

@Anand K , hope all is well!

Following up to see if you have had any success on your side in this issue.

Thanks!

Badge +3

Hello,

do you have a solution for that configuration?

We run in to the same issue, active and passive node have own volumes and only the volume on the active node is accessible by the cluster client.

“Force Sync configuration on remote nodes” & restart services didn’t work.

Thanks

Reply