다음을 통해 공유

Replacing Installation physical drive in CCR

  • 아티클

I was working with one of my customer and he had requirement of replacing the installation drive for CCR cluster as it was running out of space. As it was a physical disk he could not add more space in the drive and he decided to replace existing HDD with new HDD with more space/size.

 

 

Resolution:

==========

You are going to use a third party backup solution to basically clone the d:\. It is expected that this process will retain NTFS permissions and shares and share permissions.

 

1. Failover the Exchange databases and the Cluster owner to Node B (New Active node)

2. Suspend the storagegroupcopy for each of the storagegroups on Node A.

Example:  Suspend-StorageGroupCopy -Identity <NodeA_SERVERNAME>\SG1 -SuspendComment "Changing Drive" 

3. Make note of all of the Microsoft Exchange, and Cluster service states.

4. Stop and then disable all of the microsoft exchange services and the cluster service.

5. Reboot.

6. Using Handle.exe verify that none of the files on the D:\ are currently held in use by any processes.

https://technet.microsoft.com/en-us/sysinternals/bb896655.aspx  

7. After verifying that there are no files on the D:\ that are locked in use. Take a disk clone backup of the D:\ with your third party solution.

8. Turn off machine and add new disk.

9. Turn machine back on and format the new disk and then swap the drive letters so the new disk is now the d:\

10. Reboot for good measure.

11. Restore the backup of the d:\ to the new disk which is now the d:\

12. Verify file shares and permissions still look appropriate. (You can compare with Node B)

13. Restore the cluster and Exchange services to the same startup state they were in in Step 3 above. (Don’t start them)

14. Reboot to allow the service to start appropriately. (Monitor for service start failures here) (Keep in mind that manual services may not be started. This is expected)

15. Attempt to Resume the storagegroupcopy for each of the storagegroups on Node A.

Resume-StorageGroupCopy -Identity <NodeA_SERVERNAME>\SG1

16. If the resume from step 15 fails, attempt a restore-StorageGroupCopy.

17. Verify that the storagegroupcopies are healthy.

Get-StorageGroupCopyStatus -Identity <NodeA_SERVERNAME>\SG1

18. If they are all healthy and have 0 copy q/replay q, go ahead and fail the DBS back over to Node A (New Active node)

 

Repeat the process for Node B as well.

 

 

Thank you,

Mukut-