Stretch Failover Clustering Failover

Question

Good afternoon, we have a stretch cluster in testing. We have 2 vm servers in one site "DC1" and 2 vm servers in another site "DC2" (Running under Windows Server 2022". DC1 servers share 2 drives "data and log drive" sharing a scsi bus. DC2 also has scsi bus shared drives. They are under the same cluster and file share name. We use Storage Replica for replication between DC1 and DC2. When testing a failover in the case of DC1 being in a "dead" state I found I have to manually break replication so DC2 servers can be set into the primary role.

When I tried bringing the DC1 servers back into the mix. I found the drives were not coming online and not letting me re-establish the Storage Replica. Upon discovery the drives on DC1 were set to RAW when they were NTFS. I've tried using convert and diskpart to clean the partition to format it back to NTFS but it fails. CHKDSK and SFC find no issues with the drive.

One thing I do see is on one DC1 server one of the drives that is listed as RAW is listing it as NTFS on the other server sharing the drive.

Could this all be from the failover testing corrupting the data and log drives?

Support link used to create stretch cluster: Stretch cluster replication using shared storage | Microsoft Learn

Answer 1

If diskpart or even disk management will not let me format the disk on the vm what would be the best process?

Answer 2

Hi Lukas,

Hope you're doing well.

I'm not sure, there are many reasons why the disk may have this issue. Maybe you can check event logs for more details.

As far as I know, when this issue occurs, the only solution is to format the disk.

Best Regards