M
Mikey_boy4million
Hi everyone,
I have been scouring the forums for this issue and found many useful things, unfortunately none have resolved my issue.
The situation is there are 4 Server 2016 nodes, configured as a stretched cluster, with 2 nodes at each site (Site A active/Site B passive).
Attached to each pair of servers is a JBOD configured via HBA (multi-I/O).
Storage Replica is used in asynchronous mode to keep the redundant site up to date (SSD logs/HDD data).
There was a power failure on the primary and in order for the destination volume to brought out of read only mode I had to remove the replication. (Set-SRPartnership-NewSourceComputerName...didn't work as it couldn't talk to the primary nodes to flip the direction).
As the remove-srpartnership executed the primary nodes had been powered back on unexpectedly (data centre bypassed the faulty UPS) and the primary disk ended up online at Site A, the service resumed and the VMs are running fine again on Site A.
With replication broken I want to delete the Virtual Disks from the pool at Site B to provision a different layout to improve performance (I appreciate moving the VMs will be manual again and replication needs re-instating after Site A configuration is matched). Everything to do with Site B pool and disks has been removed from the cluster and no longer appear in Get-ClusterResource, yet I am unable to delete the virtual disks.
So far:
Making Site B pool read/write has not helped (although it was read only initally)
Using Clear-ClusterReservation has not worked (says Failed to clear disk reservation 0x80070001, hopefully because they are already cleared)
I can manage the volumes on the virtual disks (initially they were offline by administrator policy), they are read/write and I have deleted the volumes on the virtual disks but it says no:
Powershell is elevated:
Remove-virtualdisk -friendlyname "LOGS"
error received is: Remove-virtualdisk : not supported
Servers have been rebooted, updated etc still no joy.
I suspect it's because there is some sort of lock with the fact the disks were clustered as in Server Manager they still report as being part of the "Clustered Windows Storage" subsystem.
I also tried adding the pool and disks back to the cluster but that now says no disks are available for clustering.
I can't just kill the pool either as it says I need to remove the virtual disks first..
How on earth can I force remove these virtual disks?
There is no data required on any of the pool or the disks with in it, so even a kill switch for that will do.
Thanks for taking the time to read this if you made it all the way here
Continue reading...
I have been scouring the forums for this issue and found many useful things, unfortunately none have resolved my issue.
The situation is there are 4 Server 2016 nodes, configured as a stretched cluster, with 2 nodes at each site (Site A active/Site B passive).
Attached to each pair of servers is a JBOD configured via HBA (multi-I/O).
Storage Replica is used in asynchronous mode to keep the redundant site up to date (SSD logs/HDD data).
There was a power failure on the primary and in order for the destination volume to brought out of read only mode I had to remove the replication. (Set-SRPartnership-NewSourceComputerName...didn't work as it couldn't talk to the primary nodes to flip the direction).
As the remove-srpartnership executed the primary nodes had been powered back on unexpectedly (data centre bypassed the faulty UPS) and the primary disk ended up online at Site A, the service resumed and the VMs are running fine again on Site A.
With replication broken I want to delete the Virtual Disks from the pool at Site B to provision a different layout to improve performance (I appreciate moving the VMs will be manual again and replication needs re-instating after Site A configuration is matched). Everything to do with Site B pool and disks has been removed from the cluster and no longer appear in Get-ClusterResource, yet I am unable to delete the virtual disks.
So far:
Making Site B pool read/write has not helped (although it was read only initally)
Using Clear-ClusterReservation has not worked (says Failed to clear disk reservation 0x80070001, hopefully because they are already cleared)
I can manage the volumes on the virtual disks (initially they were offline by administrator policy), they are read/write and I have deleted the volumes on the virtual disks but it says no:
Powershell is elevated:
Remove-virtualdisk -friendlyname "LOGS"
error received is: Remove-virtualdisk : not supported
Servers have been rebooted, updated etc still no joy.
I suspect it's because there is some sort of lock with the fact the disks were clustered as in Server Manager they still report as being part of the "Clustered Windows Storage" subsystem.
I also tried adding the pool and disks back to the cluster but that now says no disks are available for clustering.
I can't just kill the pool either as it says I need to remove the virtual disks first..
How on earth can I force remove these virtual disks?
There is no data required on any of the pool or the disks with in it, so even a kill switch for that will do.
Thanks for taking the time to read this if you made it all the way here
Continue reading...