What is the recommended process for extending an RDM attached to Windows 2012 R2 CAB failover cluster?

[pgibbins]

What is the process for extending RDMs attached to a 2 guests in a Windows 2012 R2 failover cluster as clustered resource?

The Windows cluster is in a CAB configuation.

The RDMs are used by the Windows cluster as a Clustered Shared Volumes (CSV).

We run on vSphere version 5.5 update 3.

The cluster is used to provide HA instances of SQL Server 2014 SP2 Standard.

Thanks

[TheBobkin]

Hello,

The process varies depending on if they are Physical RDM (prdm) or Virtual RDM (rdm), the steps for both are outlined here:

Expanding the size of a Raw Device Mapping (RDM) (1007021) | VMware KB

Bob

-o- If you found this comment useful please click the 'Helpful' button and/or select as 'Answer' if you consider it so, please ask follow-up questions if you have any -o-

[pgibbins]

Thank you for the reply. They are Physical RDMs.

We read the below in Microsoft Windows Server Failover Clustering on VMware vSphere 5.x: Guidelines for supported configu... ...

Microsoft Windows Server Failover Clustering (WSFC) virtual machines use a shared Small Computer System Interface (SCSI) bus. Any virtual machine using a shared bus cannot make hot changes to virtual machine hardware as this disrupts the heartbeat between the WSFC nodes. These activities are not supported and cause WSFC node failover:

• vMotion migration
• Increasing the size of disks
• Hot adding memory
• Hot adding CPU
• Using snapshots
• Pausing and/or resuming the virtual machine state
• Memory over-commitment leading to virtual swapping or memory ballooning

So if I am reading  this correctly we should not increase the size of the RDMs as it is not supported. But I cannot believe that under any circumstances that this cannot be done. There must be lots of other customers that have similar configurations that need to increase the size of RDMs.

So can we do this if we shutdown both cluster node VMs and then increase the RDM lun sizes and bring the cluster nodes VMs back online?

Or maybe we can just shutdown one cluster node and then increase the size of the RDM luns (as the disks are not shared whilst only one node in the windows cluster is operational) and then bring offline node back online?

The cluster nodes use Para Virtual SCSI Controllers for the RDMs, the controllers are set to Physical SCSI Bus Sharing, each RDM is set to Physical compatibly mode.

[TheBobkin · Accepted solution]

No, it is supported to increase the size of pRDM, just not while the VMs are powered-on:

"cannot make hot changes to virtual machine hardware"

Process would be:

- Shut-down both VMs.

- Expand the mapped LUN

- Rescan Storage and adapters on the ESXi host(s)

- Power on the VMs, recan from within Guest OS is unlikely required

Bob

-o- If you found this comment useful please click the 'Helpful' button and/or select as 'Answer' if you consider it so, please ask follow-up questions if you have any -o-

[pgibbins]

Many thanks, that clarifies it for me.

[Eyadem]

hello,

Please did you tested the procedure of expanding drive in your environnent? it worked well ? I have a similar prod envirment and I want be sure that this work without risks ?

Thanks in advance