Hello,
I have the following configuration:
Two-node Windows 2012 R2 WSFC cluster. Nodes realized as VMs in VMware, shared disks realized as shared VMDK's on a shared bus. FileServer and MSSQL roles defined in the cluster and functional before DR tests.
My scenario is as follows:
- all shared disks including quorum are preserved
- one or both nodes (so their C:\ disk(s)) are lost
My restore "engine" is VERITAS NetBackup and its Bare Metal Restore V2V procedure. I always switch off the "damaged" node and I restore its backup (C:\ and System_State:\) under a new VM DisplayName so some manual steps like NIC and shared disks redefinition must be followed after restore.
Now:
- when I presume only one node loss, (so second cluster node is functional), then procedure is ok. After damaged node system restore, and some manual VM reconfigs (NIC's and disks), restored node becomes online ok and rejoins as a valid cluster node. Failovers etc. are working ok.
- when I presume all nodes loss, and I restore first node with the same steps, Cluster Manager console is empty and attempt to Connect to a Cluster fails (giving "RPC server is unavailable" or "Remote server has been paused or is in the process of being started"). Even if all NICs configs existing before disaster were reconstructed and shared disks are visible in Disk Management. Cluster Service is running in restored node.
Any ideas? Probably first restored node does not know where the quorum is and thus some manual step is required here?
Thanks
Michal