I have an old SQL 2008 (non-R2) two node cluster that may or may not have an issue. The two nodes are VMware virtual machines with a quorum raw disk in the standard way:
- All cluster disks including the quorum are assumed by Node2 during a complete service failure, like a server down/reboot.
- A manually triggered failover of all clusters and services from Node1 to Node2 results in Node to assuming all shared disks except the quorum. The quorum is still held by Node1. No errors are reported in Windows Event Logs.
If Node2 can assume the quorum during a reboot, then it doesn't appear to be permissions related. Where else should I look at?