We have 2 node failover cluster. When we "tested" failover, all seemingly went well. Service failed across and all drives with the exception of the Quorum failed across (did not manually failover the quorum disk). However, yesterday, our primary node was rebooted and the failover did not work. However, we were able to manually bring back the primary node and its services except for the the GxClusPlunIn instances. The only service on this cluster is SQL Server.
It suggests that we run the Validation Test. However, as I am new to Failover Clustering and did not configure this cluster I want to be sure that running this test will not down the nodes. If so, I will need to schedule the test during off hours.
Also, it was my understanding from reading many articles that in Windows 2008 the Quorum disk does not failover automatically during the "manual failover test" but should a node fail, the Quorum disk "will" failover automatically. It seems in our case the Quorum disk did not failover. Can anyone provide some insight into common causes for this or point me to some documentation that will help me troubleshoot this problem?
Cal Miyatake