I have a 2 node cluster with 2 instances on them and everything has been running for a couple months, just fine. I needed to swap the drive letters for the cluster because of SQL mirroring, so I followed the MDSN procedures to change the drive letters. In Cluster Manager I changed the drive letters, then change the sql start up parameters for both instances and click start and now it can't find the drives. Like I said I only swapped the drive letters, nothing else. The error log on the cluster drive is written too, which also seems odd, that it can find the error log file, but not the database or log file and they are on the same cluster drive.
As a test I moved the master db to a local D:, changed the startup parameters and tried to start SQL, in the error log file, it found the master.mdf, but failed with the model database because as it was on the cluster shared drive. So it seems that it is around the cluster drives, from what I can tell anyway.
Thanks for any insight.