This came to mind today when a coworker at another site had an issue with persistent SCSI reservations on SAN volumes (on an IBM DS4000 subsystem).....
Several months ago, I had an issue with a 'single node cluster' (the second node was on the shelf waiting to be installed) running Windows 2003 MSCS. The system had a lovely BSOD in the middle of the workday, and once it came back online, it couldn't successfully activate the cluster disks.
I worked with Microsoft for several hours on the phone, updating device drivers, clearing persistent SCSI reservations on my DS4000, and many other things, only to find the oddest solution:
I upgraded the firmware on the QLogic HBA in the system. This magically fixed the problem, and the cluster node immediately came online. Lesson learned: Make sure all the pieces and parts are kept up-to-date in your environment... you never know when something will cause you trouble that you don't expect.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment