Cluster Failover | SQL Server Performance Forums

SQL Server Performance Forum – Threads Archive

Cluster Failover

This morning, we moved the SQL Cluster from Node1 to Node2. The cluster administrator shows that all disks came online on Node2 but some of the drives was still missing (I checked on Windows Explorer). I managed to move the cluster from Node2 to Node1 back. It took a long time to come online but finally it came online. I notice now that, I can see all the drives mapped correctly on Windows Explorer. But if I go to the disk management, none of the shared disks appear. Did anyone has experienced this problem before ? Thanks,
Gopal
Hi ya, nope can’t say I’ve had this before, but I’d suggest going through the cluster log to see if it shows anything, also SQL error log and nt event log… The drives haven’t had their letters changed on one node and not the other, or something like that? Cheers
Twan
By any chance, are you viewing this through Terminal Services? I say this because I have seem some odd things, similar to this, when using Terminal Services. When you log on locally, there is no problem. —————————–
Brad M. McGehee, MVP
Webmaster
SQL-Server-Performance.Com
Thanks Twan and Brad for your reply. Twan – I did not find anything on NT log and Event log. But I did not check the cluster log. I will look into that tomorrow. This is the first time I have seen this problem. Brad – I did find out this by logging into server directly. Thanks,
Gopal
Have you applied Win2k SP4?
In addition, when one fails cluster over, instead of using Windows GUI stuff, one can use command line option. The command "cluster resource" will display all cluster resources and will also inform who owns those resources. Some other useful commands: Cluster node
cluster group
cluster version One can type cluster in command mode and see all the options. The Win2k SP4 has resolved many OS (memory, disk,network, memory above 4GB) and cluster related problems.
We have applied Post W2K SP4 Cluster hotfixes and I believe this problem has been resolved now. We moved the cluster to Node1 back and we can see all the drives on disk management. Thanks to everyone,
Gopal
]]>