Problem with windows cluster administrator

Hi all,

Scenario:
i> Two servers configured in a cluster
ii> I am using two databases named as 'NEFT' and 'SFMS' in cluster environment.
iii> I am using one application 'IBM MQSeries', which uses these databases
iv> Using Oracle 9i v2 on windows 2000 server

Yesterday, my client complained that databases are not coming up.I saw that the the status of both of the databases is showing 'offline' in Microsoft cluster Administrator.The startup command at SQL prompt also failed to open up the database.Then I fired 'startup force',the database opened.and the applications became functional which were using the databases.
But the status of 'NEFT' and 'SFMS' in cluster Administrator is still showing offline.I right clicked on NEFT and tried to bring it online. But a message appeared on the screen:

"No possible owners are specified for this resourse.This means that the cluster can not be brought online on any node in the cluster."

I saw the property of the NEFT resourse,on the general tab, there was no owner specified and the modify button was deactivated.
same is the case with SFMS.
I closed the Cluster administrator window and restarted it again. I found three error messages:

1> An error occurred while attemting to read properties for 'IBM MQSeries MSCS' resourse type:
Invalid class string
Error ID: 2147221005 (800401f3)

2> An error occurred while attempting to read properties for 'NEFT'.
The specified module can't be found
Error ID: 126 (0000007e)
3> An error occurred while attempting to read properties for 'SFMS'.
The specified module can't be found
Error ID: 126 (0000007e)

and again the status of the 'NEFT' and 'SFMS' appears offline. Database is working fine. What may be the reason?

Thanx in advance,

 

Thanks Arie,

But i am talking about SQL Server, I am using Oracle 9i(2)

Regards

 

When troubleshooting a cluster, you typically need to start low level by making sure the physical disk, network, and IP resources start up within the resource group of your SQL cluster. When troubleshooting this, you typically want to shutdown all but one node to ensure there isn't a physical disk conflict. Then make sure that those three resources start.
If they all start, then bring up the SQL resources, which should depend on the disk and IP resources within the same resource group.
Do you have a SCSI disk cluster, or fibre channel disk cluster?