IBM CLUSTER

[vallan]

We have an IBM Cluster system with an external storage running Raid Level 5.

Each of the srver have 2 HDD on Raid Level 5 and the IBM storage Drive also has 4 hard disk with 2 drives each on Raid Level 5. 2 of the drives are the Quorum Drive and the last 2 are the Data Drive.

We recently had a power failure and they were all powered up together. All drives in the storage came up with the amber light.

Runnig RAID Server Manager showed all drives coming up as Defunct.

How do I recover from this. We are not able to do anything. Is there a logical proces to follow to bring the HDD back online. The drives are all healthy as it was only a power failure?

Thanks

 

[pgee0909]

Check

http://www.pc.ibm.com/support.

Search for "multiple disk error"
Locate document MIGR-39144 (ServeRAID - Recovering from multiple Defunct Disk Drive (DDD) failures).

If drives were OK, just before power outage, the drives will not be damaged physically, but just set off-line (defunct).

In this case you should be able to force the disks online.
The last disk in the RAID set will automatically rebuild, when reseated in the expansion bay.

When 2 cluster servers are booted on the same time, you could have a SCSI connection conflict.
On a SCSI CLUSTER environment one of the servers is master and has access to the resources.
So shut down the secondary server and start rebuilding on the primary.
Next check all configurations, if OK boot secondary.

I hope this helps in your situation.

Kind regards,
Peter Geelen
Sr. System Engineer