P4300 Critical alarm cluster storage node not formatted

[Fred Blum]

Walked into the office this morning to find that the virtual servers had gone down.

VM's had failed but restarted without a problem.

Tracing back the chain of events I found that it seems to have started with one SAN P4300 storage node in a cluster with a critical alarm that "the storage system 1B status in cluster Cl1 is not formatted" time 6:31:24 AM followed by a message that it was Up time 6:31:43 AM and the final message that the volumes were resynching. 

This had triggered on the servers event log iSscsiPrt connection to the target lost, followed by mpio path lost and log in event errors.

IMHO the lose of one node should not have triggered this chain of events. What could have caused this?

TIA

 Fred

[Bichsel]

Hey

I had the same issue today. I know, it's long time since you've posted this message, but maybe you found a reason back then which could give me a clue why this happened?

-> OR MAYBE HP HAS AN ANSWER?

TIA

[Fred Blum]

Hi, I never found a cause. Beter is that it never happened again. I upgraded to 9.5 shortly after that situation and I added two nodes to have a 4 node SAN.

Fred 

[Bichsel]

Thanks

[Paul Miano]

Were you only running a 2 node cluster at the time of the failure? Did you have a FOM or a Virtual Manger running? If you were running a Virtual Manager and the node that went down hosted that manager, your volumes would have gone to a read only state.

Only running with one node should not have caused your VMs to reboot, I would not think, but maybe someone else has a different opinion?

[Fred Blum]

At the time of he failure I was runnning a 2 node with FOM. FOM in VMPlayer on a seperate server disks not on the SAN.

[Markkrol]

Are you running SANiQ 9.0? Maybe this applies to you?

http://jpaul.me/?tag=p4000