LHN P4300 -Unable to remove/add node in cluster, Storage Server Offline, Not in a management group

[Jignesh Gamdha]

Hi

We have Three LHN P4300 in cluster, 1 node has lost 2 disks and node gone offline and not able to access from CMC.  We have replaced the 2 disks and reimagining the SAN-iQ but still its showing  “Storage Server Offline, Not in a management group”in CMC.

We have also tried to remove from the Cluster but its showing message

“Insufficient space to make the requested change. The storage capacity of a cluster is equal to the capacity of the smallest storage node multiplied by the number of storage node in the cluster”

Regards,

Jignesh

P.S. This thread has been moved from Storage Area Networks (SAN) (Enterprise) to HP StoreVirtual Storage / LeftHand. - HP Forum Moderator

[Torsten.]

You should add the node to the cluster, then use exchange node. The old node will disappear after restriping all data.

See

http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?lang=en&cc=vn&taskId=120&prodSeriesId=3936136&prodTypeId=12169&objectID=c03749322

[Emilo]

Hello,

You should probably contact support. I can not be certain but what has probably occured it that the node that is out of the cluster is probably the node that owns that particular volume. You need to look at the mgrClientBin.log and search for the keyword satus:  it should say 'up, weak, ready if not then look for the keyword blocked if you see blocked you will have to contact support and they should be able to help you.

[Jignesh Gamdha]

Hi,

As per the your previous solution, i have try to add the node in cluster and exchanged it but no luck. I have also take the power cycle of the node but not able to exchanged the node in cluster and also showing the replication degraded/ unrecoverable I/O error on two VDI.

Pl find enclosed error snapshot

Regards,

Jignesh Gamdha

[oikjn]

try and join the management group with the node that shows it isn't in the group.  Sometimes it takes it attempting to join before it lets you place the node into repair mode.

If that doesn't work, then power off (not shutdown) the node and then CMC should typically allow it to be put into repair mode.

If you have a valid support contract you really should just call them as there is probably a simple backdoor way they can fix the problem w/ developer access to the node.

Side note:  why are you using full provisioning on your LUNs?  HP suggests thin and there is very little benefit to full.