Remove storage

[sepj12927]

Hi,

Yesterday I removed som LUN:s that had been used for RDM:s and datastores.

The procedure I used was to first detach the LUN from all nodes in the cluster and the remove it in the storage system.

However, today some 18 hours later I still have active monitor alerts for those luns.

I've done a resend topology but they are still there.

How do I solve this?

/Per

[rasmushc]

Did you made a Rescan of storage on the ESX host where you removed the disks?
Otherwhise I think you should do that.

[sergey.g]

Hello,

I would agree with Rasmus, vCenter will not remove a storage path unless you did a rescan of storage. Try to do a rescan - this should remove those disconnected luns from our topology.

Thanks.

[sepj12927]

I abviously did multiple rescans but the alerts stayed active.

[sergey.g]

Hello,

So, am I right that you do not see those LUNs in VI client anymore, but they are still present in SCOM?

Is it possible that you had multiple storage paths for those RDMs and Datastores? You've deleted one storage path, but the other one can be still active, or it could be active on another host and you might see alerts from storage paths with same names but from other hosts.

Thanks.

[sepj12927]

I've removed all paths and deleted the LUN in the storage system. It took abount 3 days for the alert to resolve.

[sergey.g]

Hi,

I'm glad to know that eventually those paths went away in SCOM. Although it's bad that it took 3 days, maximum delay for such things should be 24 hours. However VMware sometimes could keep information about storage paths for a while, did you perform storage rescans immediately after removing storage paths? If you did rescans later, then it could take another 24 hours after the rescan which removed the last path to the storage device.

Thanks.

[sepj12927]

Did rescans directly after I romved the last path.

[sergey.g]

Hi,

Sorry for a delay in our response. Do you recall which monitors were not resolved for 3 days? Was it HBA-targeted monitor or some host related rules/monitor for storage issues?

I have a thought that although you've removed several LUNs, some other storage paths may be in a critical state and because of that monitor were still present and in red, same thing could be true for host-related monitors.

Also could you remind me which version of SCOM, our product and vCenter do you have?

Thanks.

[sepj12927]

Hi,

I logged a call with Support, and apparently there is a bug in the version (5.7.1022) we're running that can cause this behaviour when the storage is pulit pathed.

/Per