loosing local datastore

[wallbreakr]

I replaced the megaraid MR10k controler of my x3950M2 ESXi server after a battery problem.

After the boot, the local datastore (vmhba1.0.0) disappear.

Is it possible to mount it manually ?

When I try to add storage with the VI Client, I have a warning message : The current disk layout will be destroyed, but I have some VM on my local storage that I don't want to loose...

Please Help !

[depping]

Did you already try a rescan of the HBA? Sometime doing it multiple times helps,

Duncan

My virtualisation blog:

If you find this information useful, please award points for "correct" or "helpful".

[wallbreakr]

some strange messages in the /var/log/messages attached to the post :

Sep 11 07:46:28 vmkernel: 0:00:00:20.734 cpu3:1261)WARNING: SCSI: 279: SCSI device type 0xd is not supported. Cannot create target vmhba1:256:0

Sep 11 07:46:28 vmkernel: 0:00:00:20.734 cpu3:1261)WARNING: SCSI: 1290: LegacyMP Plugin could not claim path: vmhba1:256:0. Not supported

Sep 11 07:46:28 vmkernel: 0:00:00:20.734 cpu3:1261)WARNING: ScsiPath: 3187: Plugin 'legacyMP' had an error (Not supported) while claiming path 'vmhba1:C0:T256:L0'.Skipping the path.

Sep 11 07:46:28 vmkernel: 0:00:00:20.744 cpu3:1261)ALERT: LVM: 4476: vml.02000000006005076b0608c340ff00008c08a91b6f536572766552:2 may be snapshot: disabling access. See resignaturing section in SAN config guide.

[wallbreakr]

I tried the rescan, no change

[wallbreakr]

I found this : http://theether.net/kb/100026

but I can't rename the datastore to the old name because the old name already exists !

[wallbreakr]

to solve the problem, I need to :

• use a new name for my datastore

• remove the VM from inventory

• browse de datastore

• add my old VM to inventory

Everything work fine now with a new datastore name.

[depping]

I should have thought of that one, the identifier changed...

Duncan

My virtualisation blog:

If you find this information useful, please award points for "correct" or "helpful".