UNSOLVED

jroczniak

updated

14 years ago

J

jroczniak

1 Rookie

•

30 Posts

0

4094

August 2nd, 2012 05:00

Cannot delete LUNs

I have encountered an issue on a 480 that I cannot fix. I am trying to delete 29 LUNs but the system will not let me, I get the pop up warning window that says the LUN is being used by something else such as a snap, mirror, RLP, etc, etc... as far as I can tell none of that is true. There are no storage groups on the system, no snaps, no clones, etc. etc. Below I have some CLI outputs for LUN 19, all the others have the same results....

Navisphere CLI>naviseccli -h x.x.x.x -user nasadmin -password nasadmin -scope 0 getagent

Agent Rev:           7.30.15 (0.44)
Name:                K10
Desc:
Node:                A-APM00104302383
Physical Node:       K10
Signature:           2781304
Peer Signature:      2773447
Revision:            04.30.000.5.524
SCSI Id:             0
Model:               CX4-480
Model Type:          Rackmount
Prom Rev:            5.30.00
SP Memory:           8160
Serial No:           APM00104302383
SP Identifier:       A
Cabinet:             SPE5

naviseccli -h x.x.x.x -user nasadmin -password nasadmin -scope 0 getlun -messner 19 -stack

Listed Driver:             K10RollBackAdmin

                            K10FarAdmin

                            K10SnapCopyAdmin

Navisphere CLI>naviseccli -h x.x.x.x -user nasadmin -password nasadmin -scope 0  mirror -async -setfeature -off -lun 19

Async mirror: Removing driver from lu stack failed.  This LUN is already allocated to a storage group. Check that this is the correct LUN, and if so, remove the LUN from the storage group and try again. (0x79008023)

Navisphere CLI>naviseccli -h x.x.x.x -user nasadmin -password nasadmin -scope 0  mirror -async -setfeature -off -lun 6

Async mirror: Removing driver from lu stack failed.   The shadow LU that we were attempting destroy was associated with a mirror.  Remove the mirror before attempting this operation.  (0x71528274)

As I stated above there are no storage groups or mirrors that I can see, this might be the result of a badly backed out mirrorview migration in the past.

Any ideas?

Thanks!

  • gangas3

    21 Posts

    896

    0

    Posted August 2nd, 2012 07:00

    In this case, the Management Server (CIMOM) had to be restarted twice for the navicli.jar command to run successfully. However, there are occasions when a layered driver such as k10faradmin (MirrorView),remains on the LUN even though no SnapView or MirrorView session is currently running.

    then run

    naviseccli -h <SP_IP_address> mirror -async -setfeature -off -lun  

  • jroczniak

    1 Rookie

    •

    30 Posts

    896

    0

    Posted August 2nd, 2012 07:00

    Do you mean just restarting the management server through http://x.x.x.x/setup ?

    We have done that many times as well as rebooting the SPs many times.

    Thanks.

  • AnkitMehta

    4 Apprentice

    •

    1365 Posts

    896

    0

    Posted August 2nd, 2012 09:00

    The step you can try from your end is SP Reboot.

    There is other option which requires login to SP and with the help of tools installed on SP CLARiiON engineering would be able to cleanup option associated with shadow LUN. Once, it is done we can try the setfeature off command after that which should remove the stack driver.

    About:

    Navisphere CLI>naviseccli -h x.x.x.x -user nasadmin -password nasadmin -scope 0  mirror -async -setfeature -off -lun 19

    Async mirror: Removing driver from lu stack failed.  This LUN is already allocated to a storage group. Check that this is the correct LUN, and if so, remove the LUN from the storage group and try again. (0x79008023)

    It seems like we have a private storage group that MV/A has! If that is the case then the cleanup I mentioned earlier clears any issues with private storage groups also!

    Step you can try:

    - SP Reboot. [One at a time]

    If this doesn't work. Please, open a Service Request with EMC CLARiiON Technical Support, we would analyze the SP Collects and if required, would engage next level

  • jroczniak

    1 Rookie

    •

    30 Posts

    896

    0

    Posted August 2nd, 2012 11:00

    We have actually spent about 12 hours with EMC Support with no resolution yet. They are the ones that ran the commands I put in the post.

    Please advise how to kick it up to Clariion Engineering.

    Thanks.

  • AnkitMehta

    4 Apprentice

    •

    1365 Posts

    896

    0

    Posted August 2nd, 2012 12:00

    Ok got the SR Number and assisting colleague with the SR with time permitting.

    She's on a call with Brian right now!

  • AnkitMehta

    4 Apprentice

    •

    1365 Posts

    896

    0

    Posted August 2nd, 2012 12:00

    Can you PM me the Service Request number?

    There is an internal process, I am sure engineer must be trying his best.

    My answer is based on What details you have provided, it is possible he must have seen something different on Web Ex or may be the SP Collects! We all are trained equally and I believe he is checking what can be done with out impacting the production.