Can not delete primary storage pool volumes

[wkbenson]

The disk array for our primary storage pool went bad and after we replaced the controller card we had to rebuild the array. All volumes were lost and we had to create new volumes (with new names). In TSM we varied all the old volumes that were on the array to offline and updated the status to destroyed before creating new volumes. I have tried to delete the old volumes using the command " DELETE VOLUME F:\TSM\SPFVOL420 discarddata=y "

The command fails with the following info: "Discard Data process terminated for volume F:\TSM\SPFVOL420 - volume still contains data"

The volumes are not being deleted even though the command is expressly saying yes.

When doing backups and migrations from the storage pool we are getting messages that the process is skipping damaged files on the destroyed volumes. The new volumes we created are working fine.

We are using TSM 7.1

I would love to get rid of the destroyed volumes and clean up my storage pools. Does anyone have a fix for this?

Thanks in advance for any assistance.

 

[moon-buddy]

I presume you are still using the TSM DB prior to the crash - see "del volhist" and you can delete the volumes after.

 

[wkbenson]

I presume you are still using the TSM DB prior to the crash - see "del volhist" and you can delete the volumes after.

I read up on the command and saw this was for sequential volumes (my problem is with the primary disk pool volumes). I tried it anyway and still could not delete the volumes. Thanks anyway to showing me something new.

 

[moon-buddy]

I believe there is an undocumented option, try: "delete volume <volume_name> discarddata=yes force=yes"

 

[wkbenson]

I believe there is an undocumented option, try: "delete volume <volume_name> discarddata=yes force=yes"

Result:

9:44:05 AM : delete vol f:\tsm\spfvol420 discarddata=yes force=yes
ANR2020E DELETE VOLUME: Invalid parameter - FORCE.
ANS8001I Return code 3.

 

[moon-buddy]

Ok - so that option is not valid, bummer.

Does the volume you are trying to delete physically still there; meaning that you can see it from the OS side?

Do you know from which node the data on the volume belongs to? Deleting the node data will free up the volume for deletion.

 

[wkbenson]

Ok - so that option is not valid, bummer.

Does the volume you are trying to delete physically still there; meaning that you can see it from the OS side?

Do you know from which node the data on the volume belongs to? Deleting the node data will free up the volume for deletion.

The volumes are physically gone. We had to rebuild the disk array from scratch when we replaced the controller card and recreate all the volumes (with new volume names). There are 20 volumes each in 2 different disk pools and approximately 150 nodes (all production servers) spread across these volumes. I would like to avoid deleting and then recreating all these nodes.

I'm waiting for IBM to call me back, but I wanted to touch base with the experts in the field while I wait.

 

[moon-buddy]

If the volume is physically gone, you would need DB2 support - delete the volume from the TSM database.

IBM TSM or DB2 2nd or 3rd level support really needs to be engaged.

Good luck

 

[wkbenson]

If the volume is physically gone, you would need DB2 support - delete the volume from the TSM database.

IBM TSM or DB2 2nd or 3rd level support really needs to be engaged.

Good luck

Thanks for your help.