Expiring Media - Timed out connecting to client... 1

[leopard96]

I'm trying to expire some tapes to re-import them due to a botched server migration. It isn't letting me expire the media. I've tried to use the -force option but it says "timed out connecting to client".

The catalog believes that the media is owned by another server because we got that far in our migration before we found out that the "new" (used) server decided to die on us. We can't bring it back to life to transfer the media back so I need to re-import them if I can.

Any ideas are greatly appreciated!

 

[lenski]

Don't expire the media but delete them. Then you will be able to import the old media as new tapes. But if you have lots of tapes its a big job.

 

[Stumpr]

Follow the procedures in this technote:
How to manually remove a tape completely from NetBackup/Media Manager.

http://seer.support.veritas.com/docs/208420.htm

Details:
First, if there are images on the tape, they need to be expired. This can be done by:
bpexpdate -ev <mediaid> -d 0

This will not completely free up the tape if the tape is frozen or suspended, so the command(s):
bpmedia -unfreeze -ev <mediaid>
or
bpmedia -unsuspend -ev <mediaid>
may be needed.

At this point there will be no image database entries for that tape and the media database should show the tape as not assigned therefore allowing the tape to be reused. However sometimes the image database and the media manager database can be out of sync, and the tape may show as still assigned, or if the tape has a NetBackup database backup on it, there would be no image entry to begin with. When a tape shows an assigned date, but no images are related to that tape, it will show up in the available_media report as dbbackup. To tell if a tape actually has a dbbackup on it, the command:
vmquery -m mediaid
will show a status, where 0x0 means normal backups, 0x1 means NetBackup database backup and 0x2 means Storage Migrator.
If it is 0x1, be certain this is not a database backup which will ever be needed.
If it is 0x2, it is in use by Storage Migrator and understanding of the use of the tape by Storage Migrator (not covered here) is required.

To free up the tape, when it has an assigned date but no images in the database, the command:
vmquery -deassignbyid <mediaid> <poolnumber> <status>
must be run, but the status will be displayed as 0x0, 0x1 or 0x2 and needs to be just 0, 1 or 2 in the vmquery -deassignbyid.

At this point the tape is available to be used for backups.

If the tape was bad or destroyed and that is why it is being removed, the command:
vmdelete -m <mediaid>
will remove the last remnants of information about it from media manager.

There are other options which may be specified for the above commands. Some are to specify server names, volume database host names, etc. If the NetBackup/Media Manager setup is complex, those options may be needed. The full paths to the commands mentioned are:

/usr/openv/netbackup/bin/admincmd/bpexpdate
/usr/openv/netbackup/bin/admincmd/bpmedia
/usr/openv/netbackup/bin/goodies/available_media
/usr/openv/volmgr/bin/vmquery
usr/openv/volmgr/bin/vmdelete
All but available_media have man pages.


Bob Stump
Incorrigible punster -- Do not incorrige

 

[leopard96]

Hey All,

Sorry for the late reply...

I've tried to expire the tape with the images on them and it says they're not present in the volume pool or MM. It is present in Media listings but it refuses to let me expire the tape or remove it from media manager.

"D:\VERITAS\Volmgr\bin>bpexpdate -ev ir0020 -d 0
Are you SURE you want to delete IR0020 y/n ? y
requested media id was not found in NB media database and/or MM volume database"

It appears in the media list and has an expiration in June. although it says 3/14/05 below.

Funny thing is the tape name is ir0020 but the label is ir0002. It used to be my catalog backup tape. I've since made a new one but I'd still like to be able to use the tape again.

VMQuery gave me:

media ID: IR0020
media type: DLT cartridge tape 3 (26)
barcode: IR0002S
media description: Added by Media Manager
volume pool: NDMP (4)
robot type: TLD - Tape Library DLT (8)
robot number: 0
robot slot: 2
robot control host: shp-us2k009
volume group: 00_000_TLD
vault name: ---
vault sent date: ---
vault return date: ---
vault slot: ---
vault session id: ---
vault container id: -
created: 4/28/2004 11:07:45 AM
assigned: 3/4/2005 7:07:00 PM
last mounted: 3/14/2005 4:58:55 PM
first mount: 5/14/2004 10:43:19 PM
expiration date: 3/14/2006 5:05:07 PM
number of mounts: 45
max mounts allowed: ---
status: 0x0

Any other ideas?

 

[leopard96]

Sorry for the double post everyone...

The vmquery -deassignbyid command did take care of it. Once that happened, I could remove it from MM.

I think I originally had the syntax either wrong or mis-spelled.

I think I'm good now.

Thanks to everyone for the help!