In regard to: Re: [Networker] not found in media index, Darren Dunham said...:

>>         mminfo -q volid=2248550913 -r 'volume,volid,location'
>>
>> and
>>
>>         mminfo -q volid=2295852801 -r 'volume,volid,location'
>>
>> Do you get output for both?
>
>Nope, only the first....

That's probably good.

>> In your situation, I would run
>>
>>         nsrim -X
>>
>> and then *delete* QF2345 from your media database, and then run
>>
>>         nsrim -X
>>
>> again.
>
>Makes sense.

:-)  But did it help?

>> >My mmdbd is constantly at the top of the CPU list, and mminfo queries
>> >are often fast, but occasionally they take 20 seconds (or worse, time
>> >out).
>>
>> I wouldn't worry about fast vs slow, but I would definitely be concerned
>> about a query that times out.
>
>It's just slow in a 'not normal' sense.  I can do a query on a single
>volume that should return almost instantly.  If I do it 5 times, I'll
>get a few 'instantly's and I'll get one or two '20 seconds'.

If it's the same query with wildy differing times, I agree that it's more
of a problem.  It could be something else on the box, but I too would
be concerned.

>From my prior experience with duplicate volids or other "duplicate media
database entries", my sense (without any strong proof) is it happens at
relabel/recycle time.  I know that when I was playing with Thierry's
reporting tool last week, it detected that our server's media database
had two sets of volumes that had the same barcode.  I fixed it by figuring
out in each case which volume was the "phantom" and which was the real
volume, and deleting the phantom.  In both cases, it looks like it
happened when NetWorker tried to recycle a volume and only partially got
the job done, leaving some stale information in the database.

I don't know if this was a holdover from 6.1.2, or if it happened after
we upgraded to 6.1.3.

The obvious way to fix the problem is to do what Legato recommends -- move
the media database out of the way with NetWorker shut down, and then bring
it back online.  My recollection is that a blank media database will be
recreated.

At that point, you have the tedious job of running `scanner -m' on all
your tapes, to repopulate your media database based on the available
tapes.

Doing an `mmrecov' in this situation isn't as good, since you don't know
when the corruption (if that's what it is) started, and you'll end up
losing information for any savesets that have happened since the last
bootstrap.

Tim
--
Tim Mooney                              mooney AT dogbert.cc.ndsu.NoDak DOT edu
Information Technology Services         (701) 231-1076 (Voice)
Room 242-J6, IACC Building              (701) 231-8541 (Fax)
North Dakota State University, Fargo, ND 58105-5164

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=