"MC Matt Cooper (2838)" wrote:
>
> Well I think I found something big.  It does not explain everything I see
> but it is definitely more than half.  The first set of servers that we were
> working with were DB2 database stuff.  I have been told that when using the
> DB2 API that DB2 ignores/overrides the copygroup parameters and everything
> is saved until the DB2 API is run to delete it.  I was told by the DB2 guys
> that they thought they had this process running and they just forgot to tell
> me about it.  Well it did not run.  So I had them run it, on all there DB2
> data bases on all there AIX nodes.  I will have to wait till tomorrow after
> we run EXPIRATION processing, see what is not in a PENDING status, and
> re-evaluate this whole process. I may just wait till monday to get any
> easier view of what the whole old tape scene look like.
>     I have found 2 'test' nodes and 1 node that has changed names with a
> hardware change (not a good idea).  I am sure there is some 'ONLY COPY'
> files associated with them.  IF I still have problems I will restate this
> topic.

...and the lesson for today, folks is:

When you use something other than the vanilla b/a client for backups,
such as TDP or the DB2 API, that normal retention rules and expirations
will not affect files backed up like that. Each TDP, and the DB2 API,
has its own methods and utilities for expiring old file versions.[1]

[1] Learned this the hard way with MS SQL files. When the mistake was
finally caught and proper deletions and expiration were performed, the
tape library lost 20% of its volume, and the TSM database went from 35Gb
to 24Gb.

--
Mark Stapleton (stapleton AT berbee DOT com)
Mark Stapleton (stapleton AT berbee DOT com)