I'm curious to know what fills up in the recovery log between
database backups. My recovery log is currently 888MB, with a 16GB
database that is currently 60% full (roughly 8.5GB). In spite of
what I consider a fairly large recovery log (approximately 10% of
the current DB size), I can't seem to make it through 24 hours
without triggering a database backup.

We have a regularly scheduled database backup around 3:30pm, yet
I continue to see a database backup triggered during the nightly
backups. My database has never grown more than 150MB in a single
night, so why is my recovery log filling to capacity? (The
current dbbackuptrigger is set to 80%.) I would *really* like to
avoid these extra triggered backups, but I can't continue to
throw disk space at the recovery log. I had to give up mirroring
(which cost me last week when I lost one of my recovery log
volumes) to increase the size of the log.

Is there something accumulating in the recovery log that I can
prune, or that I do not need to retain, that might keep the usage
to a more reasonable volume?

Current platform is AIX 4.1.4, server is 2.1.8. Running on a
SP2 wide node. Any suggestions or comments appreciated.

Thanks. -- Tom

"If I could dot the 'i' in a Michigan     Thomas A. La Porte
 game and the good lord came to take me   Archivist, Feature Animation
 the next day ... at least I could        DreamWorks SKG
 die happy." - Beano Cook, ESPN           <tlaporte AT anim.dreamworks DOT com>