cmpilato@collab.net writes:
> Ben Collins-Sussman <sussman@collab.net> writes:
>
> > I suspect the '15 minutes' thing was happening because there were
> > waaaaay too many useless log files in the repository... and db_recover
> > was needlessly running over them.
> >
> > (The 'hot-backup.py' script that we run on each commit wasn't fully
> > doing it's job: it was removing useless logs from the *backup* copy,
> > not the live repository. I think this still needs to be fixed!)
>
> Heh...I was just making this change. Of course, the Berkeley docs say
> that we should NOT be deleting those logfiles, but be archiving them
> to removable media instead for catastrophic recovery.
>
hot-backup.py should do one thing: make a backup. It does that. If
it wants to remove useless log files in the backup, that's fine with
me. If we want to be extra-paranoid and keep them around, that's ok,
I'm -0. (We've already got multiple backups lying around.)
But we need a *separate* commit-hook script that cleans up logfiles in
the live repository., don't you think?
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Oct 21 14:36:59 2006