Customisable expiry - how?
rra at stanford.edu
Mon Jan 22 07:16:11 UTC 2001
News Subsystem <newsaccount at free-pr0n.netscum.dk> writes:
>>> I can locate the articles in the spool directory easily enough, and
>>> I'm sure unlink()ing them from Perl wouldn't be too hard, but
>>> obviously that doesn't take in to account any of INN's overview and
>>> history databases. Given that I'm talking about a maximum number of
>>> about 500K articles in the spool directory at any one time (and that's
>>> an absolute maximum), I thought that I could just regenerate the
>>> indices after purging the files.
> Sounds reasonable. I'm going to assume, as you refer to being able to
> unlink individual articles, you're referring to traditional one-
> article-per-file directory-based spool. Also, I'll assume you are
> referring to INN before the 2.3 family, as the commands I'm going to
> refer to are in the 2.2.3 release, but I'm pretty sure have changed in
> the transition to the storage API...
Incidentally, if you want to do this same thing under INN 2.3, you can
still just unlink the articles that you want to have go away. Both
expireover and expire should then pick up the fact that the articles are
missing and make the history and overview entries go away.
Russ Allbery (rra at stanford.edu) <http://www.eyrie.org/~eagle/>
More information about the inn-workers