Slow expireover (again)
rra at stanford.edu
Sun May 23 07:13:32 UTC 2004
Mike Zanker <mike-sender-6677e0 at zanker.org> writes:
> Also, expireover seemed to complete quicker when I was using timecaf as
> opposed to timehash. Does that make sense, though, because I thought
> that expireover only looked at the overview database and the history
> files, not the article spool?
With CNFS and timecaf, expire will actually check to see if the article is
still in the spool because both of those storage methods can answer that
question very quickly. With timehash and tradspool, it will instead query
the history file. So I take that back; there should be few if any stats
(at least of articles in the spool) happening when expireover is running.
Could you paste a portion of the truss output just so that I can look it
over and make sure I understand what's going on?
I assume that this is tradindexed overview, as well....
Russ Allbery (rra at stanford.edu) <http://www.eyrie.org/~eagle/>
Please send questions to the list rather than mailing me directly.
<http://www.eyrie.org/~eagle/faqs/questions.html> explains why.
More information about the inn-workers