Mqueue.in huge

[Glenn Steen]

On 22/01/07, Marcello Anderlini <m.anderlini at database.it> wrote:
> Hello to all, I've tried to have a look to wiky but I did not found any good
> answer for me.
>
> I'm using mailscanner 4.50.15 with spamassassin 3.1.7.1.el4.rf
>
> Some time the messagges in mqueue.in rise without any good reason until
> 2000, I check dns and it seems work well, some time I get this
> error:SpamAssassin timed out and was killed, failure 0 of 10. this could be
> related with my problem ? I'm using CBL spam list.
>
> Thanks for any help and sorry for my worst english.
>
Hello Marcello,

The SA timeouts are likely part and parcel off your problem. There are
some likely spots where it is "failing":
- bayes expiry taking "too long".
- some more or less dead BL lookup (or other network test) timing out.
- some "big" or "wrong" ruleset(s) making it take forever.
...

Most likely is the first one (assuming you use bayes, of course:-). If
you have some files like "bayes_toks.expire123456" in the same
directory youstore your bayes database in, this is your problem (a
well-known problem that has been penetrated more than once on this
list).
Instead of me repeating Scotts words, go read
http://article.gmane.org/gmane.mail.virus.mailscanner/48068/match=bayes+expire
or search the archives for Matt Kettlers (final?!) words on
SpamAssassin timeouts/bayes expiry:
http://search.gmane.org/?query=bayes+expire&author=Matt+Kettler&group=gmane.mail.virus.mailscanner&sort=relevance&DEFAULTOP=and&xP=bayes%09expire&xFILTERS=Gmail.virus.mailscanner---A

Matts writings also cover most (if not all) other measures one should
do when SA is "timing out", so do browse the archives...:)

BTW, Your english isn that bad... Try using the best next, and it'll be fine:-D
Cheers
-- 
-- Glenn (who has to catch a train....->me running:-)
email: glenn < dot > steen < at > gmail < dot > com
work: glenn < dot > steen < at > ap1 < dot > se