Very large queue and very long delays

[Kearney, Rob]

12 seconds for spamassassin to complete sounds slow.  I average about 2-4
seconds per message using Razor/DCC/Pyzor no bayes.  I cannot however tell
you how many messages there are in your "mailejemplo" file.


-rob
-----Original Message-----
From: Ramon Acedo [mailto:ramon at LINUX-LABS.NET]
Sent: Friday, November 14, 2003 11:57 AM
To: MAILSCANNER at JISCMAIL.AC.UK
Subject: Re: Very large queue and very long delays


Hi

> Are you using Razor? If so, read the Razor documentation to find out how
> to
> refresh ("discover" in razor terms) the list of Razor servers. This is
> most
> likely the problem.

Thanks for the answers. I do not use razor and my spamassassin is not so
slow:

$time cat mailejemplo |spamassassin -D -t|grep debug
debug: using "../rules" for default rules dir
debug: using "/etc/spamassassin" for site rules dir
debug: using "/root/.spamassassin" for user state dir
debug: using "/root/.spamassassin/user_prefs" for user prefs file
debug: running header regexp tests; score so far=0
debug: running body-text per-line regexp tests; score so far=0.6
debug: running uri tests; score so far=0.6
debug: running raw-body-text per-line regexp tests; score so far=0.6
debug: running full-text regexp tests; score so far=0.6
debug: Razor is not available
debug: trying Received header date for real time:  Mon, 16 Jun 2003
17:53:05 +0200
debug: no Received headers found, not raising flag
debug: is Net::DNS::Resolver unavailable? 0
debug: looking up MX for 'microsoft.com.'
debug: MX for 'microsoft.com.' exists? 1
debug: is DNS available? 1
debug: DNS MX records found: 2
debug: checking RBL relays.ordb.org., set relay
debug: checking RBL results in set relay for 127.0.0.6
debug: checking RBL results in set relay for 127.0.0.4
debug: AWL active, pre-score: 1.5, mean: undef
debug: Post AWL score: 1.5
debug: is spam? score=1.5 required=6
debug: auto-report? score=1.5 threshold=30

real    0m15.281s
user    0m0.270s
sys     0m0.000s


And I've got in MailScanner.conf SpamAssassin Timeout = 20.

I do not understand what's happening when the queue grows so much. Now
I have just 10-50 messages but two hours ago I had 2000 messages.

I have alse set Debug SpamAssassin = yes but I do not see extra information.

Another question if my spamassassin timeout is 20 and the time used by
spamassassin to scan is 15 why do I get all the time 'SpamAssassin timed out
and was killed' in my logs?

Thanks,

Ramon Acedo
>
> At 14:40 14/11/2003, you wrote:
>>Hi,
>>
>>I'm using mailscanner 4.24-5 in debian woody with sendmail 8.12.3 and
>>kernel 2.4.22.
>>
>>Almost everyday my mqueue.in grows up to 2000 emails pending to be
>>delivered (or queued at mqueue directory) and the delay of some mails
>>reaches 3 hours!
>>
>>The mqueue directory has always about 200 mails pending what is
>> reasonble.
>>
>>I've tried to decrease the Queue Scan Interval to 2, to increase the
>>Max Children to 10 (I've got a 2,4 PIV with 1GB of RAM) and to try
>>Delivery Method to queue and to batch (without any differences). The load
>>is never over 0.7 (maybe more just when MS starts).
>>
>>I have disabled the RBL checks setting "Spam List = " and setting
>>skip_rbl_checks 1 in /opt/MailScanner/etc/spam.assassin.prefs.conf
>>and I have enabled my sendmail to check the relays.ordb.org though
>>MS keeps logging lots of "SpamAssassin timed out and was killed."
>>
>>MS seems to not to change its behaviour processing the queue and after
>>some hours after when the number of mails per hour decreases from
>>about 3000 to about 600 the queue keeps growing and growing.
>>
>>Doew anyone any ideas?
>>
>>Thanks!
>>
>>Ramon Acedo
>
> --
> Julian Field
> www.MailScanner.info
> MailScanner thanks transtec Computers for their support
>
> PGP footprint: EE81 D763 3DB0 0BFD E1DC  7222 11F6 5947 1415 B654
>