I've been trying to tune a QS + SpamAssassing + ClamAV environment for about 100 users - nothing too serious. But I find that scanning these messages is taking quite a long time. On average about 45 seconds, sometimes as long as 200 seconds.
My configuration: Server: Dual P4 Xeon 2.4 Ghz - 1 GB Memory - Debian Spamassassin: 2.63 running in spamd daemon mode startup with options: -d -a -q -x -H /etc/razor -u spamd Razor and Pyzor and Bayes enabled. ClamAV: 0.73 - Configured normally, no major config changes that I am aware of. clamd daemon running. Q-S: Version 1.22 configured with: ./configure --admin josh --domain www.netbits.us --scanners clamdscan,verbose_spamassassin --add-dscr-hdrs yes --unzip yes --block-password-protected yes --install When I watch the logs, Q-S says a message is scanned in about 50 seconds total. The Q-S log claims that SA took about 40 seconds of that time. SA in it's own logs claims it scanned the message in 1.5 seconds. So who's fibbing or is the time being spent in some other strange place: >From SA Log: Jun 15 14:50:40 fastconcepts spamd[12090]: identified spam (7.6/5.0) for [EMAIL PROTECTED]:1002 in 1.4 seconds, 8790 bytes. >From Q-S log: Tue, 15 Jun 2004 14:50:39 CDT:12070: SA: run /usr/bin/spamc -f -u "[EMAIL PROTECTED]" < /var/spool/qmailscan/working/new/fastconcepts108732903952512070 Tue, 15 Jun 2004 14:51:08 CDT:11980: SA: overwriting /var/spool/qmailscan/working/new/fastconcepts108732901952511980 with /var/spool/qmailscan/working/new/fastconcepts108732901952511980.spamc Tue, 15 Jun 2004 14:51:08 CDT:11980: SA: yup, this smells like SPAM Tue, 15 Jun 2004 14:51:08 CDT:11980: spamassassin: finished scan of dir "/var/spool/qmailscan/tmp/fastconcepts108732901952511980" in 39.668424 secs ^^^^^^^^^^^^^^ Tue, 15 Jun 2004 14:51:08 CDT:11980: scanloop: finished scan of "/var/spool/qmailscan/tmp/fastconcepts108732901952511980"... T And at the bottom: Tue, 15 Jun 2004 14:51:08 CDT:11980: cleanup: /bin/rm -rf /var/spool/qmailscan/tmp/fastconcepts108732901952511980/ /var/spool/qmailscan/working/new/fastconcepts108732901952511980 Tue, 15 Jun 2004 14:51:09 CDT:11980: all finished. Total of 49.374919 secs It looks like SA is lying based on the log output from QS, though I'm not positive. I tried to disable bayes, razor and pyzor in SA to see if performance would improve, but that didn't seem to help. Any thoughts? Thanks, Josh ------------------------------------------------------- This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND _______________________________________________ Qmail-scanner-general mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/qmail-scanner-general
