Hello, I've been training the Bayesian filter for some time now and I watch it get called upon in /var/log/maillog. Problem is that it doesn't look like it is adding to my spam scores. I have "use_bayes 1" turned on in my config file and run sa-learn nightly on spam and ham mailboxes I've set up. Using spamassassin 2.55 with qmail 1.03 and qmail-scanner 1.16 ... Here's how it goes down in /var/log/maillog:
Sep 9 10:25:54 ahnold spamd[11821]: logmsg: processing message <[EMAIL PROTECTED]> for qmailq:401. Sep 9 10:25:54 ahnold spamd[11821]: processing message <[EMAIL PROTECTED]> for qmailq:401. Sep 9 10:25:55 ahnold spamd[11821]: debug: bayes: 11821 tie-ing to DB file R/O /var/qmail/.spamassassin/.spamassassin/bayes_toks Sep 9 10:25:55 ahnold spamd[11821]: debug: bayes: 11821 tie-ing to DB file R/O /var/qmail/.spamassassin/.spamassassin/bayes_seen Sep 9 10:25:55 ahnold spamd[11821]: debug: debug: Only 132 spam(s) in Bayes DB < 200 Sep 9 10:25:55 ahnold spamd[11821]: debug: bayes: 11821 untie-ing Sep 9 10:25:55 ahnold spamd[11821]: debug: bayes: 11821 untie-ing db_toks Sep 9 10:25:55 ahnold spamd[11821]: debug: bayes: 11821 untie-ing db_seen Sep 9 10:25:55 ahnold spamd[11821]: debug: is Net::DNS::Resolver available? yes Sep 9 10:25:55 ahnold spamd[11821]: debug: running header regexp tests; score so far=0 Sep 9 10:25:55 ahnold spamd[11821]: debug: decoded MIME header: "Oh Sildenafil Citrate !" Sep 9 10:25:55 ahnold spamd[11821]: debug: decoded MIME header: "Oh Sildenafil Citrate !" Sep 9 10:25:55 ahnold spamd[11821]: debug: running body-text per-line regexp tests; score so far=-0.4 Sep 9 10:25:55 ahnold spamd[11821]: debug: running raw-body-text per-line regexp tests; score so far=2.6 Sep 9 10:25:55 ahnold spamd[11821]: debug: running uri tests; score so far=2.6 Sep 9 10:25:55 ahnold spamd[11821]: debug: uri tests: Done uriRE Sep 9 10:25:55 ahnold spamd[11821]: debug: running full-text regexp tests; score so far=2.6 Sep 9 10:25:55 ahnold spamd[11821]: debug: Razor2 is available Sep 9 10:25:55 ahnold spamd[11821]: debug: Razor2 is available Sep 9 10:25:55 ahnold spamd[11821]: debug: entering helper-app run mode Sep 9 10:26:05 ahnold spamd[11821]: debug: razor2 check timed out after 10 secs. Sep 9 10:26:05 ahnold spamd[11821]: debug: leaving helper-app run mode Sep 9 10:26:05 ahnold spamd[11821]: debug: Razor2 results: spam? 0 highest cf score: 0 Sep 9 10:26:05 ahnold spamd[11821]: debug: all '*From' addrs: [EMAIL PROTECTED] Sep 9 10:26:05 ahnold spamd[11821]: debug: all '*To' addrs: [EMAIL PROTECTED] Sep 9 10:26:05 ahnold spamd[11821]: debug: forged_rcvd_trail: entry 0: by=quarry.com from=quarry.com mismatches=0 Sep 9 10:26:05 ahnold spamd[11821]: debug: Ran run_rbl_eval_test rule RCVD_IN_NJABL ======> got hit Sep 9 10:26:05 ahnold spamd[11821]: debug: Ran run_rbl_eval_test rule RCVD_IN_UNCONFIRMED_DSBL ======> got hit Sep 9 10:26:05 ahnold spamd[11821]: debug: Ran run_rbl_eval_test rule RCVD_IN_BL_SPAMCOP_NET ======> got hit Sep 9 10:26:05 ahnold spamd[11821]: debug: running meta tests; score so far=7 Sep 9 10:26:05 ahnold spamd[11821]: debug: auto-learn? safety=4, ham=-2, spam=15, body-hits=7.4, head-hits=4 Sep 9 10:26:05 ahnold spamd[11821]: debug: auto-learn: currently using scoreset 1. no need to recompute. Sep 9 10:26:05 ahnold spamd[11821]: debug: auto-learn? no: inside auto-learn thresholds or safety zone around required_hits Sep 9 10:26:05 ahnold spamd[11821]: debug: lock: 11821 created /var/qmail/.spamassassin/.spamassassin/auto-whitelist.lock.ahnold.quarry.com .11821 Sep 9 10:26:05 ahnold spamd[11821]: debug: lock: 11821 trying to get lock on /var/qmail/.spamassassin/.spamassassin/auto-whitelist with 0 retries Sep 9 10:26:05 ahnold spamd[11821]: debug: lock: 11821 link to /var/qmail/.spamassassin/.spamassassin/auto-whitelist.lock: link ok Sep 9 10:26:05 ahnold spamd[11821]: debug: Tie-ing to DB file R/W in /var/qmail/.spamassassin/.spamassassin/auto-whitelist Sep 9 10:26:05 ahnold spamd[11821]: debug: auto-whitelist (db-based): [EMAIL PROTECTED]|ip=211.41 scores 0/0 Sep 9 10:26:05 ahnold spamd[11821]: debug: auto-whitelist (db-based): [EMAIL PROTECTED]|ip=none scores 0/0 Sep 9 10:26:05 ahnold spamd[11821]: debug: AWL active, pre-score: 7.1, mean: undef, originating-ip: 211.41.217.113 Sep 9 10:26:05 ahnold spamd[11821]: debug: add_score: New count: 1, new totscore: 7.1 Sep 9 10:26:05 ahnold spamd[11821]: debug: Post AWL score: 7.1 Sep 9 10:26:05 ahnold spamd[11821]: debug: DB addr list: untie-ing and unlocking. Sep 9 10:26:05 ahnold spamd[11821]: debug: DB addr list: file locked, breaking lock. Sep 9 10:26:05 ahnold spamd[11821]: debug: unlock: 11821 unlink /var/qmail/.spamassassin/.spamassassin/auto-whitelist.lock Sep 9 10:26:06 ahnold spamd[11821]: debug: is spam? score=7.1 required=5 tests=HTML_80_90,HTML_IMAGE_ONLY_02,HTML_IMAGE_RATIO_02,HTML_MESSAGE,MIME_HT ML_ONLY,MSGID_GOOD_EXCHANGE,RCVD_IN_BL_SPAMCOP_NET,RCVD_IN_NJABL,RCVD_IN_UNC ONFIRMED_DSBL Sep 9 10:26:06 ahnold spamd[11821]: logmsg: identified spam (7.1/5.0) for qmailq:401 in 10.6 seconds, 875 bytes. Sep 9 10:26:06 ahnold spamd[11821]: identified spam (7.1/5.0) for qmailq:401 in 10.6 seconds, 875 bytes. Sep 9 10:26:06 ahnold spamd[11821]: debug: bayes: 11821 untie-ing Is it not taking the Bayes into consideration because I'm below 200 emails in the DB right now, or am I missing something here? Thanks! James ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
