I have a question.... In the spamdyke log, you see any GRAYLIST line ?
2011/3/15, Scott Hughes <sc...@renshawauto.net>: > > > On 3/15/11 7:53 PM, Eric Shubert wrote: >> On 03/15/2011 05:47 PM, Scott Hughes wrote: >>> I'll have to do some research on Spamassassin in that case as I do not >>> know enough to "look closer into your bayes setup" right now. >>> >>> I did notice this in the spamd log: @400000004d80080b0cf39ed4 Mar 15 >>> 19:44:49.217 [5208] info: spamd: result: Y 3 - >>> AWL,PYZOR_CHECK,RDNS_NONE,SARE_UN4,T_FILL_THIS_FORM_SHORT >>> scantime=0.7,size=2124,user=clamav,uid=89,required_score=3.5,rhost=localhost.localdomain,raddr=127.0.0.1,rport=34382,mid=<efa1376656ef1851016d7716ad2c5...@www.facebook.com>,autolearn=no >>> >>> >>> >>> What I am curious about is the last bit (autolearn = no). Is this >>> something I should be worried about? >> >> No. You have autolearn turned on, but the score is such that SA is >> saying "I'm not going to autolearn this as ham or spam". There are >> thresholds for each. You should probably have these in your local.cf >> file in addition to what you already have: >> bayes_auto_learn_threshold_spam 5.5 >> bayes_auto_learn_threshold_nonspam 0.1 >> bayes_auto_expire 0 >> >> These will help to populate your bayes database. >> > > > Much better now (the last bit): @400000004d800cc132ef257c Mar 15 > 20:04:55.854 [5995] info: spamd: result: Y 6 - > AWL,HTML_MESSAGE,MISSING_MID,RAZOR2_CF_RANGE_51_100,RAZOR2_CF_RANGE_E8_51_100,RAZOR2_CHECK,RDNS_NONE,URIBL_OB_SURBL,URIBL_RED > scantime=1.8,size=10125,user=clamav,uid=89,required_score=3.5,rhost=localhost.localdomain,raddr=127.0.0.1,rport=33077,mid=(unknown),autolearn=spam > > Thanks, > Scott > > >>> As far as the wiki - I'll be more comfortable writing the wiki page on >>> this when I have a better understanding. >>> >>> Thanks, >>> Scott >>> >>> >>> On 3/15/11 7:42 PM, Eric Shubert wrote: >>>> I would look closer into your bayes setup, as none of the spam headers >>>> you posted had a bayes score at all. I would expect to see one of the >>>> bayes rules hit, which leads me to believe there's something in your >>>> sa-learn processing that isn't quite working. >>>> >>>> Good idea about the wiki. Any interest in writing it? (While all this >>>> is fresh in your mind) ;) >>>> >> >> > > --------------------------------------------------------------------------------- > Qmailtoaster is sponsored by Vickers Consulting Group > (www.vickersconsulting.com) > Vickers Consulting Group offers Qmailtoaster support and installations. > If you need professional help with your setup, contact them today! > --------------------------------------------------------------------------------- > Please visit qmailtoaster.com for the latest news, updates, and > packages. > > To unsubscribe, e-mail: qmailtoaster-list-unsubscr...@qmailtoaster.com > For additional commands, e-mail: > qmailtoaster-list-h...@qmailtoaster.com > > > -- Enviado desde mi dispositivo móvil --------------------------------------------------------------------------------- Qmailtoaster is sponsored by Vickers Consulting Group (www.vickersconsulting.com) Vickers Consulting Group offers Qmailtoaster support and installations. If you need professional help with your setup, contact them today! --------------------------------------------------------------------------------- Please visit qmailtoaster.com for the latest news, updates, and packages. To unsubscribe, e-mail: qmailtoaster-list-unsubscr...@qmailtoaster.com For additional commands, e-mail: qmailtoaster-list-h...@qmailtoaster.com