On Wed, 28 May 2003, Stuart Gall wrote: > ----- Original Message ----- > From: "Jack Gostl" <[EMAIL PROTECTED]> > To: "SpamAssassin listserve" <[EMAIL PROTECTED]> > Sent: Wednesday, May 28, 2003 5:40 PM > Subject: [SAtalk] 2.53 vs 2.55 > > > > > > I just upgraded from 2.53 to 2.55. Was there some discussion awhile back > > about 2.55 having a problem with RBL? > > > > I used to do RBL checks at SMTP time, but unfortunately over here there is an > increasing number of open relays in the end I just gave up and decided to let SA do > the checks. I just upped th score on all the blacklists I was using. (since I was > rejecting them before) I was convinced that 2.55 wasn't doing RBL checks. > Before I could sit and watch rejects in the mail log any time of day. But I didn't > seem to be getting any spam with RBL clasified. > Since I added Razor2 magically RBL checks seem to be appearing. > > Now the docs say RBL is requested at the start of the SA run, then if no answer is > recieved at the end it waits > up to rbl_timeout before giving up. > > I wonder if this is so for a message that is already clasified as spam, certianly if > the score is above the auto_learn threshhold there is little point in waiting for > RBL. > > The only thing I can think of is that rbl_timeout is too low and by adding razor I > have increased the time to do the scan so that the RBL replies come in. However that > being the case the rbl_timeout is either not working or not defaulting to 30 because > it was not taking 30 seconds to scan an email. > > I think it would help if the debug output showed when an RBL is requested, when the > reply is ok, and when SA is giving up on the RBL. ATM it only shows a positive reply. > > All I ever get is > debug: Ran run_rbl_eval_test rule RCVD_IN_NJABL ======> got hit > debug: Ran run_rbl_eval_test rule RCVD_IN_OSIRUSOFT_COM ======> got hit > debug: Ran run_rbl_eval_test rule RCVD_IN_RELAYS_ORDB_ORG ======> got hit > debug: Ran run_rbl_eval_test rule RCVD_IN_UNCONFIRMED_DSBL ======> got hit > > or nothing. > Is that right ???
Well, so far its just a memory of a thread I thought I saw combined with the fact that none of the false negatives have had any RBL hits. Its not statistically significant yet. Frankly, I'm getting spoiled. On 2.60cvs I was gettin 99.5-99.8%% spam captured. On 2.55, I seem to get about 1% less. I should be embarrassed. A month ago I would have swooned over 75%. -- Jack Gostl [EMAIL PROTECTED] ------------------------------------------------------- This SF.net email is sponsored by: ObjectStore. If flattening out C++ or Java code to make your application fit in a relational database is painful, don't do it! Check out ObjectStore. Now part of Progress Software. http://www.objectstore.net/sourceforge _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
