On Tue, 2004-03-16 at 16:22, Matt Kettler wrote: > At 05:15 PM 3/16/2004, Jens E. Madsen Jr. wrote: > > >I diff-ed the 20_dnsbl_test.cf between the 2.44 and the 2.63-1 and they > >are the same. > > I find that extreemely hard to believe.. 2.44 doesn't have a > 20_dnsbl_test.cf file at all.
You are extremely right, I must have diff-ed it to itself and of course they were the same. > 2.44's DNSBL tests are mixed into 20_head_tests.cf. Comparing the entries for the RCVD_IN_SBL entries, I get 20_dnsbl_test.cf (2.63): # SBL is the Spamhaus Block List: http://www.spamhaus.org/sbl/ header RCVD_IN_SBL eval:check_rbl_txt('sbl', 'sbl.spamhaus.org.') describe RCVD_IN_SBL Received via a relay in Spamhaus Block List tflags RCVD_IN_SBL net And for the 20_head_tests.cf (2.44) I get: # SBL is the Spamhaus Block List: http://www.spamhaus.org/sbl/ . header RCVD_IN_SBL rbleval:check_rbl('relay', 'sbl.spamhaus.org.') describe RCVD_IN_SBL Received via SBLed relay, see http://www.spamhaus.org/sbl/ tflags RCVD_IN_SBL net Which show the tests are different, or are they? I added the 2.44 version of the rules to my /etc/mail/spamassassin directory, i.e. sbl.cf with the rule name changed and a score. This new rule did not trigger. Looking specifically for SBL tags in previous spam, I have them. > >Has anything changed that would affect BL between 2.44 and 2.61? > > Nearly everything has changed between the two. Different back-end code was > added that supports DNSBL sub-queries (one lookup returns multiple > sub-lists, such as NJABL). Different DNSBLs. Many DNSBLs in the 2.44 code > are dead and have been dead for a very long time, resulting in timeouts. > The RCVD_IN_SBL rule above works for 2.44 and not for 2.63 for the email in question. > > Something tells me that DNSBLs were never enabled on your 2.44 setup. If it > was, you would have been suffering through 30 second timeouts for each > email processed. Perhaps Net::DNS is not installed, but got installed when > you upgraded??? > Is there anything that must be done to enable DNSBLs? But some seem to work. The 2.44 setup was out of-the-box Redhat. On my production machine, I started with that version and quickly upgraded to 2.60 or 2.61. I really only tested on the 2.44 machine to verify my Internet interface. Thanks, and now to diff myself, Jens
