-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
comment out the loadplugin lines in /etc/mail/spamassassin/init.pre. also, note that you can easily use multiple spamds, on different ports. - --j. Dave Stern - Former Rocket Scientist writes: > Now that I've got SA3.0.rc1 working, indeed, I like it a lot. As a > test, I have the last rule in my procmailrc take everything that wasn't > filtered by my other procmail rules and SA(2.64) and forward it to > another host running SA3.0rc1. Sure enough, the other machine catches > all the spam that the 2.64 host bled thru. > > That's the good news. Now in our environment, aside from the fact that > users must explicitly turn on SA via procmail rather than the MTA > running it, we're also running spamc. So on the old machine, in order > to have two iterations of SA, I presume I can only have one running the > 2.64 spamd and all others that would care to beta test 3.0 would have > to explicitly call spamassassin rather than spamc. Even with earlier > versions of SA, I've noticed this to take a CPU penalty of nearly an > order of magnitude. But worse, the remote connects to test for SURBL, DNS > take a rather extreme amount of time and anyone running spamassassin (the > 3.0 version of the script) directly will be really giving the CPU a workout. > (OTOH, I recognize that perhaps it's these tests that give such a robust > ratio of catching spam, nevertheless,) Is it possible to turn off all > these checks ALA > use_dcc 0 > use_pyzor 0 > ... > > If so, which rules are they? > > TIA > > =-=-=-=-=-=-=-=-=-=- generated by /dev/dave -=-=-=-=-=-=-=-=-=-=-=-= > David Stern University of Maryland > Institute for Advanced Computer Studies -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Exmh CVS iD8DBQFBKjVxQTcbUG5Y7woRAoDbAKDN0ovEVJWEhyM4yh9Y5Yin1jV2DQCgwYqb +r9udBzESrhp8XGD2jHqFec= =zc05 -----END PGP SIGNATURE-----
