report globally, check locally?

[Paul B. <pau...@gm...>]

Hi all,

On my mailserver I collect spam via ~300 spamtrap addresses, feeding
these messages directly to "pyzor report" via procmail. Recently I
have switched to running pyzord locally for performance and
reliability reasons. The "reliability" part is actually twofold --
one, I was seeing timeouts on 5 - 10% of connection attempts at times;
two, I realized that I had a very good local stream of data and was
not gaining that much from the remote checks.

However, I would also like to be useful to the pyzor community and so
I'm considering setting things up so that inbound mail (excepting
spamtrap mail) is checked against the local db, while spamtrap mail is
reported to both.

Anybody doing anything like this?  Caveats or suggestions?

pb


--=20
paul bissex, e-scribe.com -- database-driven web development
413.585.8095
69.55.225.29
01061-0847
72=B039'71"W 42=B019'42"N