I actually had no idea SpamAssassin could do bayesian training - I'm an
idiot. Lol. Thank you for pointing that out to me!
On 4/19/2017 2:56 PM, RW wrote:
On Wed, 19 Apr 2017 14:36:24 -0400
W B wrote:
Hey all! I'm having issues with SpamAssassin; it's assigning emails
scores that are way lower than it should. In addition, the scores
it's assigning as emails come in are different from the results of
running SpamAssassin -t on that same email after the fact.
I'd like to avoid posting too much text in a single email, so here's
a link to a Reddit post I made outlining the issues I'm having:
https://www.reddit.com/r/webdev/comments/660n2w/spamassassin_scores_are_oddly_low_different_from/
If anyone has any clue what the issue is or is able to help, please
let me know!
This is normal. In the interval the IP address was reported to spamcop
and a URI in the email was reported to SURBL.
I'm getting flooded with spam emails to the point that
it's hard to find my actual legitimate emails in my inbox...
There are two obvious problems:
1. URIBL_BLOCKED, you need to setup your own nameserver that does its
own DNS lookups. At the moment you are probably using a shared DNS
cache which leads to a lot of look-ups coming from the same IP
address.
2. You aren't using Bayes - this is the main problem.
