efa-project.org

Hello, I've got a few week old installation that has been working pretty well, but just this morning, there has been a huge uptick in spam getting through. The difference between is those messages don't list any sort of Bayes decision.
All of the correctly identified spam emails have the Bayes rating at the very top of the spam score tally.

Appreciate any insight and help.
Thanks.

A typical score tally for a spam email might look like this. ( no Bayes )

0.10 DKIM_SIGNED Message has a DKIM or DK signature, not necessarily valid
-0.10 DKIM_VALID Message has at least one valid DKIM or DK signature
-0.10 DKIM_VALID_AU Message has a valid DKIM or DK signature from author's domain
1.00 HEADER_FROM_DIFFERENT_DOMAINS From and EnvelopeFrom 2nd level mail domains are different
1.00 HTML_IMAGE_RATIO_02 HTML has a low ratio of text to image area
0.00 HTML_MESSAGE HTML included in message
1.11 MPART_ALT_DIFF_COUNT HTML and text parts are different
-0.01 RCVD_IN_MSPIKE_H3 Good reputation (+3)
0.10 RCVD_IN_MSPIKE_WL Mailspike good senders
-0.00 SPF_HELO_PASS SPF: HELO matches SPF record
-0.00 SPF_PASS SPF: sender matches SPF record

This would be a function of spamassassin. I would start by looking at your SA configuration and seeing what might be causing bayes to be bypassed. For example, are there any custom rules defined?