https://bz.apache.org/SpamAssassin/show_bug.cgi?id=7592
--- Comment #5 from John Hardin <[email protected]> --- (In reply to steven.lynch from comment #4) > Hi John, > > Thanks for your prompt response. > > I was hoping that the background debug I had done would be enough. It is > difficult to include a sample email for you because the email actually > belongs to one of our customers. But I can ask them if it is OK if you > "really" need to see one. Although since you've now seen this in your own > Spam folder I'm hoping that you won't really need a whole message. No, I don't need to see the whole message now, thanks. I'd still like to see the headers only, and it's fine to sanitize the recipient info as that's not relevant for what we're looking at. > Not everything that comes from Sendgrid is Spam ... github, quora, my local library, ... > and their RDNS does trigger > those rules, so I guess it is a bit of a no brainer really. I do note the only place I've seen that hostname pattern on a sendgrid MTA is on spams; I presume that's because the sendgrid client knows they are going to be kicked soon so doesn't bother setting up a domain-specific hostname+RDNS. I'm wondering whether those messages in your case *are* actually ham... If they are, it might be worthwhile to contact the sender (not sendgrid, the company who is using sendgrid for their outbound mail) and suggest they need to fix their outbound mail config. > We do have Bayes enabled and trained, although in our case SA is not our > entire Anti-Spam solution but one engine in a multi-engine filter. Since RBL > checks and various other network related things are done elsewhere, SA is > run without net tests. OK, that does affect the score set used as well. I think I'm going to add an exclusion for HELO_DYNAMIC_IPADDR to avoid the overlap and because that's the higher-scored rule. We'll need to look a bit deeper into the actual scores. -- You are receiving this mail because: You are the assignee for the bug.
