Hello, all. As someone new to Scalix, sendmail and e-mail in general, I'm quite confused about some of my maillog errors. When I do mailq, I see many thousands of entries on a small e-mail system. I am assuming these are tons of retries on bad mail.
I notice many are addresses to may own domain - atlasgroup.net. I assume it's mostly bad folks sending mail to randon names in my domain. Perhaps some is local system mail? There are lots of local configuration errors and complaints about the relay being mail1dc1.atlasgroup.net and sending to itself.
I've read several posts here and on sendmail sites that say to add an entry to /etc/mail/local-host-names with mail1dc1.atlasgroup.net. That did not solve the problem (I did remember to restart sendmail -several times!).
I also wonder if that is even correct. If I understand the flow correctly, Sendmail is not doing to final send. It is only listening on 127.0.0.1 and not the real address. In fact my /etc/hosts file has
127.0.0.1 localhost localhost.localdomain
172.x.x.31 mail1dc1.atlasgroup.net
lsof -i :25 shows that port 25 is bound by each of those names properly between sendmail and omsmtpd.
Wouldn't the correct path for local system mail be to sendmail which hands it off to mail1dc1 which treats it as local mail and discards it when there is no recipient found?
If it is bad mail, why is it in sendmail at all? I would think omsmtpd would get it from the Internet, send it for local processing and then mail1dc1 would conclude that it should be discarded as their is no local recipient.
I have made a mailertable entry to send all mail for atlasgroup.net to mail1dc1.atlasgroup.net to save the MX lookup.
So how do I stop the local configuration error messages and drain the queue? Why are these messages even making it to sendmail in the first place? Thanks - John