Postfix relay gives error 450 while it should be 550
- by dieter-be
Hi,
we use postfix to do relaying.
We get several messages like the following in /var/log/mail (slightly edited)
Apr 13 13:30:29 linserver postfix/smtpd[1064]: NOQUEUE: reject: RCPT from unknown[$ip]: 450 4.1.1 <[email protected]>: Recipient address rejected: undeliverable address: host domain.be [$ip] said: 550 <[email protected]>: Recipient address rejected: User unknown in virtual mailbox table (in reply to RCPT TO command); from=<[email protected]> to=<[email protected]> proto=ESMTP helo=<BLUESTREAK.domain.local>
Now, when the master mail servers gives a 550, claiming that the user does not exist, I want the relay to also give a 550 back. What happens now is that it seems to return a 450, causing clients to keep messages queued, keep trying and only notify users after a certain period has passed.
According to what I could find, the soft_bounce could cause this.
But we have not enabled this option (and by default it's off according to postfix docs)
It might also have something to do with the *_reject_code postconf values. Especially since the log message complains the unknown ip. But as you can see in the postconf output below, smtpd_sender_restrictions and smtpd_client_restrictions are empty.
So even if it would try to do any restrictions there, 550 is the "worst" error going on, so that's what I expect to be returned to the client.
postconf: http://sprunge.us/JYgB
Thanks,
Dieter
