We have an smtp outbound-only service on postfix. Users on exchange are relayed out through this.
Once in awhile someone has a typo in their domain name. Postfix bounces it back to exchange with 450 defer Dec 18 15:48:40 smtp postfix-internal/smtpd[21635]: NOQUEUE: reject: RCPT from labrador2.example.com[xx.xx.xx.xx]: 450 4.1.2 <t...@easlink.ca>: Recipient address rejected: Domain not found; from=<don...@example.com> to=< t...@easlink.ca> proto=ESMTP helo=<[xx.xx.xx.xx]> We have made attempts to set the reject for bad host/domains in these settings: invalid_hostname_reject_code = 556 multi_recipient_bounce_reject_code = 550 unknown_address_reject_code = 550 unknown_hostname_reject_code = 550 unknown_local_recipient_reject_code = 550 unknown_relay_recipient_reject_code = 550 unknown_virtual_alias_reject_code = 550 unknown_virtual_mailbox_reject_code = 550 unverified_recipient_reject_code = 550 The remaining matches for 450 from running postconf | grep 450 are few: defer_code = 450 plaintext_reject_code = 450 unverified_sender_reject_code = 450 We would rather it not issue a 450 because it goes back to exchange and retries for 2 days. When someone uses a mail list from outlook, and one address has a bad domain, it causes the whole bunch of queued messages with that list to bounce back and forth between the servers. I'd rather it email a bounce to the sender when the domain is bad. What am I doing wrong?
