An interesting question, can there is a use case where some classes of failure could be configured to be bounced earlier, if the domain does not exist is a good one, it wont ever improve, exyfingers.co will continue to not exist, no matter how long postfix waits
while most might say just don't accept mails with invalid dns names etc., typically such mail comes from a automated source, which can not handle bounces or relay failures. web forms, batch processes, etc., should at least have a noreply@ that is valid but goes to /dev/null, but ideally to a human to do error feedback or a script that reports to somebody in the sender support group if there is always a valid from (enforce it) you could be more forgiving on the to: (relax it) but only if that mail is being sent from a known source, but then how to enforce to: for other non no-reply sources. afaik we can say sender=x and recipient is valid separately OR, in sender/recipient checks and not AND can we say iff sender is a particular sender, then allow loose recipient, but otherwise enforce a tight to: recipient validation by default ? doing combined logic, in other words if the sender is nore...@website.com AND ip is a.b.c.d, THEN allow loose recipient ? in short : can we do and queries and bind several different (sender|recipient)_checks with an AND logic ? can we configure some classes of failures, such as dns, to fail more quickly ? Cheers Brett > On 15 Jun 2017, at 4:42 pm, Dominic Raferd <domi...@timedicer.co.uk> wrote: > > We occasionally get emails in our postfix queue that can never be delivered > but which are held in the queue for a week before postfix bounces them > (example: sender has typed gmail.co instead of gmail.com). I realise this > delay is the correct behaviour, but how can I - by exception - bounce a > queued mail immediately, with notification back to sender?
