From: Wietse Venema (wietseporcupine.org)
Date: Tue Aug 03 2004 - 15:39:27 CDT

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Hamie:
> >Postfix 2.1 connects to a backup MX after a temporary failure (not just
> >connection problem) with the primary MX, only the final status is logged
> >for each recipient. Temporary mail transaction problems with a non-final
> >MX are not logged.
> >
> >
>
> Right. So no logging of the non-fatal connect to the primary... But at
> time of writing, the primary WAS up... The secondary wasn't... This is
> the whole (Munged to prevent scraping) log for the relevant QID's...

The failure logging depends on the failure type: before or after
connecting. For example, a failed attempt to connect to the primary
is logged but without the queue ID string, so it may escape attention.
It would probably be better if the queue ID had been logged.

Aug 3 16:32:48 spike postfix/pickup[66926]: 64B30BC0D9: uid=1001 from=<wietse>
Aug 3 16:32:48 spike postfix/cleanup[68505]: 64B30BC0D9: message-id=<20040803203248.64B30BC0D9spike.porcupine.org>
Aug 3 16:32:48 spike postfix/qmgr[54801]: 64B30BC0D9: from=<wietseporcupine.org>, size=330, nrcpt=1 (queue active)
Aug 3 16:33:18 spike postfix/smtp[68461]: connect to fist.porcupine.org[168.100.189.4]: Operation timed out (port 25)
Aug 3 16:33:18 spike postfix/smtp[68461]: 64B30BC0D9: to=<wietsefist.porcupine.org>, relay=none, delay=30, status=deferred (connect to vm.porcupine.org[168.100.189.16]: Connection refused)

        Wietse

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]