On Tue, 17 Dec 2002, Miquel van Smoorenburg wrote:
> Using exim 3.35, I have my retry rules set up like this:
> However, mail to a certain domain (eos.ncsu.edu) stays stuck in the
> queue for basically forever:
>
> 12d 2.4K 18K2fC-0000t2-00 <freeradius-users-admin@???>
> cwu4@???
>
> 11d 6.9K 18KKkq-0002Dy-00 <freeradius-users-admin@???>
> cwu4@???
>
> .. etc. Those messages should have bounced a week ago.
Is there any useful information in the Exim log about those messages?
> # exim -v -M 18K2fC-0000t2-00
> delivering message 18K2fC-0000t2-00
> Connecting to uni00mr.unity.ncsu.edu [152.1.1.163.25] ... connected
> .. exim tries to the deliver this message *5 times* over SMTP
> in the same queue run (to all 5 MXes of eos.ncsu.edu), then decides
> that the message is not deliverable.
That is correct behaviour. After a temporary error, it will try another
MX.
> The question is, why is the queue runner not doing this? I have
> exim running as '/usr/sbin/exim -bd -q30m'.
Good question. Next time this happens, it may be worth trying a forced
"queue run" style delivery, either using -Mc instead of -M or using -q
with start and stop message ids (in this case, the same id), and
debugging turned on (-d9 for Exim 3).
However, note that Exim 3 is obsolescent now; I haven't time to chase up
anything other than really serious bugs (and I hope there aren't any of
them!).
Philip
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
