On Fri, 15 Mar 2002, I. Forbes wrote:
> domain.com * F,24h,5m; F,6d,1h
> * * F,2h,5m; G,16h,2h,1.5; F,6d,8h
> It works for all mail except mail destined for the head office. That is
> bounced with a premature timeout. As per these logs (this one
> spent less than a minute on the queue - it should have been there
> for up to 6 days. The queue was empty a short while before this
> failure:
>
>
> 2002-03-15 11:40:16 16loAv-0005gG-00 ==
> headofficeuser@??? T=undef_local defer (-33): failed to
> lookup IP address for smtpsmarthost.com
Use "exinext" to see what Exim has got in its retry data for
smtpsmarthost.com. Check the logs for previous history of its attempts
to resolve that domain.
> Are DNS timeouts treated differently to other timeouts?
No. They should give rise to "routing retries".
> Is there something recording previous DNS lookup failures. If so
> when does this info get wiped?
Yes. It is in the retry data. If you use exim_dumpdb to dump the retry
data, the data for routing failures is marked with "R". (Those for
transport failures are marked with "T".) It should get wiped when a
successful routing occurs. It should also be ignored if it is too old.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
