On Wed, 2006-06-07 at 14:25 +0100, Philip Hazel wrote:
> On Wed, 7 Jun 2006, David Woodhouse wrote:
>
> > Do we store the ultimate timeout in the retry hints database when we
> > first fail to deliver the mail, then bounce it later even if the
> > ultimate timeout in configuration file has been changed?
>
> No. All that Exim stores is time of first failure, time of last try,
> time before which it shouldn't try again, and an "expired" bit. In
> other words, exactly what you get out from exim_dumpdb.
Hm. That was just a guess (evidently incorrect) at what might be going
wrong. A debugging run is at
http://david.woodhou.se/retry.txt
Do you have any better ideas as to why this might be ignoring my retry
configuration? I really shouldn't be seeing this after I changed my
retry configuration as described in my previous mail, should I?
LOG: MAIN
** mike.evans@??? R=lookuphost T=remote_smtp: retry time not reached for any host after a long failure period
Further testing will be hindered by the fact that I've now removed
/var/spool/exim/db/retry*, so it's gone back to queueing the messages it
should be queueing instead of rejecting/bouncing them.
--
dwmw2
