Here's an odd one. A message on the queue has got a large number of
recipients for whom delivery attempts are rejected with entries like
the following in mainlog:
1998-08-11 12:47:33 0z45fc-0000eU-00 ** postmaster@???:
unrouteable mail domain "wedstrijd-0189.demon.nl"
1998-08-11 12:47:39 0z45fc-0000eU-00 ** postmaster@???
R=lookuphost T=remote_smtp: SMTP error from remote mailer after end of data:
host punt.mail.nl.demon.net [194.159.73.25]: 554
[ No such file or directory ] Message has travelled too many hops
[lines folded for clarity] -- there are literally hundreds of these
errors every time the message is re-tried. After these, the following
lines appear in the log:
1998-08-11 12:47:39 Error while handling error message:
excessively long header line read
1998-08-11 12:47:39 0z45fc-0000eU-00 Process failed (1) when writing
error message to owner-demon-nl-announce@???
where the address given is indeed the correct address for the errors.
Unfortunately this then seems to result in the retry rules being
ignored and the message is re-tried every time the queue is processed.
I've got the full results of processing the message with -d9 but they
don't indicate what is actually going in the excessively-long header.
Any ideas, anyone?
I can easily enough delete the message from the queue and this problem
is unlikely to recur, as the message only got there in the first place
because someone tried to experiment with the Exim config file while I
was on leave, but it would be nice to know if there is an obvious
resolution. I'd rather that the message was frozen or that the re-try
rukles were obeyed even though the sub-process crashed.
Gyan.
--
*** Exim information can be found at
http://www.exim.org/ ***
