[exim] Retry timeout exceeded, but why exactly.

Top Page
Delete this message
Reply to this message
Author: Sander Smeenk
Date:  
To: exim-users
Subject: [exim] Retry timeout exceeded, but why exactly.
Hi!

I see mails fail to deliver with the message "retry time exceeded" but i
can't really place why the retry time was exceeded in the first place.
I think this has to do with hints databases, but what exactly can be
done to prevent it is unclear.

Scenario:
We have a (mailman) mailinglist with >20 members on a list. A message
gets sent to the list, mailman sends one message to >20 recipients.
Our smtp server accepts these messages to >20 recipients and tries to
pass them on to our mx-servers as routing dictates.
Our mx-server is set to allow max 20 recipients in a single message, all
other recipients get '452 too many recipients', should be fine, but our
smtp-servers (sometimes) immediately drop mail because of 'retry timeout
exceeded'.

This is the log of said message, sorry for the long lines. This all was
logged by the same exim-process in two seconds time:

1NZPYS-0006yl-CZ <= allemaal-bounces@??? H=(mailman.example.com) [172.25.10.1]:2579 I=[172.25.0.234]:25 P=esmtp S=23850 id=20100125140428.GC18828@???
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<kris@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<michelm@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<emil@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<michel@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<henkjan@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<henk@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<sanders@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<cris@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<bastiaan@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<marks@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ SMTP error from remote mail server after RCPT TO:<bartv@???>: host mx1.example.com [172.25.0.237]: 452 too many recipients
1NZPYS-0006yl-CZ mx3.example.com [172.25.0.235] Connection refused
1NZPYS-0006yl-CZ => peter@??? R=dnslookup T=remote_smtp S=25049 H=mx1.example.com [172.25.0.237] C="250 OK id=1NZPYS-0006VP-My" QT=0s DT=0s
[ .. 18 C="250 OK id=..." msgs removed, mx accepts 20 .. ]
1NZPYS-0006yl-CZ -> bas@??? R=dnslookup T=remote_smtp S=25049 H=mx1.example.com [172.25.0.237] C="250 OK id=1NZPYS-0006VP-My" QT=0s DT=0s
1NZPYS-0006yl-CZ == kris@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == michelm@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == emil@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == michel@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == henkjan@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == henk@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == sanders@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == cris@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == bastiaan@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == marks@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ == bartv@??? R=dnslookup T=remote_smtp defer (111): Connection refused
1NZPYS-0006yl-CZ ** bartv@???: retry timeout exceeded
1NZPYS-0006yl-CZ ** marks@???: retry timeout exceeded
1NZPYS-0006yl-CZ ** bastiaan@???: retry timeout exceeded
1NZPYS-0006yl-CZ ** cris@???: retry timeout exceeded
1NZPYS-0006yl-CZ ** sanders@???: retry timeout exceeded
1NZPYS-0006yl-CZ ** henk@???: retry timeout exceeded
1NZPYS-0006yl-CZ ** henkjan@???: retry timeout exceeded
1NZPYS-0006yl-CZ ** michel@???: retry timeout exceeded
1NZPYS-0006yl-CZ ** michelm@???: retry timeout exceeded
1NZPYS-0006yp-Pr <= <> R=1NZPYS-0006yl-CZ U=exim P=local S=4721

The bounce message just says what the log says, retry time exceeded.

Not -all- failed recipients had been discarded, i logged in to this
server and manually flushed 1NZPYS-0006yl-CZ which did retry delivery
for two recipients which at first had failed to deliver correctly:

1NZPYS-0006yl-CZ => kris@??? R=dnslookup T=remote_smtp S=25049 H=mx1.example.com [172.25.0.237] C="250 OK id=1NZPpE-0006rt-4b" QT=17m20s DT=0s
1NZPYS-0006yl-CZ -> emil@??? R=dnslookup T=remote_smtp S=25049 H=mx1.example.com [172.25.0.237] C="250 OK id=1NZPpE-0006rt-4b" QT=17m20s DT=0s
1NZPYS-0006yl-CZ Completed

What could cause this retry timeout exceeded message? Why did two of the
recipients 'survive' the timeout? Exim didn't even retry any of the messages
before logging the timeout was 'exceeded'.

We have one general retry rule:
   *     *    F,4h,5m; F,16h,30m; F,7d,3h


As the QT= values indicate, none of these messages have been in queue
longer than 7d.

I 'fixed' this by setting max_rcpt on the smtp-servers, but rather had
it fixed at the origin of the problem. I just can't figure out what the
origin is.

Any hints appreciated!

With regards,
-Sander.
--
| 42.7 percent of all statistics are made up on the spot.
| 4096R/20CC6CD2 - 6D40 1A20 B9AA 87D4 84C7 FBD6 F3A9 9442 20CC 6CD2