gordanbobich.net
Date: Thu Nov 01 2007 - 12:52:03 CDT

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Thu, 1 Nov 2007, Wietse Venema wrote:

>>>> Nov 1 07:15:33 smtp-out-b postfix/cleanup[22462]: 5C6037A2683:
>>>> message-id=<!&!AAAAAAAAAAAYAAAAAAAAAMZwGI7IkCJPiEIFgLPzqtnCgAAAEAAAA
>>>> FTJx60TftpBkIoHmlwF8GsBAAAAAA==zoom.co.uk>
>>>
>>>> Nov 1 07:15:33 smtp-out-b postfix/qmgr[20307]: 5C6037A2683:
>>>> from=<foozoom.co.uk>, size=2903, nrcpt=1 (queue active)
>>>
>>>> Nov 1 07:16:52 smtp-out-b postfix/smtp[22405]: 5C6037A2683:
>>>> to=<bargriffin.com>, relay=none, delay=81, status=deferred
>>>> (connect to avt15.mail.griffin.com[85.189.39.55]: No route to host)
>>>
>>> This shows only one connection attempt, either you don't have multiple
>>> MX hosts, or you log search only looks for log entries that match the
>>> queue id of the message, and omits other log entries that show failed
>>> connection attempts with no queue id.
>>
>> It didn't occur to me that there would be connection attempts listed
>> without the queue id of the message the connects were attempted for. How
>> would one, in that case, relate a connection to a message?
>
> Searching all the postfix/smtp messages for PID 22405 would get
> you a pretty good idea of the MX connect strategy.

That pinned it down. Thanks. It looks like something else went on the
blink when the message was rejected. The MX-es were tried in the correct
order. Postfix was not the cause, it was an unrelated glitch on the
network.

Thank you all for helping me troubleshoot this. On a separate note,
Postfix doesn't try just the top 5 MX-es as was implied in an earlier
post. It tries all of them at least up to 17. ;-)

As to why it kept re-trying the 17th 6 times in 5 hours rather than trying
again from the top is a separate question, but I guess that squeezes past
the RFC under the "retry in due time" part. ;-)

Thanks.

Gordan

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]