Wakko Warner wrote:
> Graeme Fowler wrote:
>> OK, so let's have a look at your log entry...
>>
>>> 2010-05-13 11:08:03 1OCZtL-000Iko-E9 H=m190.salsalabs.net [69.174.83.190] F=<org2@???> temporarily rejected after DATA
>> "temporarily".
>>
>> And "after DATA".
>>
>> These are really important terms. Your server, for some reason, rejected
>> the message with a 4xx response after it received the DATA part of the
>> transaction; without other detail we don't know why. If the sender
>> didn't honour the 4xx response, then they are not correctly configured
>> *or* your machine subsequently sent a 5xx (permanent) response after the
>> detail you gave us.
>
> I'd like to add what I've seen on a small server using slow dialup:
>
> I've noticed some strange things happening when the network line is loaded.
> Here's my last temporarily rejected entry for the DATA phase.
> 2010-04-30 01:30:39 1O7ilc-0002eg-T3 H=tahini.csx.cam.ac.uk [131.111.8.192]
> I=[192.168.2.1]:25 F=<exim-users-bounces+animx.eu.org@???>
> temporarily rejected after DATA
>
> # grep "^acl_smtp_data" /etc/exim4/exim4.conf
> acl_smtp_data = /etc/exim4/data.acl
> # grep warn /etc/exim4/data.acl | wc -l
> 0
> #
>
> In other words, I have no stanzas in my config that have warn in the data
> phase, but still the message was temp reject. I have no clue why it does
> this. It seems to only happen when the connection's usage is at 100%. The
> session duration appears to be about 6.5 minutes long. The connection was
> closed with "QUIT".
>
> It would be nice to know why it did this, but it doesn't happen that often.
>
Most curious. I have *very* complex acl clauses in the DATA section.
Yet the only such messages I have are from the same hour of the same day a year
ago when I was mucking about with an acl clause that 'failed to expand...' due
to mis-matched brackets.
Any chance that your Exim - or that of the OP - is temporarily unable to write
the message into the queue storage area due to a clash of privs,
in-use-by-other, some maintenance pass being run, or lack of raw space?
Bill
