[Mimedefang] Trying to find cause of tempfails

[Cormack, Ken]

James,

Thanks for the feedback - it gives me a place to start.  However, I've
checked my INPIT_MAIL_FILTER statement, in sendmail.mc, and it is configured
as "F=T, T=S:5m;R=5m;E:5m".  I have used these values as-is, for 2 years
now, without problems.  I would agree that perhaps they might timeout if the
messages were particularly large, but we allow 50MB max size (which even
that, we seldom see), the messages in question are usually just a few KB in
size, and the attempts were made throughout the night... Historically, our
period of lightest traffic/workload.  Other emails received during this same
period, coming from different domains, are coming through just fine.  It's
only this one sending domain that is experiencing the problem I'm seeing.
Lastly, we have two MX hosts, and these errors are showing up on both the
primary and secondary hosts.

Might anyone else have other pointers as to where I should look?

Ken

-----Original Message-----
From: James Ebright [mailto:jebright at esisnet.com] 
Sent: Monday, March 07, 2005 9:36 AM
To: mimedefang at lists.roaringpenguin.com
Subject: Re: [Mimedefang] Trying to find cause of tempfails


I believe that is a generic sendmail "timeout while talking to the milter"
type return. Basically sendmail thinks it is talking to dead air for some
reason.

I would check your timeout string for the mimedefang milter in sendmail.mc,
make sure DNS resolved form the mail server for this user, etc. Perhaps this
user is trying to send a monster message that is taking MD unusually long to
scan?

Here is the default milter strings  in sendmai.mc for mimedefang and their
meanings:
# INPUT_MAIL_FILTER(`mimedefang',
`S=unix:/var/spool/MIMEDefang/mimedefang.sock, F=T, T=S:5m;R:5m')

    * F=T: instruct sendmail to send ALL pieces of mail through the milter.
Tempfail if cannot.
    * Timout options for milter, If the scripts are not set, the default
values are: T=C:5m;S:10s;R:10s;E:5m
    * C: is the connection time between Sendmail and the milter.
    * S: is the time for sending data from Sendmail to milter.
    * R: is the recieving time between Sendmail and the milter.
    * E: is the end of message time between Sendmail and milter.

Jim


On Mon, 7 Mar 2005 09:09:15 -0500 , Cormack, Ken wrote
> Group,
> 
> This morning, I've been trying to diagnose a type of tempfail I'm
repeatedly
> seeing, when a particular party tries emailing to us.  The tempfail
verbage
> is non-descript, and doesn't match any strings I've got in my filter.
> Perhaps one of you could help identify the reason these are tempfailing?
> 
> Just one such example looks like this (some address munging applied, 
> to protect the innocent)...
> 
> Mar  7 01:00:19 mail01 sendmail[31935]: j2760J2V031935:
> from=<SELEX-FACS at MUNGEDSENDER.COM>, size=26614, class=0, nrcpts=1,
> msgid=<02F117E1B24BA9F2 at munged.com>, proto=ESMTP, daemon=MTA,
> relay=gbhub.mungeddomain.com [151.147.XXX.32]
> Mar  7 01:00:21 mail01 sendmail[31935]: j2760J2V031935: Milter: data,
> reject=451 4.3.2 Please try again later
> Mar  7 01:00:21 mail01 sendmail[31935]: j2760J2V031935:
> 
> to=<cindyw at mungedrecip.com>, delay=00:00:02, pri=56614, stat=Please try
> again later
> 
> As you can see, the log entries don't say much, and none of my rules 
> (that I'm aware of) send a "451 4.3.2" tempfail, specifically.
> 
> Anyone recognize this?
> 
> Ken
> _______________________________________________
> Visit http://www.mimedefang.org and http://www.canit.ca
> MIMEDefang mailing list
> MIMEDefang at lists.roaringpenguin.com
> http://lists.roaringpenguin.com/mailman/listinfo/mimedefang


--
EsisNet.com Webmail Client

_______________________________________________
Visit http://www.mimedefang.org and http://www.canit.ca
MIMEDefang mailing list
MIMEDefang at lists.roaringpenguin.com
http://lists.roaringpenguin.com/mailman/listinfo/mimedefang