I'm about to lose a client because of mail delivery problems, but the
good news is I think I now understand the problem.
In the examples below, by "myself" I mean my account on the same server
as my client, running Exim 4.24.
When I send a 1.4M attachment to myself directly from a copy of sendmail
at my home, it gets delivered fine.
When I send a 1.4M attachment to myself through a lightly loaded server
I manage in Northern California, it gets delivered fine.
When I send a 1.4M attachment to myself through my ISP server it doesn't
get delivered.
When I send a 1.4M attachment to my email account on that lightly loaded
server in Northern California, from my ISP, it gets delivered just
fine.
In addition, one of my correspondents has the same problem with the same
file. When he sends it from his connection (in Texas) to my account on
that lightly loaded server in Northern California, it gets delivered
just fine.
But when he sends it from his same connection to my account on my exim
server, he gets a bounce with an error message:
<snip>
Router: Failed to connect to SMTP host NOBALONEY.NET
because 221 da1.ns-one.net closing connection
</snip>
Any ideas?
It looks as if the problem is a timing problem, that when heavily loaded
servers take too long to send a message, exim disconnects.
Is there a way to adjust this behavior so exim will allow more time
before it disconnects? I looked in the conf file and can't find
anything except rfc1413_query_timeout, which is set for 0s. My
understanding is that 0s disables it completely. Am I wrong?
Thanks.
Jeff
--
Jeff Lasman, nobaloney.net, P. O. Box 52672, Riverside, CA 92517 US
Professional Internet Services & Support / Consulting / Colocation
Our blists address used on lists is for list email only
Phone +1 909 324-9706, or see: "
http://www.nobaloney.net/contactus.html"
