[courier-users] More on: Connection time out for one specific site.

[Peter Davies]

More information regarding pacbell.net and courier:

doing a "dig" from my site I see the following mx records (end of email)
using telnet, none of the first 3 answer on port 25
2 timed out and one got "connection refused". The fourth one was working but slowly.

Is it possible that courier only attempts to connect with the first MX record returned?
If so, is there a way to encourage courier to try all the records until it meets with success?

thanks
peter


# dig mx pacbell.net

; <<>> DiG 9.2.1 <<>> mx pacbell.net
;; global options: printcmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 34947
;; flags: qr rd ra; QUERY: 1, ANSWER: 14, AUTHORITY: 2, ADDITIONAL: 1

;; QUESTION SECTION:
;pacbell.net. IN MX

;; ANSWER SECTION:
pacbell.net. 600 IN MX 0 pbimail1.prodigy.net.
pacbell.net. 600 IN MX 0 pbimail2.prodigy.net.
pacbell.net. 600 IN MX 0 pbimail3.prodigy.net.
pacbell.net. 600 IN MX 0 pbimail4.prodigy.net.
pacbell.net. 600 IN MX 0 pbimail5.prodigy.net.
pacbell.net. 600 IN MX 0 pbimail6.prodigy.net.
pacbell.net. 600 IN MX 0 pbimail7.prodigy.net.
pacbell.net. 600 IN MX 0 pbimail8.prodigy.net.
pacbell.net. 600 IN MX 0 pbimail9.prodigy.net.
pacbell.net. 600 IN MX 0 pbimaila.prodigy.net.
pacbell.net. 600 IN MX 0 pbimailb.prodigy.net.
pacbell.net. 600 IN MX 0 pbimailc.prodigy.net.
pacbell.net. 600 IN MX 0 pbimaild.prodigy.net.
pacbell.net. 600 IN MX 0 pbimaile.prodigy.net.

-------- Original Message --------
Subject: Connection time out for one specific site.
Date: Fri, 10 Jan 2003 18:12:18 -0800
From: Peter Davies <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]



I have recently noticed that email leaving my (otherwise very
succesful) courier server (0.38.0) destined for @pacbell.net
sustains a "Connection timed out" and the message is deferred for a
later delivery attempt. These attempts occur and some hours later the
sender receives the bad news.

This same email will succesfully be delivered if sent from a
different server (non courier but that is not to say there is not
something about my courier server that is misconfigured).
There are no known other problems and I have been running this
configuration for almost one year. pacbell.net is the only receiving
server to exhibit this problem.

As referenced in http://www.courier-mta.org/courier.html I created a
file, esmtptimeoutconnect, with the one line contents "0\n" but as I
suspect that there is something about the initial HELO message that the
pacbell server is disliking (rather than it being a very busy server)
I was not expecting any success. there was no change other than a
longer time elapsing between the "started" msg and the
timeout. Previousy, 1 minute, increased to value shown below in the maillog excerpt.

The next debugging step I would like to attempt is to
examine/trace the SMTP dialogue that is taking place between my courier server
and pacbell's receiving server - assuming that tracing/logging this dialogue is somehow
possible.

Are there any other suggestions folks might have to debug/fix this
problem. It is embarassing telling users that there is one destination domain they can't use!

below is an excerpt of one "cycle" of a failed delivery attempt:

Jan 10 17:46:23 chrony courierd: started,id=000047F0.3E1F6B00.00004A5A,from=,module=esmtp,host=pacbell.net,addr=
Jan 10 17:49:35 chrony courieresmtp: id=000047F0.3E1F6B00.00004A5A,from=,addr=: Connection timed out
Jan 10 17:49:35 chrony courieresmtp: id=000047F0.3E1F6B00.00004A5A,from=,addr=,status: deferred

thanks for any help
peter davies






-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com
_______________________________________________
courier-users mailing list
[EMAIL PROTECTED]
Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-users