I have been having a most frustrating problem since I upgraded to RT
3.x. From time to time an incoming email will loop showing the
following error in the mail Q:
-Queue ID- --Size-- ----Arrival Time---- -Sender/Recipient-------
6112B34D7B 177915 Thu Oct 2 23:15:23 user@domain.com
(temporary failure. Command output: An Error Occurred =================
500 Internal Server Error ) jobs@company.com
If this is a new request it will end up creating new tickets over and
over again. If it is a reply or comment to an existing ticket it will
just add to that ticket over and over again until the message is purged.
I am not sure what “500 Internal Server Error” refers to. Wether it is
reported by “rt-mailgate” or by the web server because of “–url http://rt/”
Why is “–url http://rt/” used since we only want to pipe incoming
emails to a Q ?
My environment is:
OS: FreeBSD 4.7-RELEASE
Perl: This is perl, v5.8.0 built for i386-freebsd
MySQL: Ver 3.23.55 for portbld-freebsd4.7 on i386
RT: 3.0.2 (mysql - apache 1.3.27 - fastcgi)
Mail: postfix-2.0.4,1
and my alias entry is:
support: “|/usr/local/rt3/bin/rt-mailgate --queue Support --action
comment --url http://rt/”
Any ideas on this issue would be much appreciated.
You need to understand that rt-mailgate talks to RT via the web
interface after it receives and parses the mail from sendmail (or
whatever other MTA you might be using).
The URL on the command line must point to your functional RT instance.
I’m actually pretty surprised that it gets as far as getting anything
into the database.
(Jesse, is that a bug/design flaw?)
In any case, I suspect this behaviour will go away if you fix the url.
I have been having a most frustrating problem since I upgraded to RT
3.x. From time to time an incoming email will loop showing the
following error in the mail Q:
–
-Queue ID- --Size-- ----Arrival Time---- -Sender/Recipient-------
6112B34D7B 177915 Thu Oct 2 23:15:23 user@domain.com
(temporary failure. Command output: An Error Occurred
================= 500 Internal Server Error ) jobs@company.com
If this is a new request it will end up creating new tickets over and
over again. If it is a reply or comment to an existing ticket it will
just add to that ticket over and over again until the message is purged.
I am not sure what “500 Internal Server Error” refers to. Wether it is
reported by “rt-mailgate” or by the web server because of “–url http://rt/”
pipe mail in CLI through alias string with --debug and report output
Why is “–url http://rt/” used since we only want to pipe incoming
emails to a Q ?
Thanks for the info… This setup actually works very well for 95% of
the emails sent to RT. Only a few rare emails get stuck that way…
If it is a timeout from the web server (too slow or busy) why would it
timeout every time the mailq tries to deliver the message ? If most
emails go through then the second or third time the message is
attempted to be delivered by the mailq, it should be accepted but does
not. Once an email gets stuck it is never processed successfully…
Thanks
ThomasOn Friday, October 3, 2003, at 10:21 AM, Jim Rowan wrote:
You need to understand that rt-mailgate talks to RT via the web
interface after it receives and parses the mail from sendmail (or
whatever other MTA you might be using).
The URL on the command line must point to your functional RT instance.
I’m actually pretty surprised that it gets as far as getting anything
into the database.
(Jesse, is that a bug/design flaw?)
In any case, I suspect this behaviour will go away if you fix the url.