Sendmail problems (EXIT_TEMPFAIL) with rt3 mailgate

RT Users
1

Hi,

  I have searched high and low in an effort to find a solution to

this problem, to no avail. After migrating our RT2 data into the RT3
database and doing the necessary config (see below), our mail logs
record the following result when I try to create a new ticket via email
in rt3 :

Jun 12 23:16:28 xx-xxx sendmail[21357]: h5CDGSU3021357:
from=a.watts@xxx.xxx.xx, size=48019, class=0, nrcpts=1,
msgid=C73A42308A9E56438DEDD76DB6FFC90A5B46C1@locxxx.xxx.xxx.xx,
proto=ESMTP, daemon=MTA, relay=mx2xxxxx.xxx.xxx.xx [138.77.5.126]
Jun 12 23:16:29 xxxxxxxx sendmail[21358]: h5CDGSU3021357:
to=“|/usr/local/rt3/bin/rt-mailgate --queue Helpdesk -action correspond
–url=http://my.rtserver/rt3”, ctladdr=icom-helpdesk@xxx.xxx.xxx
(8/0), delay=00:00:01, xdelay=00:00:01, mailer=prog, pri=78235,
dsn=4.0.0, stat=Deferred: prog mailer (/usr/sbin/smrsh) exited with
EX_TEMPFAIL

 OS is RH 8.0, running 'Apache/1.3.27 (Unix) mod_perl/1.27', MySQL

3.23.53, sendmail 8.12.5-7.

 1. /etc/aliases :
  icom-helpdesk: "|/usr/local/rt3/bin/rt-mailgate --queue general

-action correspond --url=http://my.rtserver/rt3";

  I have used the 'newaliases' command to rebuild sendmail's

database, and the mail log confirms this.

 2. /etc/smrsh :

[root@xx-xxx smrsh]# ls -l rt-mailgate
lrwxrwxrwx 1 root root 30 Jun 13 07:50 rt-mailgate →
/usr/local/rt3/bin/rt-mailgate

 3. httpd.conf :

Alias /rt3 /usr/local/rt3/share/html
Alias /NoAuth /usr/local/rt3/share/html/NoAuth
AddDefaultCharset UTF-8

PerlModule Apache::DBI
PerlRequire /usr/local/rt3/bin/webmux.pl

<Location /rt3>
SetHandler perl-script
PerlHandler RT::Mason

 4. Perms :
Everything under /usr/local/rt3 is owned by 'root:webserveruser',

where ‘webserveruser’ is the name of the Apache ‘User’ variable.
Everything in the web interface is fine, it’s the mail gateway which
seems to be my only stumbling block.

 The sendmail doco says something to the effect that 'EX_TEMPFAIL'

means that it is unable to deliver the mail, but will queue it for
delivery at a later date. As soon as I switch everything back to RT2,
mail begins coming in again, and the queued messages are delivered 30-50
minutes after that point. I suspect that it’s a perms issue with the
receiving mailbox, but the same user runs Apache under rt3 as rt2, and
rt2 is working fine.

  I am not a sendmail whiz, so any suggestions are most welcome.

Alternatively, tips on how I can use strace to get a better idea of what
file sendmail is trying to open would be handy - its currently just
returning the above diag without giving me much idea of what is going
on.

Thanks,

Arthur

Arthur Watts (a.watts@cqu.edu.au)

Web Developer

Faculty of Informatics and Communication | Central Queensland University
| Rockhampton, Australia

2

This sounds exactly the same as the problem I am having. Do the message(s)
causing the problem have attachments?On Fri, 13 Jun 2003, Arthur Watts wrote:

Date: Fri, 13 Jun 2003 09:56:19 +1000
From: Arthur Watts a.watts@cqu.edu.au
To: rt-users@lists.fsck.com
Subject: [rt-users] Sendmail problems (EXIT_TEMPFAIL) with rt3 mailgate

Hi,

  I have searched high and low in an effort to find a solution to

this problem, to no avail. After migrating our RT2 data into the RT3
database and doing the necessary config (see below), our mail logs
record the following result when I try to create a new ticket via email
in rt3 :

Jun 12 23:16:28 xx-xxx sendmail[21357]: h5CDGSU3021357:
from=a.watts@xxx.xxx.xx, size=48019, class=0, nrcpts=1,
msgid=C73A42308A9E56438DEDD76DB6FFC90A5B46C1@locxxx.xxx.xxx.xx,
proto=ESMTP, daemon=MTA, relay=mx2xxxxx.xxx.xxx.xx [138.77.5.126]
Jun 12 23:16:29 xxxxxxxx sendmail[21358]: h5CDGSU3021357:
to=“|/usr/local/rt3/bin/rt-mailgate --queue Helpdesk -action correspond
–url=http://my.rtserver/rt3”, ctladdr=icom-helpdesk@xxx.xxx.xxx
(8/0), delay=00:00:01, xdelay=00:00:01, mailer=prog, pri=78235,
dsn=4.0.0, stat=Deferred: prog mailer (/usr/sbin/smrsh) exited with
EX_TEMPFAIL

 OS is RH 8.0, running 'Apache/1.3.27 (Unix) mod_perl/1.27', MySQL

3.23.53, sendmail 8.12.5-7.

 1. /etc/aliases :
  icom-helpdesk: "|/usr/local/rt3/bin/rt-mailgate --queue general

-action correspond --url=http://my.rtserver/rt3";

  I have used the 'newaliases' command to rebuild sendmail's

database, and the mail log confirms this.

 2. /etc/smrsh :

[root@xx-xxx smrsh]# ls -l rt-mailgate
lrwxrwxrwx 1 root root 30 Jun 13 07:50 rt-mailgate →
/usr/local/rt3/bin/rt-mailgate

 3. httpd.conf :

Alias /rt3 /usr/local/rt3/share/html
Alias /NoAuth /usr/local/rt3/share/html/NoAuth
AddDefaultCharset UTF-8

PerlModule Apache::DBI
PerlRequire /usr/local/rt3/bin/webmux.pl

<Location /rt3>
SetHandler perl-script
PerlHandler RT::Mason

 4. Perms :
Everything under /usr/local/rt3 is owned by 'root:webserveruser',

where ‘webserveruser’ is the name of the Apache ‘User’ variable.
Everything in the web interface is fine, it’s the mail gateway which
seems to be my only stumbling block.

 The sendmail doco says something to the effect that 'EX_TEMPFAIL'

means that it is unable to deliver the mail, but will queue it for
delivery at a later date. As soon as I switch everything back to RT2,
mail begins coming in again, and the queued messages are delivered 30-50
minutes after that point. I suspect that it’s a perms issue with the
receiving mailbox, but the same user runs Apache under rt3 as rt2, and
rt2 is working fine.

  I am not a sendmail whiz, so any suggestions are most welcome.

Alternatively, tips on how I can use strace to get a better idea of what
file sendmail is trying to open would be handy - its currently just
returning the above diag without giving me much idea of what is going
on.

Thanks,

Arthur

Arthur Watts (a.watts@cqu.edu.au)

Web Developer

Faculty of Informatics and Communication | Central Queensland University
| Rockhampton, Australia

rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

Have you read the FAQ? The RT FAQ Manager lives at http://fsck.com/rtfm

3

Hi Stewart,

  Thanks for the prompt reply.

  No - I had a problem with attachments previously, but even 'echo

“Testing” | mail icom-wemaster’ fails when I move to the rt3 config in
/etc/smrsh. Same result every time - I’ve checked that sendmail knows
what to do with the mail for that user, but still no luck.

Cheers,

ArtieFrom: Stewart James [mailto:stewart.james@vu.edu.au]
Sent: Friday, 13 June 2003 10:17
To: Arthur Watts
Cc: rt-users@lists.fsck.com
Subject: Re: [rt-users] Sendmail problems (EXIT_TEMPFAIL) with rt3
mailgate

This sounds exactly the same as the problem I am having. Do the
message(s) causing the problem have attachments?

4

Maybe the problems we are having are linked.

However, I only seem to have issues with some email, not all - although
the errors we are getting seem to be similar. Although I am not getting
anything useful out of the rt logs to indicate what is actually happening
inside RT.

I am really hoping someone on the list knows more about the inside of RT
to help solve this.

StewartOn Fri, 13 Jun 2003, Arthur Watts wrote:

Date: Fri, 13 Jun 2003 10:27:17 +1000
From: Arthur Watts a.watts@cqu.edu.au
To: Stewart James stewart.james@vu.edu.au
Cc: rt-users@lists.fsck.com
Subject: RE: [rt-users] Sendmail problems (EXIT_TEMPFAIL) with rt3
mailgate

Hi Stewart,

  Thanks for the prompt reply.

  No - I had a problem with attachments previously, but even 'echo

“Testing” | mail icom-wemaster’ fails when I move to the rt3 config in
/etc/smrsh. Same result every time - I’ve checked that sendmail knows
what to do with the mail for that user, but still no luck.

Cheers,

Artie

-----Original Message-----
From: Stewart James [mailto:stewart.james@vu.edu.au]
Sent: Friday, 13 June 2003 10:17
To: Arthur Watts
Cc: rt-users@lists.fsck.com
Subject: Re: [rt-users] Sendmail problems (EXIT_TEMPFAIL) with rt3
mailgate

This sounds exactly the same as the problem I am having. Do the
message(s) causing the problem have attachments?

On Fri, 13 Jun 2003, Arthur Watts wrote:

Date: Fri, 13 Jun 2003 09:56:19 +1000
From: Arthur Watts a.watts@cqu.edu.au
To: rt-users@lists.fsck.com
Subject: [rt-users] Sendmail problems (EXIT_TEMPFAIL) with rt3
mailgate

Hi,

  I have searched high and low in an effort to find a solution to

this problem, to no avail. After migrating our RT2 data into the RT3
database and doing the necessary config (see below), our mail logs
record the following result when I try to create a new ticket via
email in rt3 :

Jun 12 23:16:28 xx-xxx sendmail[21357]: h5CDGSU3021357:
from=a.watts@xxx.xxx.xx, size=48019, class=0, nrcpts=1,
msgid=C73A42308A9E56438DEDD76DB6FFC90A5B46C1@locxxx.xxx.xxx.xx,
proto=ESMTP, daemon=MTA, relay=mx2xxxxx.xxx.xxx.xx [138.77.5.126] Jun
12 23:16:29 xxxxxxxx sendmail[21358]: h5CDGSU3021357:
to=“|/usr/local/rt3/bin/rt-mailgate --queue Helpdesk -action
correspond --url=http://my.rtserver/rt3”,
ctladdr=icom-helpdesk@xxx.xxx.xxx
(8/0), delay=00:00:01, xdelay=00:00:01, mailer=prog, pri=78235,
dsn=4.0.0, stat=Deferred: prog mailer (/usr/sbin/smrsh) exited with
EX_TEMPFAIL

 OS is RH 8.0, running 'Apache/1.3.27 (Unix) mod_perl/1.27', MySQL

3.23.53, sendmail 8.12.5-7.

 1. /etc/aliases :
  icom-helpdesk: "|/usr/local/rt3/bin/rt-mailgate --queue general

-action correspond --url=http://my.rtserver/rt3";

  I have used the 'newaliases' command to rebuild sendmail's

database, and the mail log confirms this.

 2. /etc/smrsh :

[root@xx-xxx smrsh]# ls -l rt-mailgate
lrwxrwxrwx 1 root root 30 Jun 13 07:50 rt-mailgate →
/usr/local/rt3/bin/rt-mailgate

 3. httpd.conf :

Alias /rt3 /usr/local/rt3/share/html
Alias /NoAuth /usr/local/rt3/share/html/NoAuth
AddDefaultCharset UTF-8

PerlModule Apache::DBI
PerlRequire /usr/local/rt3/bin/webmux.pl

<Location /rt3>
SetHandler perl-script
PerlHandler RT::Mason

 4. Perms :
Everything under /usr/local/rt3 is owned by 'root:webserveruser',

where ‘webserveruser’ is the name of the Apache ‘User’ variable.
Everything in the web interface is fine, it’s the mail gateway which

seems to be my only stumbling block.

 The sendmail doco says something to the effect that 'EX_TEMPFAIL'

means that it is unable to deliver the mail, but will queue it for
delivery at a later date. As soon as I switch everything back to RT2,
mail begins coming in again, and the queued messages are delivered
30-50 minutes after that point. I suspect that it’s a perms issue with

the receiving mailbox, but the same user runs Apache under rt3 as rt2,

and rt2 is working fine.

  I am not a sendmail whiz, so any suggestions are most welcome.

Alternatively, tips on how I can use strace to get a better idea of
what file sendmail is trying to open would be handy - its currently
just returning the above diag without giving me much idea of what is
going on.

Thanks,

Arthur

Arthur Watts (a.watts@cqu.edu.au)

Web Developer

Faculty of Informatics and Communication | Central Queensland
University
| Rockhampton, Australia

rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

Have you read the FAQ? The RT FAQ Manager lives at
http://fsck.com/rtfm

5

Arthur Watts wrote:

dsn=4.0.0, stat=Deferred: prog mailer (/usr/sbin/smrsh) exited with
EX_TEMPFAIL

So, what happens if you pipe the offending email directly
into rt-mailgate from the commandline?
Phil Homewood, Systems Janitor, http://www.SnapGear.com
pdh@snapgear.com Ph: +61 7 3435 2810 Fx: +61 7 3891 3630
SnapGear - Custom Embedded Solutions and Security Appliances

6

Doing ti with the problem I am getting returns:

An Error Occurred

500 read timeout

I also tarted top and noticed that one of the apache processes start
chewing 99% of CPU.

The email is inserted into RT, and apart from this error, everything seems
to function fine.

There was nothing in the apache logs, mysql logs to indicate the error
either.

StewartOn Fri, 13 Jun 2003, Phil Homewood wrote:

Date: Fri, 13 Jun 2003 10:39:39 +1000
From: Phil Homewood pdh@snapgear.com
To: rt-users@lists.fsck.com
Subject: Re: [rt-users] Sendmail problems (EXIT_TEMPFAIL) with rt3
mailgate

Arthur Watts wrote:

dsn=4.0.0, stat=Deferred: prog mailer (/usr/sbin/smrsh) exited with
EX_TEMPFAIL

So, what happens if you pipe the offending email directly
into rt-mailgate from the commandline?

7

I am still trying to figure this out.

With the sample message I am using this is what I am seeing:

Submit message to RT via rt-mailgate.
RT Acts on the email fine - generates the the messages that the scrips
define etc.

Around the 3min mark rt-mailgate returns with a 500 timeout erro
(apparently from the web server). However, the webserver (using 99% of
cpu) continue using 99% of cpu.

At around the 5 minute mark, apahche stops using 99% of cpu, and the final
log message is made…ticket created.

How do I increase the time that rt-mailgate will wait for a response from
the web server…not the best solution, but considering I am not running
the preferred perl and mysql versions I can see difficulties with figuring
this out.

Does this extra info spur solutions on from anyone?

Cheers,

StewartOn Fri, 13 Jun 2003, Stewart James wrote:

Date: Fri, 13 Jun 2003 11:53:56 +1000 (EST)
From: Stewart James stewart.james@vu.edu.au
To: rt-users@lists.fsck.com
Subject: Re: [rt-users] Sendmail problems (EXIT_TEMPFAIL) with rt3
mailgate

Doing ti with the problem I am getting returns:

An Error Occurred

500 read timeout

I also tarted top and noticed that one of the apache processes start
chewing 99% of CPU.

The email is inserted into RT, and apart from this error, everything seems
to function fine.

There was nothing in the apache logs, mysql logs to indicate the error
either.

Stewart

On Fri, 13 Jun 2003, Phil Homewood wrote:

Date: Fri, 13 Jun 2003 10:39:39 +1000
From: Phil Homewood pdh@snapgear.com
To: rt-users@lists.fsck.com
Subject: Re: [rt-users] Sendmail problems (EXIT_TEMPFAIL) with rt3
mailgate

Arthur Watts wrote:

dsn=4.0.0, stat=Deferred: prog mailer (/usr/sbin/smrsh) exited with
EX_TEMPFAIL

So, what happens if you pipe the offending email directly
into rt-mailgate from the commandline?

rt-users mailing list
rt-users@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-users

Have you read the FAQ? The RT FAQ Manager lives at http://fsck.com/rtfm