RT migration troubles (mail gateway)

I’ve been running RT in a testing capacity on my desktop machine for several
months. Now we’re installing it for real on a different server.

I have (with help from our local sysadmin guru) got RT installed and all its
dependencies satisfied. I have even moved our old data over from my machine
(which was wayyyy easier than I thought it would be). And I can now use RT
just fine on the new machine if I do it through a web browser.

But we’re having trouble getting the mail gateway to work. My desktop
machine is Mandrake 8.1, which by default uses postfix as its MTA, and the
mail gateway pretty much Just Worked out of the box. But the new machine is
Red Hat 7.2, which uses sendmail by default. I copied over my RT-specific
lines in /etc/aliases and ran newaliases, and I can tell that the aliases
are picking up incoming mail OK. But when they try to turn around and run
rt-mailgate, this fails.

The mail I send in to one of my aliases bounces with this message:

----- The following addresses had permanent fatal errors -----
"|/usr/local/rt2/bin/rt-mailgate --queue general --action correspond"
(reason: service unavailable)
(expanded from: rt-general@nairobi2.garmin.com)

----- Transcript of session follows -----
smrsh: rt-mailgate not available for sendmail programs
554 5.0.0 “|/usr/local/rt2/bin/rt-mailgate --queue general --action
correspond”… Service unavailable

I wasn’t sure what that meant, so I tried running rt-mailgate by hand
(admittedly, not really knowing how to type in a real SMTP message by hand):

[root@nairobi2 bin]# echo test| mail -s testing
rt-general@nairobi2.garmin.com |/usr/local/rt2/bin/rt-mailgate --queue
general --action correspond
Can’t locate RT/Action/.pm in @INC (@INC contains: /etc/rt2
/usr/local/rt2/lib /usr/lib/perl5/5.6.1/i686-linux /usr/lib/perl5/5.6.1
/usr/lib/perl5/site_perl/5.6.1/i686-linux /usr/lib/perl5/site_perl/5.6.1
/usr/lib/perl5/site_perl .) at (eval 52) line 3.
Require of RT::Action:: failed.
Can’t locate RT/Action/.pm in @INC (@INC contains: /etc/rt2
/usr/local/rt2/lib /usr/lib/perl5/5.6.1/i686-linux /usr/lib/perl5/5.6.1
/usr/lib/perl5/site_perl/5.6.1/i686-linux /usr/lib/perl5/site_perl/5.6.1
/usr/lib/perl5/site_perl .) at (eval 52) line 3.

[root@nairobi2 bin]#

Now this looks a bit fishy. Can’t locate RT/Action/.pm? Shouldn’t there be
something between Action/ and .pm? Is there something still missing in our
setup, or is this just because of something weird with me trying to run it
by hand? Or is there something special we need to do to sendmail? I have
determined that the message “Require of RT::Action:: failed” is generated by
/usr/local/rt2/lib/RT/ScripAction.pm just in case that helps turn on a light
bulb for anyone.

Here’s hoping someone finds this somewhat familiar…

Kendric Beachey

http://fsck.com/rtfm/factoid.html?id=111 should tell you what you need to know.

-jOn Mon, Jan 07, 2002 at 09:08:43AM -0600, Beachey, Kendric wrote:

I’ve been running RT in a testing capacity on my desktop machine for several
months. Now we’re installing it for real on a different server.

I have (with help from our local sysadmin guru) got RT installed and all its
dependencies satisfied. I have even moved our old data over from my machine
(which was wayyyy easier than I thought it would be). And I can now use RT
just fine on the new machine if I do it through a web browser.

But we’re having trouble getting the mail gateway to work. My desktop
machine is Mandrake 8.1, which by default uses postfix as its MTA, and the
mail gateway pretty much Just Worked out of the box. But the new machine is
Red Hat 7.2, which uses sendmail by default. I copied over my RT-specific
lines in /etc/aliases and ran newaliases, and I can tell that the aliases
are picking up incoming mail OK. But when they try to turn around and run
rt-mailgate, this fails.

The mail I send in to one of my aliases bounces with this message:

----- The following addresses had permanent fatal errors -----
"|/usr/local/rt2/bin/rt-mailgate --queue general --action correspond"
(reason: service unavailable)
(expanded from: rt-general@nairobi2.garmin.com)

----- Transcript of session follows -----
smrsh: rt-mailgate not available for sendmail programs
554 5.0.0 “|/usr/local/rt2/bin/rt-mailgate --queue general --action
correspond”… Service unavailable

I wasn’t sure what that meant, so I tried running rt-mailgate by hand
(admittedly, not really knowing how to type in a real SMTP message by hand):

[root@nairobi2 bin]# echo test| mail -s testing
rt-general@nairobi2.garmin.com |/usr/local/rt2/bin/rt-mailgate --queue
general --action correspond
Can’t locate RT/Action/.pm in @INC (@INC contains: /etc/rt2
/usr/local/rt2/lib /usr/lib/perl5/5.6.1/i686-linux /usr/lib/perl5/5.6.1
/usr/lib/perl5/site_perl/5.6.1/i686-linux /usr/lib/perl5/site_perl/5.6.1
/usr/lib/perl5/site_perl .) at (eval 52) line 3.
Require of RT::Action:: failed.
Can’t locate RT/Action/.pm in @INC (@INC contains: /etc/rt2
/usr/local/rt2/lib /usr/lib/perl5/5.6.1/i686-linux /usr/lib/perl5/5.6.1
/usr/lib/perl5/site_perl/5.6.1/i686-linux /usr/lib/perl5/site_perl/5.6.1
/usr/lib/perl5/site_perl .) at (eval 52) line 3.

[root@nairobi2 bin]#

Now this looks a bit fishy. Can’t locate RT/Action/.pm? Shouldn’t there be
something between Action/ and .pm? Is there something still missing in our
setup, or is this just because of something weird with me trying to run it
by hand? Or is there something special we need to do to sendmail? I have
determined that the message “Require of RT::Action:: failed” is generated by
/usr/local/rt2/lib/RT/ScripAction.pm just in case that helps turn on a light
bulb for anyone.

Here’s hoping someone finds this somewhat familiar…


Kendric Beachey


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

http://www.bestpractical.com/products/rt – Trouble Ticketing. Free.

From: Jesse Vincent [mailto:jesse@bestpractical.com]

http://fsck.com/rtfm/factoid.html?id=111 should tell you what
you need to know.

It was indeed the smrsh. Thanks, everybody.

BTW: after I fixed that, it died saying it couldn’t write to /var/log/rt2.
So I made it world-writeable. This machine will not be on the internet,
just our intranet…does anyone know of any good reasons why this kluge is
unsafe?

Hi,

I think the first thing to verify is that /usr/local/rt2/bin/rt-mailgate is
symlinked to /etc/smrsh/rt-mailgate and that it has the correct
permissions (setgid bit set?) This is a common problem though I
didn’t find any reference to it when digging around the FAQ.

The long story is that Sendmail’s security model requires mail
gateways, etc, to be symlinked under /etc/smrsh (smrsh =
sendmail restricted shell) and have particular permissions; Postfix
avoids this by running as a non-privileged user [insert
Sendmail/Postfix security model flamewar here.]

I’m tempted to suggest ripping out Sendmail and replacing it with
Postfix on the Red Hat box, but that may not be a welcome
option… :slight_smile:

– BobOn 7 Jan 2002, at 9:08, Beachey, Kendric wrote:

I’ve been running RT in a testing capacity on my desktop machine for
several months. Now we’re installing it for real on a different
server.

I have (with help from our local sysadmin guru) got RT installed and
all its dependencies satisfied. I have even moved our old data over
from my machine (which was wayyyy easier than I thought it would be).
And I can now use RT just fine on the new machine if I do it through a
web browser.

But we’re having trouble getting the mail gateway to work. My desktop
machine is Mandrake 8.1, which by default uses postfix as its MTA, and
the mail gateway pretty much Just Worked out of the box. But the new
machine is Red Hat 7.2, which uses sendmail by default. I copied over
my RT-specific lines in /etc/aliases and ran newaliases, and I can
tell that the aliases are picking up incoming mail OK. But when they
try to turn around and run rt-mailgate, this fails.

The mail I send in to one of my aliases bounces with this message:

----- The following addresses had permanent fatal errors -----
"|/usr/local/rt2/bin/rt-mailgate --queue general --action correspond"
(reason: service unavailable)
(expanded from: rt-general@nairobi2.garmin.com)

----- Transcript of session follows -----
smrsh: rt-mailgate not available for sendmail programs
554 5.0.0 “|/usr/local/rt2/bin/rt-mailgate --queue general --action
correspond”… Service unavailable

I wasn’t sure what that meant, so I tried running rt-mailgate by hand
(admittedly, not really knowing how to type in a real SMTP message by
hand):

[root@nairobi2 bin]# echo test| mail -s testing
rt-general@nairobi2.garmin.com |/usr/local/rt2/bin/rt-mailgate --queue
general --action correspond Can’t locate RT/Action/.pm in @INC (@INC
contains: /etc/rt2 /usr/local/rt2/lib /usr/lib/perl5/5.6.1/i686-linux
/usr/lib/perl5/5.6.1 /usr/lib/perl5/site_perl/5.6.1/i686-linux
/usr/lib/perl5/site_perl/5.6.1 /usr/lib/perl5/site_perl .) at (eval
52) line 3. Require of RT::Action:: failed. Can’t locate RT/Action/.pm
in @INC (@INC contains: /etc/rt2 /usr/local/rt2/lib
/usr/lib/perl5/5.6.1/i686-linux /usr/lib/perl5/5.6.1
/usr/lib/perl5/site_perl/5.6.1/i686-linux
/usr/lib/perl5/site_perl/5.6.1 /usr/lib/perl5/site_perl .) at (eval
52) line 3.

[root@nairobi2 bin]#

Now this looks a bit fishy. Can’t locate RT/Action/.pm? Shouldn’t
there be something between Action/ and .pm? Is there something still
missing in our setup, or is this just because of something weird with
me trying to run it by hand? Or is there something special we need to
do to sendmail? I have determined that the message “Require of
RT::Action:: failed” is generated by
/usr/local/rt2/lib/RT/ScripAction.pm just in case that helps turn on a
light bulb for anyone.

Here’s hoping someone finds this somewhat familiar…


Kendric Beachey


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

BTW: after I fixed that, it died saying it couldn’t write to /var/log/rt2.
So I made it world-writeable. This machine will not be on the internet,
just our intranet…does anyone know of any good reasons why this kluge is
unsafe?

It’s bad form. and completely unnecessary. see:
http://www.fsck.com/rtfm/article.html?id=5#110

seph

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Just curious, is this normal:

I don’t seem to be getting a unified rt.log file. Rather I am getting
a directory full of little imcomplete log files. As follows:

- -rw-r--r-- 1 www-data rt 0 Jan 2 09:53 rt.log.5317.33 - -rw------- 1 www-data www-data 0 Nov 14 10:28 rt.log.5428.0 - -rw-r--r-- 1 www-data rt 0 Nov 1 15:35 rt.log.559.33 - -rw-r--r-- 1 www-data rt 0 Nov 1 15:42 rt.log.582.33 - -rw-r--r-- 1 www-data rt 0 Nov 28 10:42 rt.log.5882.33 - -rw------- 1 www-data www-data 0 Nov 28 11:35 rt.log.6071.0 - -rw------- 1 www-data www-data 0 Nov 15 12:38 rt.log.6185.0 - -rw------- 1 www-data www-data 0 Nov 1 16:10 rt.log.645.0 - -rw-r--r-- 1 www-data rt 0 Dec 6 07:54 rt.log.6464.33 - -rw-r--r-- 1 www-data rt 156 Nov 15 13:32 rt.log.6465.33 - -rw-r--r-- 1 www-data rt 134 Nov 15 13:38 rt.log.6494.33 - -rw-r--r-- 1 www-data rt 0 Dec 15 06:56 rt.log.655.33 - -rw------- 1 www-data www-data 0 Nov 28 13:18 rt.log.6625.0 - -rw------- 1 www-data www-data 0 Nov 28 13:20 rt.log.6648.0 - -rw------- 1 www-data www-data 0 Nov 28 13:22 rt.log.6669.0 - -rw-r--r-- 1 www-data rt 0 Nov 1 16:10 rt.log.671.33 - -rw------- 1 www-data www-data 0 Nov 28 13:24 rt.log.6711.0 - -rw-r--r-- 1 www-data rt 0 Nov 28 13:25 rt.log.6727.33 - -rw-r--r-- 1 www-data rt 0 Jan 4 10:16 rt.log.6729.33 - -rw------- 1 www-data www-data 1050 Jan 3 17:49 rt.log.6831.0 - -rw-r--r-- 1 www-data rt 0 Dec 6 10:27 rt.log.6925.33 - -rw-r--r-- 1 www-data rt 0 Dec 6 11:22 rt.log.7117.33 - -rw-r--r-- 1 www-data rt 0 Dec 6 11:32 rt.log.7159.33 - -rw-r--r-- 1 www-data rt 0 Jan 2 14:38 rt.log.8665.33 - -rw-r--r-- 1 www-data rt 0 Jan 4 12:41 rt.log.9069.33 - -rw-r--r-- 1 www-data rt 0 Jan 2 15:24 rt.log.9204.33 - -rw-r--r-- 1 www-data rt 0 Dec 6 23:01 rt.log.9218.33 - -rw-r--r-- 1 www-data rt 0 Jan 2 15:49 rt.log.9417.33 - -rw-r--r-- 1 www-data rt 0 Jan 4 13:44 rt.log.9895.33 - -rw-r--r-- 1 www-data rt 0 Jan 4 13:48 rt.log.9912.33

It makes it hard to track down problems…


Jeremy

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.0.2i

iQA/AwUBPDn2UsWSutURApyZEQLlOACgtJefhIha4Iootbe1fuslqOzq1NMAniFD
oGa1+Vas9e+Bjbkro/Xu0mIG
=uLw4
-----END PGP SIGNATURE-----

Yes, it is the default.

If you want to change it, edit config.pm and change the following line to
suit your purposes and stop and start apache:

$LogToFileNamed = “$LogDir/rt.log.”.$$.".".$<; #log to rt.log..

NOTE: I haven’t actually tried this and don’t think it will break anything
but YMMV

Chris Scott
Host Orlando, Inc.
http://www.hostorlando.com/-----Original Message-----
From: rt-users-admin@lists.fsck.com [mailto:rt-users-admin@lists.fsck.com]On
Behalf Of Jeremy Freeman
Sent: Monday, January 07, 2002 2:26 PM
To: 'Rt-Users (E-mail)'
Subject: [rt-users] log files

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Just curious, is this normal:

I don’t seem to be getting a unified rt.log file. Rather I am getting
a directory full of little imcomplete log files. As follows:

- -rw-r--r-- 1 www-data rt 0 Jan 2 09:53 rt.log.5317.33 - -rw------- 1 www-data www-data 0 Nov 14 10:28 rt.log.5428.0 - -rw-r--r-- 1 www-data rt 0 Nov 1 15:35 rt.log.559.33 - -rw-r--r-- 1 www-data rt 0 Nov 1 15:42 rt.log.582.33 - -rw-r--r-- 1 www-data rt 0 Nov 28 10:42 rt.log.5882.33 - -rw------- 1 www-data www-data 0 Nov 28 11:35 rt.log.6071.0 - -rw------- 1 www-data www-data 0 Nov 15 12:38 rt.log.6185.0 - -rw------- 1 www-data www-data 0 Nov 1 16:10 rt.log.645.0 - -rw-r--r-- 1 www-data rt 0 Dec 6 07:54 rt.log.6464.33 - -rw-r--r-- 1 www-data rt 156 Nov 15 13:32 rt.log.6465.33 - -rw-r--r-- 1 www-data rt 134 Nov 15 13:38 rt.log.6494.33 - -rw-r--r-- 1 www-data rt 0 Dec 15 06:56 rt.log.655.33 - -rw------- 1 www-data www-data 0 Nov 28 13:18 rt.log.6625.0 - -rw------- 1 www-data www-data 0 Nov 28 13:20 rt.log.6648.0 - -rw------- 1 www-data www-data 0 Nov 28 13:22 rt.log.6669.0 - -rw-r--r-- 1 www-data rt 0 Nov 1 16:10 rt.log.671.33 - -rw------- 1 www-data www-data 0 Nov 28 13:24 rt.log.6711.0 - -rw-r--r-- 1 www-data rt 0 Nov 28 13:25 rt.log.6727.33 - -rw-r--r-- 1 www-data rt 0 Jan 4 10:16 rt.log.6729.33 - -rw------- 1 www-data www-data 1050 Jan 3 17:49 rt.log.6831.0 - -rw-r--r-- 1 www-data rt 0 Dec 6 10:27 rt.log.6925.33 - -rw-r--r-- 1 www-data rt 0 Dec 6 11:22 rt.log.7117.33 - -rw-r--r-- 1 www-data rt 0 Dec 6 11:32 rt.log.7159.33 - -rw-r--r-- 1 www-data rt 0 Jan 2 14:38 rt.log.8665.33 - -rw-r--r-- 1 www-data rt 0 Jan 4 12:41 rt.log.9069.33 - -rw-r--r-- 1 www-data rt 0 Jan 2 15:24 rt.log.9204.33 - -rw-r--r-- 1 www-data rt 0 Dec 6 23:01 rt.log.9218.33 - -rw-r--r-- 1 www-data rt 0 Jan 2 15:49 rt.log.9417.33 - -rw-r--r-- 1 www-data rt 0 Jan 4 13:44 rt.log.9895.33 - -rw-r--r-- 1 www-data rt 0 Jan 4 13:48 rt.log.9912.33

It makes it hard to track down problems…


Jeremy

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.0.2i

iQA/AwUBPDn2UsWSutURApyZEQLlOACgtJefhIha4Iootbe1fuslqOzq1NMAniFD
oGa1+Vas9e+Bjbkro/Xu0mIG
=uLw4
-----END PGP SIGNATURE-----