Two RT intances in two Apache servers in the same box. Is it possible?

Hi guys,

I�m setting up a second RT instance, for testing, in the same box where we
have the production RT instance.
Often, we need stop and restart the Apache server to show the most recent
changes, but we can�t to do this with the production site. So, I�ve
installed another Apache server (both Apache/1.3.31) and another instance of
RT (3.0.10 for production and 3.0.11 for testing).
The production RT instance runs still great, but RT test instance causes
always an �Internal Server Error�.

When, from RT_Siteconfig, I erase a line

SetHandler perl-script
PerlHandler RT::Mason �erased!

I can see the index.html page (but not compiled with Mason�), so I imagine
than the error comes from this sentence.

After googeling, search in wiki page (there is
Request Tracker Wiki but the two
instances have the same Apache server) and search in users list, I�ve found
an error with mason permissions who return �Internal Server Error � too
(, but I think my
case isn�t a permissions problem.

Finally, I�m completely lost�
Any ideas?

---- Configuration:
Debian Woody Kernel 2.4
RT 3.0.11

---- Apache relevant configuration:

ServerName mydomain
DocumentRoot /opt/test/rt3/share/html
ErrorLog /opt/test/rt3/var/log/error_log
TransferLog /opt/test/rt3/var/log/access_log
AddDefaultCharset UTF-8
PerlModule Apache::DBI
PerlRequire /opt/test/rt3/bin/

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

---- Tail of the error_log:
[Mon Jun 21 17:47:57 2004] [notice] caught SIGTERM, shutting down
[Mon Jun 21 17:48:03 2004] [notice] Apache/1.3.31 (Unix) mod_perl/1.29
configured – resuming normal operations
[Mon Jun 21 17:48:03 2004] [notice] Accept mutex: sysvsem (Default: sysvsem)

---- Web interface
Internal Server Error (500)
The server encountered an internal error or misconfiguration and was unable
to complete your request.
Please contact�

Thanks a lot!


Marc Boix i Ricart
Dept. de Comunicacions
Tf: 93 205.64.64 - ext 4454
CESCA (Centre de Supercomputaci� de Catalunya)

If you are going to run 2 apaches on the same box,
you must be running on different ports.

The way I described running 2 RT’s on the same apache,
each had it’s own set of fastcgi processes. I suspect
(but haven’t tried) that you could actually kill one
set of fastcgi’s, without killing the other set.
Apache would restart them, and this would give you
the same affect as restarting apache does now.

(That is, the whole point of restarting apache is
to reload code changes. If you are using mod_perl,
you need to do this. But if you are using fastcgi,
the code is loaded into fastcgi, not into apache itself.)