Re: [Exim] Urgent: Exim failing on production server

Top Page
Delete this message
Reply to this message
Author: Jeffrey C. Ollie
Date:  
To: Whitney Williams
CC: exim-users
Subject: Re: [Exim] Urgent: Exim failing on production server
On Sun, Feb 11, 2001 at 08:32:08PM -0800, Whitney Williams wrote:
>
> The return code 134 is constant -- although at times the command referred to
> is an in-house Perl script (which has worked flawlessly for months). Most of
> the time, however, the failing command is usually /usr/bin/env!?


A common trick among script writers is to put:

#!/usr/bin/env perl

at the top of the script rather than:

#!/usr/local/bin/perl

This trick improves script portability because you don't have to edit
the script if you install a script on a system that has perl installed
somewhere other than /usr/local/bin.

I'm pretty certain that someone has done something that has changed
the search path that exim passes on to subprocesses. Perhaps some
security patch or something. With the changed path, the script
interpreter cannot be located by env.

The best fix is to find the script that is being executed (look in the
command option in the remote_pipe transport) and change any references
to files to absolute paths.

In the meantime, add a path option to the remote_pipe transport that
contains the path to the script interpreter. See:

<http://www.exim.org/exim-html-3.20/doc/html/spec_18.html#SEC526>

for documentation on the path option.

Jeff