Ahem, I´d really like to help, but without any hint on the type of failure, it will be impossible. Surely you´ll agree that the possibilities for failure are countless, even on a properly configured system (let alone a bad or BOFH system;
<BOFH>
think of it: configure cron to run with a chroot environment, the users will never figure out what´s the problem! {veg}
</BOFH>
).

To aid your search for information:

• There should be a mail, sent to root, after the failure of one or more cron jobs. Look there for exit codes and/or messages.
• Rewrite your script to send a message to the syslog (or a file in /tmp, use a unique name including a timestamp) first thing (even before useing libraries. If you get the message, your script did start up and found a perl interpreter, so put in new debug messages. Otherwise, your perl installation might not be in the search path.

If you got more details, post a followup. We´ll be here :-).

Have fun ...

Andreas