Thank you for taking the time to answer so fully. I have rejigged the script to use strict, and -w. After a bit of cleaning up, the script is working again, but I am a bit confused at how to approach the error handling.

The script will be a part of a set of scripts to take snap shots of LUN's(Logical Units of Storage) on an SAN. Your check_call approach is exactly what I need for the standard unix commands, but EMC does not appear to use non-standard return codes in their software i.e. With admsnap(for taking snapshots), 13 is returned when a snapshot is activated, which complicates the issue.

The other complicating factor is that on some machines, the scripts will be run by operators from a menu system. This means that if they encounter any problems, we have to provide them with a plain english explanation of the problem, before returning them to the menu.