I read the article and looked forward to trying out the scripts. It seemed like just what we need. Some very good ideas in here but the scripts should probably be tested on a more common distro and checked for compliance with FHS.

I spent a few minutes to try testing the scripts and soon decided it was not worth the effort. Please let us know if the scripts have been fixed.

First... have the scripts gone through a unix2dos filter?? Why?

For example:

debian@:~/test$ ./monitor_web.sh
: bad interpreter: No such file or directory
debian@:~/test$ hexdump -C monitor_web.sh |head -n 1
00000000 23 21 2f 62 69 6e 2f 62 61 73 68 0d 0a 23 0d 0a |#!/bin/bash..#..|

Note the "0d 0a" AKA CR-LF which confused the bash shell interpreter.

After using unix2dos to remove these... I noticed some necessary editing to do in the monitor_disk.sh script:

GREP="/sbin/grep"
MOUNT="/sbin/mount"
AWK="/usr/bin/awk"

... why bother.... normally these are in the path... if it is really necessary to put the commands in variables... why not make it easy for readers and use locations that will just work. or even better, use "which" to set your variables. grep and mount are usually not in /sbin.

After this, I noticed that one script requires ksh. Not a big deal but, not mentioned in the article.

Thanks for the ideas though. I may consider writing a simple ssh based monitoring system.