I just read about SMART (Smart Monitoring And Rebooting Tool) in a Linux magazine recently so I tinkered with it a bit.
It’s a system monitoring tool, kind of like Nagios, but much, much simpler (no Web UI, no flap detection, etc.). I think if I cared about something enough to monitor it, it’s probably worth it to set up Nagios, as Nagios is more sophisticated and has been thoroughly battle-tested. But for some reason if you wanted something much more lightweight and simpler than Nagios, than SMART could be of interest. If you hate Nagios because it’s written in Perl, I’m not what sure what you’ll think of SMART, since it’s written as a bunch of bash scripts – worse, if you ask me, but different strokes for different folks.
You can download SMART from here (Universitat Internacional de Catalunya).
Here’s some sample output (after a bit of hacking on the configuration files):
~/sw/smart$ ./smart SERVICE PID PROCS STATUS PROBLEM ------- ----- ----- ------ ------- APACHE2 6188 4 [OK] ATD 4926 1 [OK] BACKUPPC 4335 1 [OK] CRON 4939 1 [OK] CUPS 6279 1 [OK] DISK ? 0 [OK] No start command. KLOG 4079 1 [OK] NFS ? 8 [OK] NMB 4698 1 [OK] NTP 4861 1 [OK] PORTMAP ? 1 [OK] POSTFIX 4668 1 [OK] RPC.MOUNTD ? 1 [OK] No start command. RPC.STATD 4834 1 [OK] No start command. SMB 4700 2 [OK] SSH 4719 3 [OK] SYSLOG 7430 1 [OK]