[Discuss] Simple server monitoring

[blu at nedharvey.com (Edward Ned Harvey (blu))]

> From: discuss-bounces+blu=nedharvey.com at blu.org [mailto:discuss-
> bounces+blu=nedharvey.com at blu.org] On Behalf Of Tim Callaghan
> 
> I need to add basic monitoring to a server and am wondering how I can
> easily accomplish the following:
> 
> 1. Keep track of CPU, Memory, and Disk IO / Space.
> 2. Allow me to set alerts and send emails if certain thresholds are passed.

Just one system?  Script it yourself in cron.

I recently deployed zabbix.  (A few months ago).  I find, it does everything you need it to do, and it's pretty reasonable on all accounts.  But the web interface is kind of confusing.  You know what you want it to do, you know it can do it, but you just can't find the right button to click, because the wording and/or layout is weird.

If you do use zabbix, make sure all your servers and agents are above version 2.0.  That's when they added the ability to generate alerts when free disk space drops below a threshold.  This basically means, since 2.0 isn't distributed in any OS by default, just go download the precompiled binaries from their website.  You have to create the service user account yourself, and create the config file and startup scripts yourself.