[Xymon] network traffic monitoring

[Mark Felder]

On Dec 6, 2013, at 8:31, Scot Kreienkamp <SKreien at la-z-boy.com> wrote:

> MRTG has the ability to monitor much more than just throughput.  MRTG will collect and graph any number of variables (up to 4 I believe, I’ve never tried more than two) that can be collected by SNMP or scripts.  I have used it to monitor temperature, CPU load, length in minutes of daylight during the day, and several other things as well as throughput over the years I’ve used it.  I’ve also used RRDTool directly to monitor and maintain graphs of various things when MRTG didn’t allow enough variables in a single graph or if I wanted more control over the process. 
>  
> The bottom line is both MRTG and Devmon will work, as well as doing something directly with RRDTool like I have if Kris cares to do that.  I advocate for MRTG simply because I think setup/integration is easier with MRTG and Xymon. 


This may be true, but the MRTG integration cannot assist Xymon with sending *alerts* for things like:

- high cpu
- high memory
- high temp
- interface errors, discards, or collisions
- interface under a high load
- interface down, but administratively up
- power supply failure
- fan failure
- wrong number of route prefixes


These are all things devmon can have xymon alert you for, but MRTG can only graph them.