MRTG: how to monitor multiple devices?

I'm able to monitor my Dell Powerconnect 48 switch, but I'd like to monitor five others. What files need to be changed to do this? Do I need to run cfgmaker on mrtg again?
 
Create separate config files for each device you want to monitor.

Then create a "master" mrtg config file that is nothing but global config entries, and Include lines that point to each of the individual device config files.

I'd recommend, though, switching to using the RRD backend for MRTG, separating your "data collection" from "data display", and look into using something like Routers2 to dynamically create the graphs as needed.

We do this, polling for network stats and load averages every 1 minute, disk I/O every 2 minutes, CPU stats, memory stats, and load averages every 5 minutes, and disk usage every 15 minutes.

For over 127 servers across the district. All nicely laid out in cascading menus by Routers2, with graphs only being created when viewed.
 
I'm new to snmpd and mrtg. I'm assuming when you say create separate config files do you mean snmpd.con files?

Is there a doc on creating the global and include entries in the mrtg.cfg file?

Thanks.
 
phoenix said:
We do this, polling for network stats and load averages every 1 minute, disk I/O every 2 minutes, ...

I will be happy if you can tell me how do you do to have a polling time of 1 or 2 minutes with mrtg ?
I red in docs and/or forums that minimum polling time in mrtg is 5 minutes which i think is too much to have something realistic and with visual alerts.

Thanks :)
 
The lowest polling time that MRTG can do with it's internal logging is 5 minutes, yet.

However, if you switch the log format to rrd, then you can go down to 1 minute. :)

There's lots of docs on how to do this on the MRTG website.
 
Create separate config files for each device you want to monitor.

Then create a "master" mrtg config file that is nothing but global config entries, and Include lines that point to each of the individual device config files.

I'd recommend, though, switching to using the RRD backend for MRTG, separating your "data collection" from "data display", and look into using something like Routers2 to dynamically create the graphs as needed.

We do this, polling for network stats and load averages every 1 minute, disk I/O every 2 minutes, CPU stats, memory stats, and load averages every 5 minutes, and disk usage every 15 minutes.

For over 127 servers across the district. All nicely laid out in cascading menus by Routers2, with graphs only being created when viewed.

Hi Phoenix,

I have create individual mrtg.cfg file for each node, how to create master mrtg.cfg file and how to include the different node mrtg.cfg file.
 
Back
Top