Metric graphs not working

[cityofrefuge]

Hey guys I was just wondering if anyone could shed some light on an issue I am having with the metrics addon in Nagios XI.

When I click it I want to see all the graphs for certain services on every host. Right now I have Disk Usage, CPU Usage, Memory Usage, Load and Swap that I am monitoring. When I click Disk usage, Load, and Swap I get graphs for every host. However when I click CPU Usage or Memory Usage nothing comes up.

I went to check if I could see those graphs elsewhere like in service detail and I was able to see the graphs.

Any help would be amazing!

Thanks!

[mguthrie]

The metrics component looks at the actual plugin output from each of the services to determine which ones are for CPU usage and memory. The graphs you're looking for coming from linux or windows servers? Which wizard did you run to setup the checks?

[cityofrefuge]

Thanks for the reply,

All of the hosts I am monitoring right now are Ubuntu 9 and 10 and I ran the "Linux Server: Monitor a remote Linux server." Wizard - I kept all of the default values that the wizard had set by default.

[cityofrefuge]

Oh and a little bit of extra information about nagiosxi is below.

My Nagios XI installation is a 32-bit VM that I downloaded from the website.

It said there was an update so I updated yesterday to see if it would fix anything so right now its currently running on version: Nagios XI 2012R1.0

I am running on the trial period right now, my company is determining whether we want to purchase it or not.

One thing I did notice is in the services its named "CPU Stats" and in the command line it calls "check_cpu_stats"...is this a different plugin than the one that is being called to in the metrics? (I only ask because its called CPU Usage in the metrics section)

[mguthrie]

OK, so your question got my hunting on this one because I was somewhat curious about it myself. So it looks like the CPU Usage is intended for windows servers, and has a metric that looks like the following:

'5 min avg Load'=2%;80;90;0;100

While CPU usage for a linux server is measured with the following metric:

CPU STATISTICS OK: user=35.31% system=3.73% iowait=0.00% idle=60.71%

The metric that gets captured for Linux servers and can be used to compare CPU is the Load metric, which looks like the following:
load1=0;5;10;; load5=0;4;10;; load15=0;4;7;;

Because the measurements are different across the platforms they can't be accurately compared on the same table. So, for linux machines, use Load, for Windows use the CPU usage. : )

[cityofrefuge]

Awesome! It was right under my nose haha. Thanks a lot, you helped big time.

One more thing though, this is exactly what we are looking for but it seems as if the metric section is set to show graphs within a 3 day range by default. Is there a way to change this range so that I can customize it to like a month or something along those lines while still seeing this specific metric (load) for every host on the same page?

(Sorry if I am quite the picky user haha)

[mguthrie]

We've actually got a feature request logged for this exact item.
http://tracker.nagios.com/view.php?id=280
We just got through our final push to get 2012 out the door, so we have a bit of a development queue at the moment for some of the features like this, but it's good to ask because our development queue gets prioritized based on user requests and needs.

[cityofrefuge]

Great! Okay thanks a lot for all of your help!