Nagios Support Forum

On XI 5.2.5, I have nrpe running some memory checks on a remote server. The graph that highcharts outputs looks a little off however:
I

I have tried uploading a different php template that worked under pnp 0.6.25 on our core server (the check command was imported from core, and the nrpe command defined on the remote server hasn't changed,) but that doesn't seem to affect the highcharts graph at all. It does break the standard rrd graph however, if we uncheck "use highcharts for perfdata graphs".

If the perfdata hasn't changed, what is the best way to straighten out the inaccuracies of the graph? I'm not sure how highcharts is handling this.

What about the above graph looks a little off to you? I'm having trouble seeing the issue that you're describing - is it the fact that CACHED goes down once in awhile and nothing appears to make up for that cache space?

Well, the KB KB KB along the side is a little odd, but the fact its reporting a total of 12MB of memory is incorrect..

The graph lists 'KB' for each memory value (FREE uses KB, CACHED uses KB, etc) - while the 12MB of memory you're referencing is actually 12 million KB, which equals about 12GB of memory - does that seem reasonable? I understand the graph is a little confusing, especially in this case. If you modify your check to report on megabytes instead of kilobytes I bet it'd look better. Thanks!

Ahh.. duh. That does make more sense. Thanks for the help, This is safe to lock!