I am currently using pnp4nagios as a graphing utility but have found that because it uses rrdtool, it has some limitations. Most of my services only run once a week which means that the graph doesn’t actually end up displaying anything due to the limited number of data points.
Does anyone know of a performance data tool which could work for my purpose? Even just displaying the data in a table would be an improvement.
Thank you for the response, I will read through the man pages and post what I find for others. I had previously posed this question on the pnp4nagios site and they responded saying rrdtool was not built for infrequent checks like this.
EDIT (21 Jul 2011 11:41 GMT-5):
The file with the default heartbeat and step size for pnp4nagios is set here: …/pnp4nagios/etc/process_perfdata.cfg
My understanding is that the heartbeat will set the maximum time to wait before an UNKNOWN value is returned and step will set the minimum interval between accepting a new value so I have left step alone and have changed the default heartbeat to be 1week (604800). I will repost later once I am able to see if it works. From what I understand, the only downside for a large heartbeat is that the graph will “hold” a value for a long time if no new data is received rather than set it to UNKNOWN.
that MIGHT be… but i honestly don’t see a big difference between adding a point every 300 or every 36000 secs…
might take some time to see a correctly working graph
Just an update for anyone else in this situation:
The solution posted above in my edit worked. All you have to do is change the heartbeat as described. I also received an email from the guys at pnp4nagios regarding which files contain information about heartbeat and step configuration. They noted the same file (process_perfdata.cfg) along with a second “rra.cfg” in the same directory. I do not have a rra.cfg file, but if you do you may want to check/edit that file as well.