Non uniform data intervals

[jonjsilver]

I am looking at an application to monitor phone call quality.
During the call, I will recieve jitter, lost packet... data every 30 seconds or every minute.
During non-call periods, I will receive no data.
Can I do alarming with nagios with this kind of data?

I'd like to look at data within a call and also aggregate data for multiple calls.
thanks much,
jon

[slansing]

Are you using any software to monitor your calls currently? It is possible that you could configure said software to return error codes and packet loss data to nagios for alerting.

[jonjsilver]

That is actually what I am trying to figure out. I know ganglia, but no nothing about nagios.
We can collect the data (it actually comes as trap information). We gave a manager that collects this raw data and stores it.

Instead of us storing it and doing our own processing, I was looking at using ganglia / nagios.
I can pass it to ganglia (gmond) using the gmetric utility.
ganglia graphs it ok - it just shows zeros for the in-between call periods.
However, how about monitoring and alerting? Will I be able to use Nagios for that since the data is not constant?
Where can I read up on the rrules and algorithyms that I can configure for Nagios?

thanks,
jonathan

[slansing]

We have a pretty steep read on developer guidelines found here:

http://nagiosplug.sourceforge.net/devel ... lines.html

The plugin would have to be passed at least one metric, and within that have a warning "-w" and critical "-c" value and a return code of 0, 1, 2, 3 [Unknown, Ok, Warning, Critical] for nagios to properly read it.