FR: Need Performance Graph for Services/Processes

[vtracnagios]

A client want to have a performance graph for checking "services" and "processes".

Since checking for "services" only return "OK" for "Running" and "Failed) for "Stopped". There isn't any "values" returned from the service check for graphing.

It would be nice to have "graph", which will let the client know if or when the NCPA agent stopped.

[ericloyd]

Are you talking about active checks performed by the Nagios server or passive checks generated by the remote host that are sent to the Nagios server through NRDP or other methods?

[vtracnagios]

The client is talking about active check data coming from NCPA to Nagios XI not displaying performing graph.

Example: /usr/local/nagios/libexec/check_ncpa.py -H 192.168.x.x -t 'ncpaToken' -M 'services' -q 'service=ncpa_listener' OK: ncpa_listener is running

As you can see from above, there is no value(s) returned from the check, so no graphs were generated.

[ericloyd]

This check returns a boolean value of "running" or "stopped" with no other values, that I'm aware of. What possible performance data could it return?

[vtracnagios]

Which is why I opened this FR. Make it return some number like "1" or "0" or "3" (unknown) .... so that graph can be generated.

[ericloyd]

This is what availability reports are for. You should educate your customer on how to use them properly. A world of reporting possibilities will be opened before their very eyes!

[ericloyd]

Also, why run a service check looking for the listener status? We typically run a check looking for the NCPA server version number, then tie that as a dependency for all the active NCPA based service checks. If that fails to respond, then the others don't trigger, plus you get to track the version number over time so you know when it's time to upgrade.

Lastly, isn't it "ncpalistener" not "ncpa_listener"?

[ccztux]

To get peformance data you can do this:

OK example:

08:37:18 ✓ LAB-CL01 root@cl01 ~ # /usr/lib64/meg/plugins/check_ncpa.py -H localhost -t mytoken -M services -q 'service=gssproxy' --performance
OK: gssproxy is running | 'status'=0;1;2;;

UNKNOWN example:

08:37:20 ✓ LAB-CL01 root@cl01 ~ # /usr/lib64/meg/plugins/check_ncpa.py -H localhost -t mytoken -M services -q 'service=gssproxyS' --performance
UNKNOWN: No services found for service names: gssproxyS | 'status'=3;1;2;;

Snippet from the help output of check_ncpa.py:

  -p, --performance     Print performance data even when there is none. Will
                        print data matching the return code of this script

[ccztux]

Also, why run a service check looking for the listener status? We typically run a check looking for the NCPA server version number, then tie that as a dependency for all the active NCPA based service checks. If that fails to respond, then the others don't trigger, plus you get to track the version number over time so you know when it's time to upgrade.

Lastly, isn't it "ncpalistener" not "ncpa_listener"?

It is ncpa_listener

08:41:07 ✓ LAB-CL01 root@cl01 ~ # /usr/lib64/meg/plugins/check_ncpa.py -H localhost -t mytoken -M services -q 'service=ncpa_listener' --performance
OK: ncpa_listener is running | 'status'=0;1;2;;

[vtracnagios]

Thanks guys!! :-) I think this issue can be close as there is an option "--performance", which can be used for this.

[vtracnagios]

I have just learned that there is an option "--performance", which can be used to generate performance graph for "services".

[ericloyd]

Ooops. Windows sees the service name as "ncpalistener" while Linux sees it with the dash.

[ccztux]

Ooops. Windows sees the service name as "ncpalistener" while Linux sees it with the dash.

You are absolutely right. I forgot about the inconsistent naming.