Problem getting performance metrics graph with custom nagios style service. "No historic metrics recorded but performance data is available. Maybe performance data processing is disabled."

Troubleshooting
1

**CMK version:2.2.0p24
**OS version:Redhate Enterprise Linux 8.10

**Error message:No historic metrics recorded but performance data is available. Maybe performance data processing is disabled.

I created a custom shell script like an old style nagios check and installed it. It outputs perfomance data like this:
‘Temp (F)’=10.6;5.0:36.0;1.0:38.0;30;120
I dont get a perfomance graph, performance data is enabled in master control. Im not sure what else I need to do to get the performance data in there? Is a range not acceptable here like it is in nagios?

My script output line is like this echo “TEMP OK - $temp F|‘Temp (F)’=$temp;$warningrange;$criticalrange;30;120” the $warningrange is done like 5.0:36.0 and $criticalrange is 1.0:38.0

2

use “temp” as metric name and not “Temp (F)” and use Celsius as unit. Checkmk uses SI units for all the values.

1 Like
3

First, thank you so much for the answer. I added code to return in celcius and changed the output variable name to temp= which seems to at least start filling in the service perf-o-meter line in the service details and have temperature be recognized, but I still see the No historic metrics recorded but performance data is available. Maybe performance data processing is disabled error.

Do I need to wait for 4 hours of data to be collected maybe?

service performance data (source code) now shows as temp=10.1;1.7:10.0;-1.1:12.8;0;50

4

I updated our site to v2.2p32 in case that helped but see no change

5

Just my few cents…

I would suggest letting go the old ‘Nagios-style’ and use the CheckMK native way of writing a Local check.
It is not that different, just the location of your script and the format of the line(s) is in the output.
IMHO this would also make your Local check more ‘future-proof’

Please see Local checks for how to do this.

  • Glowsome
6

I don’t exactly know if the rrd files will get adjusted accordingly if you change your scripts post-haste.
Therefor I think that if you delete the corresponding .xml and .rrd file on the filesystem and restart your monitoring core, the files should get recreated and that might fix your issue at hand.

1 Like
7

Thanks for the response. I would love to use a local check type of check, but from my understanding, that requires an agent and these are simple SNMP devices that I am polling an OID and doing some logic to check within ranges. I can’t put an agent on these devices to run any local code.

8

Thank you also for this, I will have to research where to find the corresponding .xml and .rrd files for these checks are and then I will try deleting them and restarting to check. Really do appreciate the help though.

9

That was a piece of information which was not included in the original topic-starter. :eyes:
And yes, that changes a/the conditions as to making a Local check. :slight_smile:

I think your route would point towards a datasource (been playing around with that myself https://forum.checkmk.com/t/howto-p1-meter-monitoring-in-cmk/49317)

Reference: https://docs.checkmk.com/latest/en/datasource_programs.html

  • Glowsome
2 Likes
10

If you want to fetch single OIDs as metric from an SNMP agent we got you covered: Checkmk Exchange

2 Likes
11

This is exactly what I was trying to do with my script, thank you so much for this.

Chris

12

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed. Contact an admin if you think this should be re-opened.