Graph not displayed for Invoke-IcingaCheckCPU

Icinga Web Modules Visualization, Metrics & Reporting
1

Hi team,
I created a new check using Invoke-IcingaCheckCPU but no graphs are shown.

The whisper files were created but the graph debug log shows that the metrics were excluded:

+ Icinga check command: 'Invoke-IcingaCheckCPU'
+ Obscured check command: NULL
+ Applying templates for check command 'Invoke-IcingaCheckCPU'
++ Applying template 'load-windows'
+++ Fetched 0 metric(s) from 'http://192.168.9.77:81/metrics/expand?query=icinga2.KTC01_LABOR_domain_com.services.Windows_CPU.Invoke-IcingaCheckCPU.perfdata.%2A.value'
+++ Fetched 0 metric(s) from 'http://192.168.9.77:81/metrics/expand?query=icinga2.KTC01_LABOR_domain_com.services.Windows_CPU.Invoke-IcingaCheckCPU.perfdata.%2A.crit'
+++ Fetched 0 metric(s) from 'http://192.168.9.77:81/metrics/expand?query=icinga2.KTC01_LABOR_domain_com.services.Windows_CPU.Invoke-IcingaCheckCPU.perfdata.%2A.warn'
+++ Excluded 0 metric(s)
+++ Combined 0 metric(s) to 0 chart(s)
+ Applying default templates, excluding previously used metrics
++ Not applying template 'default-host'
++ Applying template 'default-service'
+++ Fetched 0 metric(s) from 'http://192.168.9.77:81/metrics/expand?query=icinga2.KTC01_LABOR_domain_com.services.Windows_CPU.Invoke-IcingaCheckCPU.perfdata.%2A.value'
+++ Excluded 0 metric(s)
+++ Combined 0 metric(s) to 0 chart(s)
++ Not applying template 'multi2-host'
++ Applying template 'multi2-service'
+++ Fetched 0 metric(s) from 'http://192.168.9.77:81/metrics/expand?query=icinga2.KTC01_LABOR_domain_com.services.Windows_CPU.Invoke-IcingaCheckCPU.perfdata.%2A.%2A.value'
+++ Excluded 0 metric(s)
+++ Combined 0 metric(s) to 0 chart(s)
++ Not applying template 'multi3-host'
++ Applying template 'multi3-service'
+++ Fetched 10 metric(s) from 'http://192.168.9.77:81/metrics/expand?query=icinga2.KTC01_LABOR_domain_com.services.Windows_CPU.Invoke-IcingaCheckCPU.perfdata.%2A.%2A.%2A.value'
+++ Excluded 10 metric(s)
+++ Combined 0 metric(s) to 0 chart(s)

My cpu_windows_powershell_framework.ini template file looks like this:

[load-windows.graph]
check_command = "Invoke-IcingaCheckCPU"

[load-windows.metrics_filters]
load.value = "$service_name_template$.perfdata.$load$.value"
crit.value = "$service_name_template$.perfdata.$load$.crit"
warn.value = "$service_name_template$.perfdata.$load$.warn"

[load-windows.urlparams]
areaAlpha = "0.5"
lineWidth = "2"
min = "0"
title = "CPU $load$ %"
yUnitSystem = "none"

[load-windows.functions]
load.value = "alias(color($metric$, '#1a7dd7'), 'CPU usage(%)')"
crit.value = "alias(color($metric$, '#ff0000'), 'Crit (%)')"
warn.value = "alias(color($metric$, '#ff8d00'), 'Warn (%)')"

How can we futher investigate the issue?
The check is called “Windows CPU” to avoid problems with special chars

Thanks and kind regards
Markus

2

Hi Markus,

are you sure about

check_command = "Invoke-IcingaCheckCPU"

?

Is the name of the CheckCommand (not the Service) correct?

Greetings.

3

Hi,
yes I am sure:
https://icinga.com/docs/icinga-for-windows/latest/plugins/doc/plugins/04-Invoke-IcingaCheckCPU/
The check is working as it should and I am receiving performance data.

4

Did you get this sorted out as I also have this problem ?

5

No, it’s still open…

6

OK, seems like it must just be some simple error?

7

I get this in the debug:

+ Icinga check command: 'Invoke-IcingaCheckCPU'
+ Obscured check command: NULL
+ Applying templates for check command 'Invoke-IcingaCheckCPU'
++ Applying template 'load-windows'
+++ Fetched 0 metric(s) from 'http://127.0.0.1:81/metrics/expand?query=icinga2.server01_local.services.windows-Invoke-IcingaCheckCPU.Invoke-IcingaCheckCPU.perfdata.%2A.value'
+++ Fetched 0 metric(s) from 'http://127.0.0.1:81/metrics/expand?query=icinga2.server01.services.windows-Invoke-IcingaCheckCPU.Invoke-IcingaCheckCPU.perfdata.%2A.crit'
+++ Fetched 0 metric(s) from 'http://127.0.0.1:81/metrics/expand?query=icinga2.server01.services.windows-Invoke-IcingaCheckCPU.Invoke-IcingaCheckCPU.perfdata.%2A.warn'
+++ Excluded 0 metric(s)
+++ Combined 0 metric(s) to 0 chart(s)
+ Applying default templates, excluding previously used metrics
++ Not applying template 'multi2-host'
++ Applying template 'multi2-service'
+++ Fetched 0 metric(s) from 'http://127.0.0.1:81/metrics/expand?query=icinga2.server01.windows-Invoke-IcingaCheckCPU.Invoke-IcingaCheckCPU.perfdata.%2A.%2A.value'
+++ Excluded 0 metric(s)
+++ Combined 0 metric(s) to 0 chart(s)
++ Not applying template 'multi3-host'
++ Applying template 'multi3-service'
+++ Fetched 3 metric(s) from 'http://127.0.0.1:81/metrics/expand?query=icinga2.server01.services.windows-Invoke-IcingaCheckCPU.Invoke-IcingaCheckCPU.perfdata.%2A.%2A.%2A.value'
+++ Excluded 3 metric(s)
+++ Combined 0 metric(s) to 0 chart(s)
++ Not applying template 'default-host'
++ Applying template 'default-service'
+++ Fetched 0 metric(s) from 'http://127.0.0.1:81/metrics/expand?query=icinga2.server01.windows-Invoke-IcingaCheckCPU.Invoke-IcingaCheckCPU.perfdata.%2A.value'
+++ Excluded 0 metric(s)
+++ Combined 0 metric(s) to 0 chart(s)

I see that i applies the template ‘default-service’… It seems not to apply the template ‘cpu_windows_powershell_framework.ini’ which I suppose should be the correct one? Is that maybe the problem?

8

or is it actually the “Applying template ‘load-windows’” row that tells which template it’s using?

9

I had the same problem you simply need to change the name of the command to

“Invoke-IcingaCheck-cpu”

cpu must be lowercaps.

that worked for me

10

OK, you mean in the template? If so, in which template file? Or do you mean in the actual check command ?

11

in the actuall command

image
image1621×424 33.4 KB

1 Like
12

Yes!!! :grinning: Thanks @ExeLeNtCrypt !

image
image1164×1023 53.5 KB

13

Perfect - Thanks @ExeLeNtCrypt !

14

Another question that maybe any of you can answer… what is causing the graphs to have “gaps” in them… like this for some:

image
image903×207 12.2 KB

but for instance the CPU looks like this (better!)

image
image1095×207 13.9 KB

I assume it has to do with how the template is configured ? Or have I made an mistake in some config?

15

Noone has any clue of the above issue?

16

this could be the configuration of your storage-schema.conf

17

OK, I was also looking into that… This is how it looks right now:

pattern = ^carbon.
retentions = 60:90d

[icinga-hosts]
pattern = ^icinga2..*.host.
retentions = 5m:4w

[icinga-services]
pattern = ^icinga2..*.services.
retentions = 5m:12w,15m:26w,30m:1y

[icinga2_default]

intervals like PNP4Nagios uses them per default

pattern = ^icinga2.
retentions = 1m:2d,5m:10d,30m:90d,360m:4y
[default]
pattern = .*
retentions = 60s:1d,5m:30d,1h:3y

18

I actually also see that I cant see graphs older than 2 days… But according to the pattern the data should be there, right?

19

I saw some notes regarding storage-aggregation.conf, but I actually don’t have any such file at all in the /etc/carbon directory… I was following different setup guides so maybe I’ve missed some part?

I see older data on the CPU graphs, which was the ones we fixed before… so maybe the only thing I need to do is to set storage-schema to:
[icinga2_default]
pattern = ^icinga2.
retentions = 1m:2d,5m:10d,30m:90d,360m:4y

Then run the whisper-resize tool ? but how should I then run that?

20

so i am not really an expert on the gaphite stuff.

i run a less fine-grained schema:

[icinga_default]
pattern = ^icinga\.
retentions = 1m:2d,5m:10d,30m:90d,360m:4y

[default]
pattern = .*
retentions = 60s:1d,5m:30d,1h:3y

not sure what you mean with the whisper-resize tool?

1 Like