Skip to main content
Solved

CMA Windows 24.10.18 - Execute command failed

  • August 6, 2026
  • 5 replies
  • 45 views
S
Forum|alt.badge.img+4

Hello,

I've just updated my Windows agents in 24.10.18 version and my Centreon Engine in 24.10.23 version.

Now the CMA services sometimes switch to unknown status with “Execute command failed” result.
If I look at the centengine.log as @matty69v propose in another post I have this : 

[2026-08-06T16:28:29.989+02:00] [runtime] [warning] [1794793] Warning: The results of service 'CPU' on host 'XX' are stale by 0d 0h 2m 0s (threshold=0d 0h 2m 0s).  I'm forcing an immediate check of the service.
[2026-08-06T16:28:29.998+02:00] [otl] [error] [1794793] open telemetry services must be passive
[2026-08-06T16:28:29.999+02:00] [runtime] [warning] [1794793] Error: Service check command execution failed: open telemetry services must be passive
[2026-08-06T16:28:35.000+02:00] [runtime] [warning] [1794793] Warning: The results of service 'CPU' on host 'XX' are stale by 0d 0h 2m 5s (threshold=0d 0h 2m 0s).  I'm forcing an immediate check of the service.
[2026-08-06T16:28:35.002+02:00] [events] [info] [1794793] SERVICE ALERT: XX;CPU;UNKNOWN;SOFT;2;(Execute command failed)

Where can I modify the default threshold of 2 minutes ?

Thanks

Best answer by fmattes

Hi ​@sps ,

Thanks for the details, that helped a lot. What you're seeing is a known limitation on the 24.10.x line: when "Check Freshness" is enabled on services collected by the CMA, the platform sometimes tries to actively re-run the check, which the agent rejects, hence the intermittent UNKNOWN "(Execute command failed)". It clears on its own at the next data push, which matches exactly what you describe.

Freshness for CMA is only available in 25.10 version.

Here's the workaround on 24.10:

1. Go to Configuration > Services (or the corresponding Service Template used by your CMA services).
2. Open the Data Processing tab.
3. Set "Check Freshness" to No.
4. Save, then export the configuration (Configuration > Pollers > Export configuration, with "Restart" enabled).

The UNKNOWN flapping should stop right away. The only trade-off is that you'll no longer be automatically alerted if an agent stops sending data. If that detection matters to you, the clean solution is to move to 25.10.

Hope this helps!

    5 replies

    fmattes
    Centreonian
    Forum|alt.badge.img+11
    • fmattes
    • Centreonian
    • August 11, 2026

    Hi, 

    Could you share:
    - the CMA log on the Windows host around the time of an UNKNOWN,
    - whether the UNKNOWN clears on its own or stays?
    - and whether it affects all services on the host or only some?

    That will tell us whether it's an agent-side regression in 24.10.18 or a configuration issue. Thanks!

    Florian MATTES - Group Product Manager - Data collection
    S
    Forum|alt.badge.img+4
    • sps
    • Author
    • Steward **
    • August 12, 2026

    Hi ​@fmattes,

    There are no logs on the Windows agent, the most recent are from last week :

    [2026-08-06 15:52:32.637] [centreon-monitoring-agent] [error] [check_drive_size.cc:74] D:\ unable to get file system type

    [2026-08-06 16:06:39.623] [centreon-monitoring-agent] [error] [check_drive_size.cc:74] D:\ unable to get file system type

    The unknown disappears and come back a few minutes later.

    It affects all services but I noticed CPU most often.

    Thanks for your help 

    S
    Forum|alt.badge.img+4
    • sps
    • Author
    • Steward **
    • August 12, 2026

    We tested it using the 24.10.17 agent and got the same result, I think it could be coming from centreon-engine 24.10.23.

    fmattes
    Centreonian
    Forum|alt.badge.img+11
    • fmattes
    • Centreonian
    • Answer
    • August 12, 2026

    Hi ​@sps ,

    Thanks for the details, that helped a lot. What you're seeing is a known limitation on the 24.10.x line: when "Check Freshness" is enabled on services collected by the CMA, the platform sometimes tries to actively re-run the check, which the agent rejects, hence the intermittent UNKNOWN "(Execute command failed)". It clears on its own at the next data push, which matches exactly what you describe.

    Freshness for CMA is only available in 25.10 version.

    Here's the workaround on 24.10:

    1. Go to Configuration > Services (or the corresponding Service Template used by your CMA services).
    2. Open the Data Processing tab.
    3. Set "Check Freshness" to No.
    4. Save, then export the configuration (Configuration > Pollers > Export configuration, with "Restart" enabled).

    The UNKNOWN flapping should stop right away. The only trade-off is that you'll no longer be automatically alerted if an agent stops sending data. If that detection matters to you, the clean solution is to move to 25.10.

    Hope this helps!

    Florian MATTES - Group Product Manager - Data collection
    S
    Forum|alt.badge.img+4
    • sps
    • Author
    • Steward **
    • August 12, 2026

    It works thanks, but as you said freshness is therefore unavailable for our CMA agents.

    We’ll wait until the upgrade :)

     

      Terms & ConditionsAccessibility statement