Windows exporter stopped working after a few minutes

[ajarillo1981]

Current Behavior

Windows exporter stops working after some minutes, no more than 10 minutes. It happens running windows exporter as a service or executing the app manually.
I have enabled debug in logs and there is no error, just stop providing logs and url http://localhost:9182/metrics become unaccessible. The service keeps as started but cannot be stopped (only possible finding the PDI and killing the process with "taskkill /f")

I have tried also running with only one collector, but the same results-
It is happening the same in the 2 servers that we need to monitor.

¿Is there any other test I can do to see why windows exporter freezes?

Expected Behavior

The application windows exporter should be providing metrics and http://localhost:9182/metrics should be accesible

Steps To Reproduce

Environment

• windows_exporter Version: 0.30.6
• Windows Server Version: 2016

windows_exporter logs

time=2025-05-07T12:33:46.611Z level=DEBUG source=collect.go:220 msg="collector memory succeeded after 5.9986ms, resulting in 35 metrics"
time=2025-05-07T12:33:46.612Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 6.9981ms, resulting in 609 metrics"
time=2025-05-07T12:33:46.613Z level=DEBUG source=collect.go:220 msg="collector net succeeded after 7.9976ms, resulting in 27 metrics"
time=2025-05-07T12:33:46.628Z level=DEBUG source=collect.go:220 msg="collector service succeeded after 23.0823ms, resulting in 2872 metrics"
time=2025-05-07T12:34:01.605Z level=DEBUG source=collect.go:220 msg="collector cs succeeded after 0s, resulting in 3 metrics"
time=2025-05-07T12:34:01.606Z level=DEBUG source=collect.go:220 msg="collector os succeeded after 1.0137ms, resulting in 11 metrics"
time=2025-05-07T12:34:01.608Z level=DEBUG source=service.go:282 msg="failed collecting service config" collector=service err="El sistema no puede encontrar el archivo especificado." service=CloudAvUpdaterEx
time=2025-05-07T12:34:01.609Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9898ms, resulting in 609 metrics"
time=2025-05-07T12:34:01.610Z level=DEBUG source=collect.go:220 msg="collector logical_disk succeeded after 4.9873ms, resulting in 68 metrics"
time=2025-05-07T12:34:01.611Z level=DEBUG source=collect.go:220 msg="collector system succeeded after 5.9913ms, resulting in 9 metrics"
time=2025-05-07T12:34:01.612Z level=DEBUG source=collect.go:220 msg="collector memory succeeded after 7.0304ms, resulting in 35 metrics"
time=2025-05-07T12:34:01.612Z level=DEBUG source=collect.go:220 msg="collector physical_disk succeeded after 7.0304ms, resulting in 24 metrics"
time=2025-05-07T12:34:01.614Z level=DEBUG source=collect.go:220 msg="collector net succeeded after 8.9903ms, resulting in 27 metrics"
time=2025-05-07T12:34:01.631Z level=DEBUG source=collect.go:220 msg="collector service succeeded after 25.9959ms, resulting in 2872 metrics"
time=2025-05-07T12:34:16.605Z level=DEBUG source=collect.go:220 msg="collector cs succeeded after 0s, resulting in 3 metrics"

Anything else?

No response

[jkroepke]

Start the exporter from an Admin CMD/Powershell and look how it behave.

Some logs can't transport the Windows Event Logs because of how Event Logs are working.

[ajarillo1981]

I have started exporter with Admin Powershell and the behaviour is the same. Now I see the logs "collector succeeded" in Application logs of event viewer and suddenly stop appearing events and URL http://localhost:9182/metrics become unavailable. No other trace. App running and visible in task manager but doing nothing. Do you have any other clue?

[jkroepke]

could you try with --log.level=debug?

[ajarillo1981]

Debug was already configured in config.yaml. However, I have tried --log.level=debug directly in command line and power shell and the result is exactly the same; it collects metrics and they are available in the URL http://localhost:9182/metrics for promethus. After a few minutes no more logs and http is not available anymore, but no log saying why stop working. This is happening in the 2 servers that I manage.

[jkroepke]

If it's stop working, does the program terminate?

Could you reduce the --collectors.enabled to an minimum? e.g. --collectors.enabled=cpu maybe one specific collector the faulty.

Additionally. please add --debug.enabled an open http://localhost:9182/debug/pprof/goroutine and post the content here.

[ajarillo1981]

Tested with the proposed parameteres, just one collector and debug enabled and it still stop working after a couple of minutes:

C:\Users\admin2\Desktop>windows_exporter-0.30.6-amd64.exe --collectors.enabled=cpu --log.level=debug --debug.enabled
time=2025-05-16T09:30:09.278Z level=DEBUG source=main.go:153 msg="logging has Started"
time=2025-05-16T09:30:12.271Z level=INFO source=main.go:265 msg="Running as HYPER1\admin2"
time=2025-05-16T09:30:12.271Z level=INFO source=main.go:185 msg="Enabled collectors: cpu"
time=2025-05-16T09:30:12.272Z level=INFO source=main.go:203 msg="starting windows_exporter in 2.9950023s" version=0.30.6 branch=HEAD revision=db60c78f32185083354f16fb9d534a021f0d85f9 goversion=go1.23.4 builddate=20250406-11:43:27 maxprocs=32
time=2025-05-16T09:30:12.274Z level=INFO source=tls_config.go:347 msg="Listening on" address=[::]:9182
time=2025-05-16T09:30:12.274Z level=INFO source=tls_config.go:350 msg="TLS is disabled." http2=false address=[::]:9182
time=2025-05-16T09:30:15.433Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 2.9971ms, resulting in 609 metrics"
time=2025-05-16T09:30:30.429Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.0006ms, resulting in 609 metrics"
time=2025-05-16T09:30:45.430Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9971ms, resulting in 609 metrics"
time=2025-05-16T09:31:00.431Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 5.0098ms, resulting in 609 metrics"
time=2025-05-16T09:31:15.430Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9869ms, resulting in 609 metrics"
time=2025-05-16T09:31:20.229Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9981ms, resulting in 609 metrics"
time=2025-05-16T09:31:30.428Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 2.9951ms, resulting in 609 metrics"
time=2025-05-16T09:31:40.322Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 4.0001ms, resulting in 609 metrics"
time=2025-05-16T09:31:41.504Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.001ms, resulting in 609 metrics"
time=2025-05-16T09:31:42.386Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.0132ms, resulting in 609 metrics"
time=2025-05-16T09:31:43.168Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.0001ms, resulting in 609 metrics"
time=2025-05-16T09:31:43.856Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 4.0025ms, resulting in 609 metrics"
time=2025-05-16T09:31:45.009Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9947ms, resulting in 609 metrics"
time=2025-05-16T09:31:45.429Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9971ms, resulting in 609 metrics"
time=2025-05-16T09:31:45.696Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.0024ms, resulting in 609 metrics"
time=2025-05-16T09:31:46.497Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9986ms, resulting in 609 metrics"
time=2025-05-16T09:31:47.408Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 4.0367ms, resulting in 609 metrics"
time=2025-05-16T09:31:48.161Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.0025ms, resulting in 609 metrics"
time=2025-05-16T09:31:48.944Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 2.9971ms, resulting in 609 metrics"
time=2025-05-16T09:31:50.448Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.0494ms, resulting in 609 metrics"
time=2025-05-16T09:31:51.345Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9624ms, resulting in 609 metrics"
time=2025-05-16T09:31:52.130Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 4.0015ms, resulting in 609 metrics"
time=2025-05-16T09:31:52.928Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3ms, resulting in 609 metrics"
time=2025-05-16T09:31:53.696Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9971ms, resulting in 609 metrics"
time=2025-05-16T09:31:54.817Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.0015ms, resulting in 609 metrics"
time=2025-05-16T09:31:57.410Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9497ms, resulting in 609 metrics"
time=2025-05-16T09:31:58.112Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 2.9927ms, resulting in 609 metrics"
time=2025-05-16T09:31:58.817Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 4.0001ms, resulting in 609 metrics"
time=2025-05-16T09:31:59.552Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 2.9981ms, resulting in 609 metrics"
time=2025-05-16T09:32:00.337Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 3.9624ms, resulting in 609 metrics"
time=2025-05-16T09:32:00.427Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 2.0503ms, resulting in 609 metrics"
time=2025-05-16T09:32:01.054Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 1.9985ms, resulting in 609 metrics"
time=2025-05-16T09:32:15.427Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 2.9971ms, resulting in 609 metrics"
time=2025-05-16T09:32:30.427Z level=DEBUG source=collect.go:220 msg="collector cpu succeeded after 2.9986ms, resulting in 609 metrics"

I cannot upload the file downloaded here. So I share via wetransfer: https://we.tl/t-oCcjq878cV

Any help would be appreciated.

[jkroepke]

Sorry, I need the text ouput of /debug/pprof/goroutine?debug=1

[ajarillo1981]

Output pasted:

goroutine.txt

Thanks for the effort to solve the issue.

[jkroepke]

Is the output from that path after it has been stop working?

[ajarillo1981]

It was the output when start, now find below the output after it has failed:

goroutine2.txt

[jkroepke]

Thanks, that helps a bit.

Are you using the collect[] while call the metrics endpoint?

And could you gather the output twice in a 10 second interval?

[ajarillo1981]

This is the config that I have in Prometheus:

• job_name: hyper-1
  honor_timestamps: true
  scrape_interval: 15s
  scrape_timeout: 10s
  metrics_path: /metrics
  scheme: http
  follow_redirects: true
  enable_http2: true
  static_configs:
  • targets:
    • hyper1:9182

I have been trying to collect 2 outputs as requested but http server also crash and I cannot get it.
Do you have any idea about where the issue can be? or anything to try in order to find de cause of the crash?

Thank you for your helpl

[jkroepke]

What did you mean by crash? is there a log?

[ajarillo1981]

I mean there is no any logs, stops the logs collector telling "cpu succeded" without no error in cmd debug. The url http://localhost:9182/debug/pprof/goroutine?debug=1 and http://localhost:9182/metrics stops resplying when I access from a web browser. It is like the app is running but crashed. Everything works for a few minutes but after that crash.