Lots of warnings: No Network Stats for container

[bip7686]

I'm running the windows exporter through wins for my kubernetes cluster. The windows exporter is printing a lot of messages like:

time="2021-06-15T18:28:29+02:00" level=info msg="No Network Stats for container: 52026722c95ef2d36d3bc894dfe4ee1e910277c315e4878fa43b3dddc533e5e2" source="container.go:231"

WMI Setup:

time="2021-06-15T14:26:06+02:00" level=info msg="Enabled collectors: cache, system, os, cpu, memory, logical_disk, cs, container, service, net" source="exporter.go:347"
time="2021-06-15T14:26:06+02:00" level=info msg="Starting windows_exporter (version=0.16.0, branch=master, revision=f316d81d50738eb0410b0748c5dcdc6874afe95a)" source="exporter.go:399"
time="2021-06-15T14:26:06+02:00" level=info msg="Build context (go=go1.15.6, user=appveyor-vm\\appveyor@appveyor-vm, date=20210225-10:52:19)" source="exporter.go:400"
time="2021-06-15T14:26:06+02:00" level=info msg="Starting server on 0.0.0.0:9100" source="exporter.go:403"
time="2021-06-15T14:26:06+02:00" level=info msg="TLS is disabled." source="gokit_adapter.go:38"
time="2021-06-15T14:26:06+02:00" level=error msg="Failed to start service: The service process could not connect to the service controller." source="exporter.go:337"

Windows Node: BuildNumber: 17763 Version: 10.0.17763

[carlpett]

Just to be clear, is this about missing data that should be there, or about excessive logging? Does the containers actually have network interfaces attached to them?

[bip7686]

The data seems to be there. How would I check if the containers have network interfaces attached to them?

[jsturtevant]

checkout how this can be done with hostprocess containers. There are also examples on wiring all this up with the rest of the Prometheus stack in https://github.com/prometheus-operator/kube-prometheus/blob/main/docs/windows.md

[dineshsharmads]

This is still an issue even when exporter is running as hostprocess. I have deployed the exporter using hostprocess to few AKS clusters integrated with kube-prometheus.

https://github.com/prometheus-operator/kube-prometheus/pull/2048

I can see cpu and memory metrics for containers in Windows dashboards but nothing for network.

time="2023-03-21T14:01:48Z" level=info msg="No Network Stats for container: fc7bcf92522b968ccc1f4cde542c37360a9cbbb1bec9ef2e85be29f2bc6435b0" source="container.go:262" time="2023-03-21T14:01:48Z" level=info msg="No Network Stats for container: fd803dea296ebb94722e95691696bb056d6b2aec18779d3a55bd33121487bd3b" source="container.go:262" time="2023-03-21T14:01:48Z" level=info msg="No Network Stats for container: febe4ed5b0a9e67f717f230c15e70f70f29ffefb815b3db17ebef06d09abd5fa" source="container.go:262"

[jsturtevant]

@dineshsharmads See https://github.com/prometheus-community/windows_exporter/issues/1129 which .

Are you using contianerd? The issue for networking is related to how stats are queried in HCS v1/v2. Let's continue the conversation in #1129