hcnDeleteNamespace failed in Win32: The specified request is unsupported (0x803b0015)

[jsturtevant]

We are seeing errors in the upstream Kubernetes tests that have this error in containerd. I've posted some more details in kubernetes/kubernetes#125195 (comment)

We are also seeing this error in the logs:

``
[1429] {"Timestamp":"2024-05-29T00:52:27.9870075+00:00","ProviderName":"ContainerD","Level":4,"Keywords":0,"Payload":{"ID":1,"EventName":"LogrusEntry","TraceEventProcessID":5084,"TraceEventProcessName":"","Message":"Container to stop "b6eacb5bab6ce42f8f0d173a9b0fae2175623c982adbf2cd8f3b9f89825e8215" must be in running or unknown state, current state "CONTAINER_EXITED"","Time":"2024-05-29T00:52:28.0224034+00:00"}}

The hns service crashed:

5/29/2024 12:21:00 AM 7031 Error The Host Network Service service terminated
unexpectedly. It has done this 1 time(s). The
following corrective action will be taken in 120000
milliseconds: Restart the service.

[jsturtevant]

We have setup crash dump reporting but it appears the service didn't generate dumps this time

[marosset]

kubernetes-sigs/windows-testing#438 should enable (or re-enable) crash dump collection for most of our Windows CI jobs in K8s. I really thought this was already enabled...