Resource Health reports server as Unavailable despite successful Arc heartbeats (OpenTelemetry)

Hi,

Anyone else started to have issues with Azure Resource Health or experienced similar issue?

Recently (like 2 days ago) we started to get alerts via our Activity log alert rule connected to our Arc enabled servers. Looking in Azure Monitor we can see multiple down alerts for all on-premises Arc enabled servers repeating through out the day despite the servers are online and sending Arc heartbeats as expected.

Our servers are configured to use the OpenTelemetry metrics (Using Azure Monitor Workspace, DCR and Prometheus)

Looking at the Resource Health history for a server we can see flapping (availabilityState) between Available and Unavailable state. This despite severs being operational, net connectivity loss etc.

az rest --method get --url "https://management.azure.com/subscriptions/<sub>/resourceGroups/<rg>/providers/Microsoft.HybridCompute/machines/<server>/providers/Microsoft.ResourceHealth/availabilityStatuses?api-version=2025-05-01" -o jsonc

"value": [
    {
      "id": "/subscriptions/<sub>/resourceGroups/<rg>/providers/Microsoft.HybridCompute/machines/<arc server>/providers/Microsoft.ResourceHealth/availabilityStatuses/current",
      "location": "westeurope",
      "name": "current",
      "properties": {
        "availabilityState": "Available",
        "category": "Not Applicable",
        "context": "Not Applicable",
        "occuredTime": "2026-09-24T10:51:58Z",
        "reasonChronicity": "Transient",
        "reasonType": "Planned",
        "reportedTime": "2026-09-24T11:35:24.0626821Z",
        "summary": "There aren't any known Azure platform problems affecting this machine.",
        "title": "Available"
      },
      "type": "Microsoft.ResourceHealth/AvailabilityStatuses"
    },
    {
      "id": "/subscriptions/<sub>/resourceGroups/<rg>/providers/Microsoft.HybridCompute/machines/<arc server>/providers/Microsoft.ResourceHealth/availabilityStatuses/2026-09-24+09%3a46%3a53Z",
      "location": "westeurope",
      "name": "2026-09-24+09%3a46%3a53Z",
      "properties": {
        "availabilityState": "Unavailable",
        "category": "Not Applicable",
        "context": "Not Applicable",
        "occuredTime": "2026-09-24T09:46:53Z",
        "reasonChronicity": "Transient",
        "reasonType": "Unplanned",
        "resolutionETA": "2026-09-24T10:06:53Z",
        "summary": "We are sorry, your resource is unavailable.",
        "title": "Unavailable"
      },
      "type": "Microsoft.ResourceHealth/AvailabilityStatuses"     } 	
... 
]

For the same time frame, on the server and looking in the himds.log we can see that the server is sending hearbeats and gets 200 responses when Resource Health is saying the server is Unavailable. Availability and other metrics shows all is OK.

time="2026-09-24T08:54:16+02:00" level=info msg="Applied user-assigned identity heartbeat diff: version=<nil>, additions=0, deletions=0." additions=0 deletions=0 location=uami.ApplyHeartbeatDiff serverList=0 type=Security version="<nil>"
time="2026-09-24T08:54:16+02:00" level=info msg="Exiting DoHeartbeat."
time="2026-09-24T08:59:19+02:00" level=debug msg="Send HeartBeat to service via HTTP PATCH @ https://gbl.his.arc.azure.com/neu/his/machine/<guid>/metadata?api-version=4.0-preview&location=northeurope"
time="2026-09-24T08:59:19+02:00" level=info msg="Applied user-assigned identity heartbeat diff: version=<nil>, additions=0, deletions=0." additions=0 deletions=0 location=uami.ApplyHeartbeatDiff serverList=0 type=Security version="<nil>"
timtime="2026-09-24T08:59:19+02:00" level=debug msg="Received response from service with StatusCode - 200"
time="2026-09-24T08:59:19+02:00" level=info msg="Applied user-assigned identity heartbeat diff: version=<nil>, additions=0, deletions=0." additions=0 deletions=0 location=uami.ApplyHeartbeatDiff serverList=0 type=Security version="<nil>"
time="2026-09-24T08:59:19+02:00" level=info msg="Exiting DoHeartbeat."

As mentioned this happens for all servers and is repeating every hour or second hour randomly. We have updated the agents but without any difference.