Azure VM server shows unhealthy

Hello Charles Daniels - Admin

Thank you for the details.

The "Unhealthy" status you're seeing is an application-level health signal, not an indication that the VM itself is down. Azure marks an instance Unhealthy when the configured health probe (either the Application Health Extension or a Load Balancer health probe) does not receive a valid response on the configured port or request path — even if the VM is fully booted and otherwise operational. For HTTP/HTTPS probes, only an HTTP 200 response is treated as healthy.

The most common cause of this exact pattern — VM restarts, comes back up, but stays Unhealthy — is that the application or service listening on the probe port did not start automatically after the reboot, or started in a failed state.

Could you please help us with the following so we can confirm the cause:

1. A screenshot showing where the "Unhealthy" status is displayed (VM blade, scale set instance list, or Load Balancer backend health).
2. The health probe configuration in use: protocol, port, request path, interval, and threshold.
3. From inside the VM, the output of:
   • Windows: netstat -an | findstr <probePort>
     • Linux: ss -lntp | grep <probePort> The probe port must be in a LISTENING state.
     1. If an HTTP/HTTPS probe is configured, the output of curl -v http://127.0.0.1:<port>/<path> from inside the VM — we need to confirm it returns HTTP 200.
     2. The NSG rules applied to the VM's NIC and subnet. Health probes originate from the Azure platform IP 168.63.129.16, and this address must be allowed inbound on the probe port.
     3. System and Application event logs (or journalctl -b -1 on Linux) covering 30 minutes before and after the restart.
     4. The exact date, time, and timezone of the restart, along with the VM name, resource group, and subscription ID.

In parallel, please also review VM > Resource health > Health history in the portal for the time of the restart. The annotation there will tell us whether the reboot was initiated by the platform, by a user, or from within the guest OS.

Once we have this information, we can confirm the root cause and provide the corrective action.

Thanks,
Suchitra.

Hello @Charles Daniels - Admin

A VM can appear operational after restarting while Azure still reports it as Unhealthy because the health status may be based on a different signal than whether the guest OS is simply running.

I suggest checking the following first:

1. Go to VM → Help → Resource health and review the current status and health history. Azure Resource Health can indicate whether the issue is platform-initiated, such as degraded host hardware, an unexpected host reboot, or another infrastructure event.
2. Check Activity Log around the time the VM restarted to determine whether the restart was customer-, VM-, or platform-initiated.
3. Review Boot diagnostics and the Azure VM Agent/extension status for errors.
4. If you're referring specifically to Application Health, check the configured health probe. Microsoft notes that a VM can be reported Unhealthy when the configured HTTP/HTTPS endpoint doesn't return the expected response or when a TCP health probe cannot successfully connect.

Since you mentioned that the VM restarts by itself, I would pay particular attention to Resource Health and the Activity Log before performing another restart. They can help determine whether Azure initiated the reboot because of an underlying host/platform condition.

If Resource Health reports degraded hardware or recommends a redeployment, follow the remediation shown there. If the VM continues restarting without a corresponding explanation, open an Azure support request and provide the VM resource ID and the timestamps of the unexpected restarts.

Please "Accept the Answer" if this information helped you. This will help us and others in the community as well.