VM machine automatically rebooted

Question

hostname kernel: [1009087.849187] hv_utils: Shutdown request received - graceful shutdown initiated

That's a production server and therefore my celery jobs went down. I need an explanation on why there were no warnings. Plus, looking at the Azure Health Dashboard everything looked ok.

Facing about issue 3rd - 4th times in a month.

Please provide an explanation ASAP.

Answer 1

Hello, @krishna wattamwar !
I'm sorry to hear that your VM is rebooting unexpectedly. I'll summarize the list of causes down below but if you've already checked Azure Health and this is the 3rd or 4th time you are seeing this over the course of a month, I would check both the activity log and the system log to make sure that nothing stands out and then confirm that the workload/role on the VM is not triggering a crash/restart. If both of those look normal, then the next step would be to create a support ticket to take a closer look and I'd be happy to help you with that process.

Why did my VM reboot unexpectedly?
There are a list of reasons why your VM might reboot (as well as a quick reminder that configuring your VM for high availability is the best way to protect against reboots and downtime) and the following documentation covers this scenario in detail:

Understand a system reboot for Azure VM

Azure virtual machines (VMs) might sometimes reboot for no apparent reason, without evidence of your having initiated the reboot operation. This article lists the actions and events that can cause VMs to reboot and provides insight into how to avoid unexpected reboot issues or reduce the impact of such issues.

• Check resource health information (you can skip this as you've already done it)
• Check the activity log (note that VM downtimes won't show in the activity log when a VM is created or migrated to a new host, if the VM availability state changes from Available to Unavailable and back within 35 seconds, or if the VM health changes from a state to Unknown and back to the original state as this is filtered out).
• Planned maintenance sometimes requires a reboot. You can read more about planned maintenance for VMs as well as how to schedule planned maintenance.
• Memory-preserving updates should not have an impact on users however the VM will enter the paused state and then be resumed typically within 30 seconds.
• User-initiated reboot/shutdown is in the activity log if it is initiated from Azure and in the system logs if it's initiated in the OS.
• Microsoft Defender for Cloud and Windows will apply updates as configured.
• Host server faults can cause a reboot as the host server attempts a recovery (usually caused by hardware failure).
• Auto-recovery is rare but needed if a faulty host needs to be taken out of rotation and the VMs are automatically relocated to a different, healthy host server.
• VM crashes can be caused by the VM itself if the workload or role triggers a bug check.
• Storage-related forced shutdowns will happen if the availability or connectivity is affected for more than 120 seconds.
• Check the Azure Service Health Dashboard (link) to see if something has affected VMs in your datacenter (rare).
• Diagnose VM restarts using the Diagnose and Solve blade on the VM blade to run additional diagnostics.
• Contact support by creating a support request if you've gone through everything and still are unable to determine why the VM is restarting.