Problem Description
One day I logged into the server and noticed that the tasks I had kept running with screen were gone. I checked the uptime and found the system boot time didn’t match my expectations. Then I checked the Azure VM Activity log and found the following entry:
| |
The timestamp of this log exactly matched the server’s reboot time.
Next, I went to Operation > Updates and clicked on Update settings at the top. I noticed that Patch orchestration was set to Azure Managed - Safe Deployment.
Changing it to Customer Managed Schedules should solve the problem.
I feel this default setting is quite problematic and really catches people off guard.