We have an Azure VM (size A2 Standard, ie. 2 Cores, 3.5 GB RAM) which worked fine for almost a month. But since a few days, at least once a day it keeps failing and is not reachable via RDP any more. Also the our service running on this VM is not reachable any more. I have to restart the VM via the management portal, which takes unusually long (at least half an hour, until the server is reachable again).
When I compare our own application logs with the windows event log, at the exact same time our log stops, there are also some warnings like the following:
The IO operation at logical block address 0x5bf7d8 for Disk 0 (PDO name: \Device\Ide\IdeDeviceP0T0L0-0) was retried. (Event ID 153)
and
The system failed to flush data to the transaction log. Corruption may occur in VolumeId: C:, DeviceName: \Device\HarddiskVolume1. (The I/O device reported an I/O error.) (Event ID 140)I also tried resizing and restarting the virtual machine, like it was suggested in an other thread in the forum, but the error persists. I also tried to stop our service and created a simple script which just writes the datetime into a textfile every two minutes. Didn't help much, either. Although the VM is reachable via RDP, the generated file is not accessible and there are disk errors in the event log. The datetime log stops at the same time as the disk errors occur.
Could it help if I duplicate this VM or should I try creating it completely from scratch?
Any help is appreciated. Thanks in advance.