azure virtual machine disk errors

Question

I've been using 3 identical VMs on Azure for a month or more without problem.

Today I couldn't Remote Desktop to one of them, and restarted it from the Azure Portal. That took a long time. It eventually came back up, and the Event log has numerous entries such as:

"The IO operation at logical block address 70 for Disk 0 ..... was retried"

"Windows cannot access the file C:\windows\Microsoft.Net\v4.0.30319\clrjit.dll for one of the following reasons, network, disk etc.

There are lots of errors like this. To me they seem symptomatic that the underlying disk system is having serious problems. Given the VHD is stored in a triple replicated Azure blob, I would have thought there was some immunity to this kind of thing?

Many hours later it's still doing the same thing. It works fine for a few hours, then slows to a crawl with the Event log containing lots of disk problems. I can upload screen shots of the event log if people are interested.

This is a pretty vanilla VM, I'm only using the one OS disk it came with. The other two identical VMs in the same region are fine.

Just wondering if anybody has seen this before with Azure VMs and how to safeguard against it, or recover from it.

Thanks.

Answer 1

Thank you for providing all the details and we apologize for the inconvenience. We have investigated the failures and determined that they were caused by a platform issue. Your virtual machine’s disk does not have any problems and therefore you should be able to continue using it as is.