Hi everyone, can you tell me if this makes sense and what you might try to do about it?
My environment is Server 2012 R2 on a Dell Poweredge Rack server. There is just one MS Hyper-V VM hosted on it which is running Exchange 2013 also on the Server 2012 R2 O/S. The host server is also a file share and AD, DHCP, DNS, etc. (I know that is not ideal) There is 8 gig of ram and 1 processor assigned to the VM. The host has 32 gig total with plenty of free space. The servers have been running great for about 8 months.
The Exchange 2013 server stopped working over the weekend. By stopped working I mean mail stopped coming in and Outlook running on the desktops showed outlook as disconnected. According to the VM console it was still running and I could log into it but found the performance to be brutally slow. So slow in fact that it took 10 minutes or more just to open mmc and look at services. CPU was spiked at 100% There are no disk errors or anything like that in the logs of either the VM or the host.
So when the problem was first discovered. I rebooted just Exchange. Problem continued - Exchange services were not started (they did not respond in a timely fashion) and the CPU was inexplicably at 100% utilization. The Noderunner, MMC and Task Manager were blamed.
Rebooted both the VM and the host. No change. What could be going on with this server??
What I did find was that the server had run LOW on disk space. The total volume size was about 250gb and there was only 17gb left. It would make sense that Exchange would shut down for this reason as it requires a certain % of free space to stay running.
I tried booting safe mode. I thought maybe that would rebuild paging files if they were corrupt, etc. No change.
Next I increased the size of the VM disk in Hyper-v, then increased the size of the C: partition (the VM has only one partition) and that left about 150gb free of 400gb. Rebooted the server and was surprised to find it in the same state! Still no exchange, and still with a maxed out CPU.
I'm in the process of restoring a pre-crash backup of the virtual hard drive. This will set them back a few days in email.
I don't think of this as an exchange problem because none of the exchange services are running and CPU is still maxed.
Does anyone have any suggestions as to what I might try to restore server functionality before I take the drastic step of going to my VM Hard drive backup?
I'm hesitant to add processor as I'm afraid that will just get sucked up to 100% too.