VM dissappears when start fails
- I've recently developed this problem in Hyper-V RTM server where VMs fail to start becuase the synthetic network adapter fails to 'power on'.
This can happen to a VM that has worked fine for some time or a brand new one just created. A message is displayed but nothing is written to the event log so I can't give you the exact text at present (it quite verbose and I'd have to retype it).
The VM stays in the Hyper-V Manager windows with 'starting' state until the display is refeshed and then it is gone. The config files are still there but I haven't been able to find a way to reuse them, they won't import and the manager seems to have no knowledge of them.
Creating a new VM using the leftover VHD is the only solution I have and that may fail starting and delete the new VM too. THis can happen any number of times.
I found that restarting the Hyper-V Virtual Machine Management service and/or the Hyper-V Networking Management service fixes the problem temporarily so I can get a VM started. The problem comes back without warning. I'm at the stage where I'm scared to restart a VM because I might loose it completely and some are not straight forward simple configs.
The other thing that happens is I get all new network connections in the VM so thay have to be reconfigured again each time. Some are up to Local Area Connection #8 now. I can't rename the connection to what I want to call them because the originals are still hidden somewhere (not in device manager hidden devices by the way).
Seems a bone-fide bug since reatrting services make it work again for a while. Didn't happen prior to RTM either.
Hapens most often to VMs that have two virtual networks, one external for user access and one internal to chat amoungst themselves, but not always.
Any ideas how I could permatently fix this?
Cheers,
Mark.
解答
- Separate post says this is AV scanning VHD/XML files. In particular Forefront (which I have) & Trend AV . Exclude those file types from scanning.
http://forums.technet.microsoft.com/en-US/winserverhyperv/thread/fc3a5724-5dc2-44ea-83db-9b7fd8b8eade
所有回覆
- John Howard has a blog posting about the NIC problem when you create a new vm from an existing vhd.
http://blogs.technet.com/jhoward/archive/2008/07/22/hyper-v-why-is-networking-reset-in-my-vm-when-i-copy-a-vhd.aspx
I would not configure each vm with two NICs for this. Multihomed machines (physical or virtual) can cause all sorts of odd problems. A better plan is to configure the vms with just one NIC and connect them to a virtual network. Set up just one vm with two NICs and use that vm to route between the virtual and physical networks. Server 2003 or 2008 with RRAS as a NAT router works fine.
I am not sure what is happening with the "fails to power on" problem. It could be a problem with the NIC driver for the physical NIC in the host. Does it happen in a vm which connects to a virtual switch only?
Bill - Having a virtual switch only connection exascerabtes the problem considerably. Resetting the Hyper-V service in the host temporarily fixes the issue as previously mentioned. Thus, I suspect a Hyper-V issue is the root casue not a driver since other VMs continue to run using the driver sucessfully while this problem is evident.
- Separate post says this is AV scanning VHD/XML files. In particular Forefront (which I have) & Trend AV . Exclude those file types from scanning.
http://forums.technet.microsoft.com/en-US/winserverhyperv/thread/fc3a5724-5dc2-44ea-83db-9b7fd8b8eade
