Protecting VMs against host power-offs

Question

Every now and then and for numerous reasons, my Windows 8 desktop encounters a loss of power. I am running a Debian 6 virtual machine via HyperV and when I restore power, the VM kernel panics every single time. Rebooting the VM results in the disk being
set to read-only mode, and fixing it takes quite a long time. I have a number of snapshots from states such as these.
Is there a feature of HyperV that makes VMs more tolerant to sudden host power-offs? I could script something together to take periodic snapshots of my VMs and revert to those should I encounter a loss of power, but I wonder if there is something already
present in HyperV to facilitate such protection.
I have one of the simplest of simplest setups: Windows 8, HyperV with two VMs (Solaris and Debian), each with a single attached drive and network adaptor. I have no clustering setup nor do I use any HyperV integration (the Debian kernel module leads to more
kernel panics, and I don't have time to debug that at the moment).
Is there something already present, or should I just write a script to take periodic snapshots?
Edit: I now realize this is the Windows Server forums, yet my setup uses Windows 8. Is there a more appropriate forum for my post?

Answer 1

Every now and then and for numerous reasons, my Windows 8 desktop encounters a loss of power. I am running a Debian 6 virtual machine via HyperV and when I restore power, the VM kernel panics every single time. Rebooting the VM results in the disk being
set to read-only mode, and fixing it takes quite a long time. I have a number of snapshots from states such as these.
Is there a feature of HyperV that makes VMs more tolerant to sudden host power-offs? I could script something together to take periodic snapshots of my VMs and revert to those should I encounter a loss of power, but I wonder if there is something already
present in HyperV to facilitate such protection.
I have one of the simplest of simplest setups: Windows 8, HyperV with two VMs (Solaris and Debian), each with a single attached drive and network adaptor. I have no clustering setup nor do I use any HyperV integration (the Debian kernel module leads to more
kernel panics, and I don't have time to debug that at the moment).
Is there something already present, or should I just write a script to take periodic snapshots?
Edit: I now realize this is the Windows Server forums, yet my setup uses Windows 8. Is there a more appropriate forum for my post?
The only reliable way to go would be:
1) having at least two individual hosts running server versions of Windows with Hyper-V role enabled (technically you can have ANY hypervisor but this is Windows forum)
2) configure guest VM cluster between VMs you have (yes, that would require some sort of a physical or virtual shared storage, typically...)
That would be very low RTO (virtually no downtime) and RPO (virtually no data loss). See:
Guest VM Cluster for HA
http://technet.microsoft.com/en-us/library/dn440540.aspx
Using some third-party software like VEEAM will get you reasonable RTO and RPO but no transparent failover. This one is not free but may work with a desktop version of Windows (never checked). Free Hyper-V is definitely supported. See:
VEEAM Backup & Replication
http://www.veeam.com/blog/rdp-rsat-free-hyper-v-server-2012-r2-remote-desktop-appliance-linux.html
Hyper-V Replica may also work (2012 R2 req'd). See:
Hyper-V Replica Explained
http://blogs.technet.com/b/yungchou/archive/2013/01/10/hyper-v-replica-explained.aspx?utm_source=feedburner
Hyper-V Replica Overview
http://technet.microsoft.com/en-us/library/jj134172.aspx
Good luck!
StarWind VSAN [Virtual SAN] clusters Hyper-V without SAS, Fibre Channel, SMB 3.0 or iSCSI, uses Ethernet to mirror internally mounted SATA disks between hosts.