Virtualization - The Ultimate in Disaster Recovery

Virtualization is one of the fastest growing areas in computer networking.  The ability to run multiple virtual machines on one box is something that has obvious practical advantages for larger corporations running dozens of servers.  Buying one powerful server that can be broken up into multiple boxes can reduce overall hardware costs and improve power efficiency.  It also makes it easy to make lab environments and test things on a virtual computer before rolling it out into your network infrastructure.  Many I.T. professionals point to large enterprises as being the primary target group for virtualization.  After all what would be the advantage of having a virtual server if you only have one server to begin with?

Well there are actually some legitimate reasons why you might consider implementing this.  The first and strongest argument for virtualization in my opinion is recoverbility.  Virtual PC's essentially store all of their data in VHD files on the computer.  VHD files are basically virtual hard drives.  So the virtual computer treats this VHD file exactly like it were it's hard drive.  By backing up this file on a regular basis we can ensure that you can recover your data much faster than standard back up solutions.

Suppose we have two servers, one server has a virtualized server running on it, the other is just a standard server.  Now suppose that both of those servers suffer a major hardware failure and need to be rebuilt.  

With the virtualized server we simply need to load an operating system, install VMware or Microsoft's Hyper-V and point the virtualization software at the backed up VHD file.  That's it, no additional configuration is needed because it's as if we just plugged in a clone of the original drive.

With a non-virtualized server you will need to load your operating system, do all the system configuration (assign I.P. address, reinstall software, configure your domain, etc..) and then start to restore your standard back ups. 

The end result is that a virtualized server can be back up and running inside of an hour where a non-virtualized server could take most of the day to get back up and running.  Now if you have two servers both running virtualization software we would just move the broken server's virtual machine over to the other server and you'd be back up and running with virtually no down time.  No pun intended.

Now all this comes at a cost.  Virtualization requires that the virtualized computer has it's own legitimate operating system, which increases costs for each virtual machine you are running.  But for some businesses the added cost is worth the security.