Virtualization Delivers Disaster Recovery Capabilities and Saves You Money

Brian R. Hoffman

August 28, 2008

Disaster recovery (DR) is taken much more seriously today than prior to 9/11, and the majority of today's companies have some kind of DR plan. Plans range from recovery via tape backup to an outsourced facility with infrastructure capacity provisioned at the time of disaster. Other plans involve dual data centers running production applications capable of redirecting the entire workload in the event of a data center failure. Today, even these traditional DR solutions are becoming less common as VMware succeeds in convincing the Intel community, which is considered the industry standard by most IT professionals, that virtualization is the answer.

"Virtualization" has been the buzz word for the last few years, and yet it seems to mean something different to everyone. The funny thing is virtualization isn't new; it has been around for more than 30 years. VMware popularized virtualization when it was able to capitalize on windows operating system inefficiencies and was able to develop a market for consolidation. Now that the majority of consolidations have been done, and VMware has released their hypervisor for free, VMware has begun to focus on other applications for virtual machines.

While this article focuses primarily on Intel environments, it is important to note that mainframe operating systems have been "virtualized" since the 1960's, and Unix has supported virtual machines in the form of virtual partitions for years.

When you combine virtualization with advanced storage technology (e.g. storage area network (SAN), network boot, replication, clones/snapshots), you can create a DR solution that is not only fully functional, but can be tested as often as changes are made. No longer are you relegated to annual or semiannual test week, spinning tapes, configuring servers, networks and storage and data that only test a small fraction of your applications. Through virtual machines - booted from network volumes, replicated to the DR location and made available to another virtual server - entire enterprises can be recovered in minutes.

This is the recipe for DR capability:

1.    Diskless servers
2.    Network boot
3.    Replication of virtual machines to DR location
4.    Replication of data to DR location
5.    Leverage "hot snapshots" for testing (each storage vendor has their own nomenclature)

Cost savings
But if an effective disaster recovery system is not enough of a reason to virtualize, consider the cost savings. The average power consumption of an imbedded 2.5"/10,000 revolutions per minute (RPM) drive is 11.2 watts (IBM). The additional, unnecessary, power consumption by a server with two drives barely used seems small at roughly 22 watts. But as an example, a Dell 860 server consumes 110 watts. If you boot from network storage and eliminate the 2 internal disks, power consumption is reduced by 22 watts or 20 percent. This does not factor the additional power you save by not cooling the additional heat generated by the drives or cost of the actual drives. Multiply the power saved from all your servers and it stands to be a significant reduction in cost.

Aside from the reduced operational cost of the data center, there are many operational advantages that make a strong argument for this case. You can find some of the benefits listed in any storage vendor's material. The below is from a Dell whitepaper (Understanding Boot-From-SAN Technology, 2005):

Boot-from-SAN benefits include:

1.    Improved disaster tolerance
2.    Centralized administration
3.    Reduced total cost of ownership (TCO) through diskless servers
4.    High-availability storage
5.    Enhanced business continuance
6.    Rapid server repurposing
7.    Consolidation of image management

Additional benefits of diskless servers are:

1.    Less power consumed per server
2.    Less heat generated
3.    Higher disk utilization (eliminates stranded disk)
4.    Increased server reliability
5.    Reduced cost of the server

With the ability of several storage arrays to "de-dupe", and the fact that a boot volume is 99 percent read only, the reduction in spindles is enormous. One thousand servers with two disks (mirrored) can be reduced from 2,000 spindles to two spindles in most cases.

Virtualization has been a deservedly hot topic of late. With proper implementation, it can save your company money, conserve a lot of energy and provide an efficient and effective disaster recovery system.

Brian Hoffman, vice president/managing director of the technology solutions group at Corus Group, LLC, has nearly 30 years of IT experience in data center operations, physical buildouts, data center consolidations and moves, virtualization and more. At Corus, his group is responsible for all aspects of data center operations consulting, including data center moves, consolidations, efficiency assessments, storage assessments, disaster recovery planning and operations. He can be reached at bhoffman@corusllc.com and/or 770-840-5017. For more information visit www.corusllc.com.