ToutVirtual Enterprise Virtualization - A Proper Virtual World

Short Distance Disaster Recovery?

August 8th, 2007

Virtualization, Fine, Well Sort Of? - Chapter 02

Many of you are thinking, I hope, that you know what short distance disaster recovery is. The concept should be familiar, even if the term short distance disaster recovery is not. A few typical scenarios would be:

• The fire sprinklers went off downstairs in the legal department, the entire floor has now been declared a wetland area. Since all lawyers are sharks? Is this really a problem?
• The construction next door to your most productive branch office just cut all the power and/or connectivity cables. This happens more often than one might think.
• Heads up, John Doe associate, in the security compliance department, walks into your office and says that he just erased all the files on the file server that do not match corporate policy, and every other file is missing for some reason.

Of course there are as many possible scenarios as you can devise, however that is not the point. As many reading the above scenarios, have already gathered I am sure, these disasters do not represent significant loss of infrastructure, the building is intact, computing resources still work, etc. Each of these issues should be met by the official no problem comment from your respective information technology team. Or it should be that straight-forward.

So where does the term short distance come from? Short distance disaster recovery scenarios or situations are the class of issues where you move access to data around the problem, or recover access to the data with minimum effort. Moving people to alternate workspace in the same site, rerouting network traffic around an issue, or restoring data from alternate storage will be easy from a computing infrastructure perspective, providing proper planning and implementation is done.  Successful virtualization makes this critical! Redundancy and virtualization go hand-in-hand. This is not to say that all this does not cost time or money, it does, but remember the old axiom? Everything spent now, saves even more later on, duh!

What makes short distance recovery painless? The following points are keys to successful short distance recovery:

• Always have a stand-by virtual host, a total production host server plus one (1), is a common approach. This does not mean this server is idle, it is often used as the staging server for create new virtual machines for example, or a safe location to test updates to the platform, or destination for, cough, when you have to go to auxiliary storage recovery methods.
• Always have redundant paths in both the communication and storage networks. It is common to advise clients to have north/south, or east/west network trunk lines to the same building, and if possible have different network vendors for each, if the site is critical enough for such expense. Does not hurt to have the vendors know they are both in play as well, right?
• Always have virtual instance snapshots, storage array snapshots, or some type of recent archival of the virtual instances, or at least critical data files per virtual instance in a secure location in the same site, but not necessarily the same room or building in the site.
• And the most import key requirement of all, make sure your computer infrastructure staff practice all contingency procedures and such procedures are in complete alignment with all recovery policies. Publish. Train. Validate. Train. Certify. Train. Evaluate. Train. Get the point?

In virtualization, short distance disaster recovery is often integrated by design of the product, and should be easy to implement and use. Since virtualization infrastructure should already be redundant, in reference to servers and power, networks paths, storage paths and allocated storage, these unforeseen but relatively minimal infrastructure issues should be easy to survive. Live migration technologies, high availability, and to some extent resource sharing features, are applicable to short distance recovery, and must have this same redundancy to function. The more transparent the impact to the client, the more successful your short distance disaster recovery is, after all the most painful part of the recovery is not in the technology, it is in your ears, as your clients are yelling about productivity lost, no?

A Proper Virtual World, alternate workspace, best practices, disaster recovery, pre provisioned virtualization, redundancy, short distance, short distance disaster recovery, successful short distance recovery, Virtual System Management, virtualization, virtualization 201, virtualization planning, virtualization savings, virtualization use case

Entry Filed under: A Proper Virtual World, Virtual System Management