Disorganized Disaster Recovery?
January 20, 2014
Ever notice how airlines board passengers on a plane? Some board by zone number, some by seating sections, etc. but in essence, rather than asking passengers to board whenever they are ready (stampede approach), most airlines try to make the process more efficient by using a grouping system.
Along the same line of thought, does your IT Disaster Recovery plan group business applications and systems in Tiers of recovery or are they recovered in a less structured manner? With a recent increase in extreme weather events around North America such the Calgary flooding, Denver area floods, California wildfires and Eastern US arctic blast, many companies have had to exercise their DR Plans and we can expect more challenges in the future.
Defining Tiers of recovery doesn’t have to be overly complex and you don’t need more than 4 or 5. Tiers are not defined based on characteristics of the IT systems and application, but rather, based on the criticality of the business process they support. Tiers are defined based on common attributes such as RTO/RPO, impact of an outage, dependencies, etc. which enables a structured and predictable recovery based on business priority.
You may already have a good idea of what your key systems are based on past outage experiences – but how about the rest of them? Grouping IT systems in Tiers of recovery helps with the design and sizing of the technology solution to support recovery; while a more traditional recovery from backups might be adequate for low criticality systems, a small number of highly critical systems may require a failover solution. Tiers help delineate which systems benefit from a specific recovery capability rather than attempting to deploy a potentially costly or mostly inadequate “one size fits all” recovery solution. Tiers of recovery also help budgeting; you may not have the luxury of addressing DR for all IT systems in a single funding effort. With Tiers, you can focus efforts on the most critical systems first.
Will your IT recovery effort be like a well-organized plane boarding process with on time departure or more along the lines of a stampede to the best seats?