Cloud Disaster Recovery still Needs to be Analysed and Tested

While it’s comforting to think that a professional organisation can now hold your data safe and sound for you in the cloud, cloud DR planning still needs the same careful attention as a solution using any other technology. There are significant business advantages available, not least in terms of financial flexibility and hugely scalable resources, but basic risk analysis and identification now can save tears later. Think business common sense, common understanding and agreement with providers, and above all proper testing.

What experience does the cloud provider have of disaster recovery?

Global computing resources and network connectivity does not necessarily equate to a capability to provide disaster recovery facilities. The current strategy of some large IT distributors to bundle together service offerings of cloud compute power with specific DR applications is an indication that they too recognise that it takes more than just multiple data centres to be properly prepared.

What expectations are they prepared to live up to?

The key elements of DR also still need to be defined and agreed on: for example, the maximum time you can tolerate for your organisation to be without an application (your RTO – recovery time objective), and the maximum amount, in terms of time period, of data you can stand losing, if any (your RPO – recovery point objective).

“First we try, then we trust”

This line from the film “Entrapment” is highly relevant to DR planning in general, and to cloud DR planning in particular. It doesn’t matter what the cloud provider’s reputation, specification or even contractual promises are – you still need to carry out a realistic, practical DR simulation to check that if it comes to the crunch, your organisation can really continue to function properly.