IT is great at some things, but out of its league in many cases. Business continuity planning is an example of the latter: No matter how well we set up our applications and systems, the human element is always a roadblock. Sure, we can build a complex system to return our CRM system to operation in Duluth, but will anyone be able to use it? Even the best disaster recovery (DR) infrastructure is useless without a business continuity (BC) strategy for everything else.
All IT can offer is to do its best to hold up its side of the deal. IT can design systems with return-to-operations in mind, replicating data and documenting configurations. IT can deploy remote systems and keep them warm and ready should we need them. And IT can create operational plans to rapidly get everything working when disaster strikes.
Although technology alone cannot solve the BC/DR conundrum, new technical solutions to help close the gap do occasionally appear. Data replication was one such key technology, as was server virtualization. Cloud computing will soon be added to the BC/DR hot list.
What do cloud computing and cloud storage services offer to help DR?
- Cloud resources are inherently flexible, giving needed capacity on demand. This is especially important for compute resources, since BC operations often have unpredictable usage spikes as systems come online and resume operations.
- Cloud resources scale based on usage, reducing the expense when there is no disaster. This is one of the main reasons companies don't invest in disaster recovery capacity: It's so expensive on a daily basis "just" to be prepared!
- Cloud resources are available anywhere. Rather than trying to keep displaced employees in close proximity to technology, public cloud systems can be used from anywhere during a disaster.
This last bit is especially critical. Most disaster recovery plans include hours spent bringing up replacement equipment, restoring data, and provisioning access. Steve Duplessie of ESG says cloud and physical presence can partner in this way:
"Perhaps the real opportunity for the cloud will be not only the way it enables the masses to keep different copies of data in different locations, but that it can provide a realistic set of options that enables a business to bring up their applications quickly and seamlessly, provide the operational knowledge sophistication to sustain the effort, and also be smart enough to provide for the non-IT services."
In other words, cloud resources don't require a time-consuming recovery effort. It just works. But the best cloud-based DR solutions also include consideration of how people will access these systems. Private in-house cloud systems will still need to consider enabling access in the event of a disaster. But public cloud services are inherently accessible over the public Internet, allowing employees to get up and running wherever they have connectivity rather than being relocated.
What if your CRM application could be brought up on Amazon's or Rackspace's cloud server infrastructure? What if your image or document repository was accessible through the Nirvanix SDN? You could maintain a hot-standby system for very little recurring cost and turn it up instantly when a disaster struck! But watch out! You might get spoiled by the ease of use and low cost of these services and decide to move your production systems there, too!
This won't work for some applications, of course, but many systems don't require close proximity or special hardware. Any worker who can effectively telecommute can be productive in this public cloud DR scenario.