CloudBursting

Last week Jeff Barr from Amazon Web services came up with a great new term to describe the need to handle sudden and extreme spikes in demand by enabling a hybrid cloud computing model which combines both private data center resources and remote cloud resources such as Amazon Ec2. He called this model "cloudbursting". This term seems to be quickly becoming the preferred way to describe a major problem or for some the perfect "opportunity" in the emerging cloud computing space (myself included).

In doing some research earlier I came across the original definition of a cloudbust. In meteorology a cloudburst is an extreme form of rainfall, which normally lasts no longer than a few minutes but is capable of creating flood conditions. Similarly in IT a sudden and unexpected rise in demand can quickly overwhelm a data center. In coming up with the term cloudbursting, Jeff has give a simple name to a rather complex problem. At Enomaly this is a problem we've been debating for awhile; How do you effectively enable a kind of cloud overflow in a secure yet efficient manor?

Provisioning instances in Amazon EC2 for example is relatively easy, moving live workloads across a wide area is not. In most modern dynamic applications the idea of having a "hot cloud standby" or a prebuilt virtual machine that is basically waiting in the wings would solve a lot of problems. But in reality there are a number of complexities that need to be overcome. These complexities range from network optimization to secure data transfer & replication to load balancing across geographically diverse hosting environments, just to name a few.

To truly enable a capable cloudbursting infrastructure, I feel there needs to be a common consensus on how this may be archived and by what means. So the question in the short term is; what are some of the practical approaches, technologies and architectures needed to make this kind of hybrid cloud infrastructure feasible?