Those moving to cloud computing have their work cut out of for them. They need to pick a parcel of data, applications, or both to migrate to a cloud-based service. Or, perhaps build a system from the ground up on a cloud platform.
In any event, you need a few things to insure success, including a good architecture, a deployment plan, and a sound data integration strategy.
Those who take a pass on the data integration strategy often tell themselves that the amount of data that needs to flow in and out of the system is so small that a data integration plan is not needed. Or, they assume that the migrated or new cloud-based system will always be an island unto itself. We all know that the volumes of data will increase over time, and business systems rarely have value if they are designed to be a silo.
So, what are the core reasons you need to consider a data integration strategy when moving to cloud computing?
First, the cost of retrofitting a sound data integration plan and technology after the fact is a great deal more than if you planned the use of the technology beforehand. Many assume that data integration, security, and governance can be something that’s dealt with later, perhaps after a few months of operations. However, the costs skyrocket when you factor in the amount of work, and more importantly, the risk to the business.
Second, if cloud computing is indeed a direction your enterprise needs to take, then creating a data integration strategy is something that becomes of value to a number of applications and databases, thus it becomes strategic. I don’t recommend that you design data integration approaches and technologies to eliminate a single silo, but create a strategy that rolls up into the enterprise architecture. In other words, solve many problems once, if you can.
Third, data integration strategies are typically bound to security and governance strategies, thus you need to define data integration approaches and mechanisms along with your approach to both security and governance. I’ve seen many organizations attempt to define security, governance, and data integration as standalone strategies, and then attempt to link them down the line. You’ll find that this leads to many costly mistakes. Think about the amount of dependencies between these concepts, especially when you consider the importance of security and governance in the world of cloud computing.
Finally, you need to consider the value of data integration technology at the planning stages to properly consider the business case for cloud computing. Without a plan as to how the data will find its way into the cloud, or from the cloud back to the enterprise, the use of emerging cloud computing technology has a much-diminished value.