I spoke at the Cloud Computing Expo this week in New York. As you would think, based on the hype, the event we well attended.
The best part of events like these is not the presentations, but sitting around and speaking with those rank-and-file organizations looking to move to the clouds. I gather the best data points that way. As you would guess, most are at the planning stages, or have put up a small prototype around a very simple use case.
As I listened to those moving core information systems to the cloud, I was taken back by how many did not have a data integration strategy, indeed many have not yet considered it. The typically response was "Oh yeah, we need that don't we."
While the hype around cloud computing is high, the risk is that many systems will migrate to cloud-delivered platforms without a lot of forethought as to how they will synchronize key data back into the enterprise. Thus, in many cases it's going to be an afterthought, and you know how afterthoughts work...not very well.
The good news is that data integration for cloud computing is a pretty well understood science, it's just a matter of taking a few easy steps before you relocate to the cloud, including:
• Create a semantic level understanding of both the cloud-based and on-premise systems. This means defining the metadata, including security and governance.
• Determine the acceptable latency, or how often the data should sync from system to system, and how transformation, routing, and translation will occur.
• Select a data integration vendor that's able to support cloud computing, there are a number of them out there with pre-built adapters and special sub-systems supporting cloud-to-enterprise integration.
I suspect that data integration will be a force in the world of cloud computing. Not only cloud-to-enterprise, but cloud-to-cloud. However, like many issues with integration it will be an afterthought, but it does not have to be. Do your planning now.
Leave a comment