If your organization has moved past all the excuses to use the public cloud, congratulations! But did you think through your data-integration strategy before deployment? If not, you'll find it difficult to maintain your corporate data in a public cloud.
The best bang for the cloud computing buck comes from using public cloud resources, such as Amazon Web Services, Google, Rackspace, IBM, and Microsoft. But that means you have to move some of your corporate data to the public cloud to take advantage of its cheap storage and per-use rental of compute cycles.
[ Get the no-nonsense explanations and advice you need to take real advantage of cloud computing in InfoWorld editors' 21-page Cloud Computing Deep Dive PDF special report. | Stay up on the cloud with InfoWorld's Cloud Computing Report newsletter. ]
When you move even a smidgen of data to a public cloud, you quickly understand the need for some sort of synchronization with on-premise enterprise systems. Otherwise, users will rekey data, overnight USB drives, and take other ugly approaches to data movement -- it happens more often than most IT organizations realize.
But when IT acknowledges the data-sync problem, cloud deployments grind to a halt until IT figures out the synchronization issue.
It doesn't have to be this way, nor should it. After all, the synchronization issue can be solved easily with a bit of upfront planning. Data integration technology is in its fifth or sixth generation these days, so the technology typically has a quick ROI and, ironically, is often available on-demand like the cloud itself.
Like many architectural issues that need some face time when you move to the public cloud (such as security, governance, and performance), data integration requires upfront thinking and planning. Here's what you need to consider:
- The amount of data to be placed in the public cloud
- The kind of data moving between the enterprise and the cloud provider and the frequency of that transmission
- The content and structure changes required for cloud storage, including encryption and other security requirements
- The logging and auditing requirements
- The exception-handling needs
Unfortunately, I suspect the data-integration gotchas will not go away any time soon. Many enterprises don't like to think about data integration for the same reason we don't like to think about indoor plumbing. But if you can hold your nose and do the upfront work, your public cloud deployment will work well.
This article, "Deploying to a public cloud? Deal with data integration first," originally appeared at InfoWorld.com. Read more of David Linthicum's Cloud Computing blog and track the latest developments in cloud computing at InfoWorld.com. For the latest business technology news, follow InfoWorld.com on Twitter.