Editor’s note: This post by Chip Childers of CumuLogic captures context we believe of special relevance to all enterprise IT professionals, especially those with data to collect/store/process/analyze. He ends with a topic more of us from the technology world should be weighing in on. – bg
As public and private cloud adoption continues to skyrocket, it’s becoming more and more clear that Database-as-a-Service (DBaaS) is hugely important to users. This isn’t really news, as much as it’s a confirmation of what many were seeing much earlier in the cloud industry’s maturation. DBaaS is the next killer app for cloud services.
As an example, consider that DynamoDB from AWS was, after only five months in production, the fastest growing AWS service all the way back in 2012. Also consider the 451 Research report authored by Matt Aslett back in August of 2013. By Matt’s estimation, DBaaS services based on MySQL will grow at a CAGR of 81% from 2012 to 2016. That’s some serious growth, representing a clear user demand.
Why do users gravitate to these services? Well each service has it’s own specific draw when compared to others: perhaps it’s just the service offered by your favorite cloud provider, or perhaps it’s a nifty new technology that your developers fell in love with. Generally though, I believe that the reason for the draw is that databases, while critical to your applications, are generally a bit difficult to configure, scale and operate correctly. When presented with the opportunity to off-load that work to a provider, developers jump at the chance.
Application owners need to be careful to select the right DBaaS provider. If you select a single purpose public DBaaS cloud offering, how do you ensure that your applications are close enough to their data to avoid unnecessary latency (or even sometimes costly bandwidth consumption)? If you select a DBaaS service being offered as part of a suite of services from a provider, the application locality becomes less of a problem, yet you can still fall into the trap of being locked into a single provider if there’s no compatible alternative available. How can you move from that provider if they go out of business or become more expensive than alternative providers? Last, what about data sovereignty concerns? Is your data being stored in a provider that you can trust, and whose government can be trusted?
Some of this might seem like a bit of fear, uncertainty and doubt (FUD), but these are quite practical concerns. As I’ve heard from enterprises and services providers around the globe, these exact questions are issues that they either struggle with directly, or hear from their customers, on a regular basis.
The time is right for a solution that can solve many of these issues. DBaaS is the next killer app for cloud services, but it needs to be done in a way that can meet the challenges of locality, operability and provider choice.