I Have Seen the Future of Database, and It Is …

by Jeff Cogswell

The biggest thing we’ll see by far in the next year or two is an increasing amount of database storage moving “into the cloud.” We’re already seeing it, but we’ll see more of this moving forward. Databases will be distributed worldwide, and the hardware will support it. For example, Microsoft acquired DATAllegro, which builds data warehouse appliances. These babies will be fast and support the needs of cloud databases.

Because of the move into the cloud, smaller companies will be able to afford to rent the space and bandwidth on these massive cloud-based storage systems. One- and two-person shops will be able to have massive applications that rival those of Amazon.

Although the databases will be off-site, you’ll have immediate access to them right from your desktop computer. It will be as if they are sitting right in front of you. Microsoft has announced plans to add even more “self-service” tools with its SQL Data Services.

With bigger and more powerful databases comes the need for more business intelligence and reporting features. Oracle and Microsoft both have plans to add more features that will allow executives to continue using their favorite tool, Excel, to quickly drill down into data.

Facebook may be redefining the way data is stored. With only a couple of administrators, Facebook is managing thousands of MySQL installations. But it also has created new ways of looking up and instantly accessing large BLOB data. Look at how quickly images come down for the millions of people using Facebook at any given time. This isn’t just dumb luck. Facebook is pioneering the future.

Although you’re going to see more REST access, allowing Web applications to quickly access and manipulate data through simple HTTP calls, Microsoft and others are also creating new protocols to allow servers better access to Web-based data. Microsoft’s new Tabular Data Stream protocol even supports triggers and stored procedures.

Google is changing the way data is stored with its BigTable approach, which is anything but relational. This thing is fast. Developers are recognizing that there’s a need for speed in lookups, but not quite as much in writing, which is the idea behind BigTable. Amazon has created a similar type of database that is flat. But developers still want relational access. While Microsoft’s Azure supports a flat data approach, your software can still access SQL Services in the cloud. We predict that we’re going to see new technologies in the next two years that will allow for incredibly fast relational databases.