The whole idea of innovation is to build a better mousetrap, but Sybase’s latest product shows you can also focus on simply building a better piece of the mousetrap.
Sybase is a database management company so boring it’s almost never included in the same breath as Oracle, IBM’s DB2 or SQL Server. What it lacks in hype, however, it makes up for in sheer survival. The company maintains a strong development presence in Canada, where its focus has primarily been on the iAnywhere line of mobile enterprise products. This week’s launch of its Adaptive Server Enterprise Cluster Edition is as close as it’s come to shocking the industry by a) drawing attention to itself at all and b) going up more directly against a key Oracle technology.
Adaptive Server Enterprise Cluster Edition is Sybase’s take on Oracle’s Real Application Clusters (RACs). When they came out earlier this decade, RACs were Oracle’s break-the-business bet that companies would load-balance their database loads across a couple of machines. It’s taken a while, but it works. I’ve spoken in the past with customers like Best Buy who have either implemented or are considering it as a way to manage the growth of their information management needs.
In what sounds a lot like storage virtualization, Adaptive Server Enterprise Cluster Edition abstracts the physical machines with all their various database sessions and allow them to put in place different service-level requirements. All of this would be managed in one interface, which would mean DBAs could allocate resources more efficiently.
There’s just one thing: Oracle moved on from its RAC strategy at least four years ago when it announced its 10g database. As Larry Ellison explained at the time, RACs are great for working with two or three machines, but the grid-enabled version of its flagship product could create a manageable architecture for a fleet of database systems. Oracle still offers RAC technologies, of course, but that’s not where the big money is, and it’s not where its team is focusing.
Sybase has come out with a RAC competitor for late adopters, the kind who aren’t prepared to run grids within the next several years. That slow, steady approach to winning the database race is consistent with Sybase’s history, and may make it more successful than some analysts expect. It also raises an issue that my colleagues at Computerworld U.S. already explored – namely, that Sun Microsystems might have made a much more credible choice by acquiring Sybase instead of MySQL. When you consider the fact that Adaptive Server Enterprise Cluster Edition will likely work best in Solaris environments, you would think Sybase might have been the more complementary platform, but one that might have rankled longtime partner Oracle more.
That still leaves a number of big players who lack a database, such as HP or SAP, or even those that don’t (IBM didn’t blink at buying Informix in 2000, despite its investments in DB2). The only thing that would surprise me more than Sybase continuing to stand on its own five years from now would be if survived not by starting new database trends, but by doing a better follow-up job on existing ones than anyone else.