Why Did SpringSource Buy GemStone: A Discussion with Rod Johnson
By Cameron McKenzie
Last week I got an opportunity to speak with Rod Johnson, the former CEO of SpringSource, and the current General Manager of the SpringSource product division of VMWare, while he was out doing the press tour, announcing VMWare’s recent acquisition of GemStone. For those who missed the news, GemStone is a vendor in the data management space, specializing in sophisticated tools that do, among other things, distributed caching.
So, VMWare bought GemStone; the obvious first question is “why?”
“We think this is an important enhancement to our middleware portfolio” was Rod Johnson’s response.
“First of all, this addresses a problem that is relevant to many large users of Enterprise Java. If you look at the GemStone customer base, there are hundreds of customers, and those customers are in some of the most demanding industries, such as financial services and defense; and the GemStone product is very much a proven product in terms of its ability to solve hard, distributed caching, problems. But really, that is only part of the story. The main theme here relates to our move towards cloud computing.
“We are putting a lot of effort into providing a path for enterprise Java users to embrace cloud computing. This includes both a public platform as a service initiative, such as VMForce, and it also includes building a private cloud product with SpringSource middleware.
“We believe that in making the transition to cloud computing, sophisticated data management is essential. Firstly, it allows applications to scale horizontally and elastically, in the middle tier, often without needing to change or scale the database, (which is far more difficult). So, GemStone solves an important challenge of cloud computing.
“Secondly, it can significantly improve performance by reducing latency in many applications, which can make a dramatic difference in performance.
“But it can also help to un-tether application logic from data at rest. So, very often in a public or hybrid cloud, there may be some parts of an application’s data, for security, regulatory or other reasons, that needs to be stored somewhere different from where the application’s compute resource is. Clearly, this is a significant challenge, because it is very difficult to move data at rest; it’s much easier to move compute capacity. And a sophisticated data management product like GemStone can actually help to solve this problem by significantly reducing latency when accessing data at rest.”
So, that answers the questions “Why?” And of course, figuring out the why wasn’t too difficult.
Everyone that’s been paying attention knows that SpringSource and VMWare have their eyes on the cloud. And to compete effectively in the cloud, you need a good messaging infrastructure, which SpringSource obtained when they acquired RabbitMQ a few weeks ago, and of course, you need a distributed caching solution, which SpringSource now has with GemStone. But it also begs the question: why GemStone?
To be honest, I had anticipated the purchase of either Terracotta or GridGain. GemStone wasn’t at the top of my acquisition list, although they were probably one of the less expensive vendors that were waving a for sale sign at the industry. But when asked why SpringSource picked up GemStone, as opposed to Terracotta or GridGain, Rod Johnson made it clear that GemStone had the technology that they were after.
“I would say that a GridGain comparison isn’t apples for apples.” Was Rod’s response, not really seeing a spot for GridGain in their current portfolio of acquisitions. “I see GridGain’s strength being in a slightly different area around computer grids.
"Terracotta, you could argue, is a somewhat more direct comparison.
“Basically, we felt we knew this space very well, we’ve obviously been in this space for many years with Spring, and we’re familiar with the solutions here.
“We really wanted to acquire a technology in this space; it’s the next thing we would have done as an independent company at SpringSource. One of the great things that happens at VMWare, is we really have the ability to look across the options and not be constrained with regards to which options we could choose. That wouldn’t be possible if we were a smaller, independent company.” In other words, VMWare has deep pockets, and the decision to acquire GemStone was less about ‘what can we afford’ and more about ‘what exactly is it that we want.’
“Frankly, we were very, very impressed by the quality of GemStrone technology.
“This is a space where getting the solution right takes literally years. The algorithms are hard; then there’s years of toughing it out in the trenches in order to give customers a really rock solid implementation.
“We were very impressed by the due diligence of the GemStone technology. We were very impressed with the extensive conversations we had with existing customers of GemStone, such as JP Morgan.
“This was the technology we felt was at the most appropriate level of capabilities and also, at the appropriate level of quality.”
And that really is the key. In order for SpringSource to be competitive in the cloud, they need a proven, distributed caching technology on which they can rely. And they’re betting on GemStone’s existing technology to fill the void they previously had in that space.
12 May 2010