Shifting into the Next Gear

March 17, 2005

Shifting into the Next Gear

I'm struck by how dramatically things have changed since last summer when I began writing this column. Early in 2004, I was spending a lot of my time with prospects explaining the need for SOA infrastructure and drawing what I hoped would be a cogent mental picture of an enterprise service bus (ESB). I'd talk through the various constituent components and how they come together into a unified architecture for connecting, mediating and controlling business services and their interactions. These conversations generally went quite well because the need for such an infrastructure is very real - the technology solves a problem that needs to be solved. The tricky part, at the time, was that software pundits at large weren't talking about SOA infrastructure, and if they were they would generally comment that there was little convergence around how such infrastructure should be built or exactly what it should do.

Today, SOA is treated as a given, the need for SOA infrastructure is much more broadly accepted, and the shape and capabilities of an ESB are discussed in more concrete terms. If you've been reading this series since last year you may recall that I spent one installment discussing - some might say griping about - the bewildering array of viewpoints provided by industry analysts on the topic. Interestingly, one of the most clear analyst reports on ESB to date arrived shortly after I penned that piece. And from none other than one of the analyst firms I'd called out in my Insight. I could say "ouch," but of course I was relieved to read Mike Gilpin of Forrester's essay on all things ESB. In it, he observes:

"An enterprise service bus (ESB) is software infrastructure that enables service-oriented architecture (SOA)
by acting as an intermediary layer of middleware through which a set of reusable business services are made
widely available. An ESB helps enterprises obtain the value of SOA by increasing connectivity, adding flexibility
that speeds change, and providing greater control over use of the important resources it binds."¹

He goes on to describe ESB capabilities at length, and in terms that are now becoming much more commonplace. Similar observations in varying depths are being offered by other analysts and industry-watchers.

Many organizations didn't wait for such guidance in 2004. They immersed themselves in SOA technologies and principles. And they discussed SOA with a variety of vendors to see how they thought about it. They attended webinars and seminars. Many read Dave Chappell's ESB book. They built prototypes and proof-of-concept SOA projects. And most importantly they gained a true understanding of how SOA could fit into their project plans and how its benefits would map to the requirements of their organizations. For more and more IT organizations the initial tire-kicking is over - they've moved on to develop long-term SOA strategies. I've had a remarkable number of IT leaders tell me that for them SOA with an ESB is the underpinning of a 10- to 15-year architecture plan. Not that they think that it will take them 10 years to "get to SOA", but that they are building a foundation now that they believe will support them for the next decade. And not for lightweight applications either, but for such mission-critical and risk-averse environments as large-scale manufacturing businesses and airport operations. Clearly the market now believes what Sonic has believed for five years - ESB is real.

So, I'm still having those conversations with technologists and IT leaders (as do my fellow colleagues at Sonic Software). In fact, we have way more of them than we did even six months ago. We still discuss SOA, the various constituent components of an ESB and how they come together in a unified architecture. We talk about the ESB's role and how it can bring both flexibility and control to IT environments. But conversations now start from a very different place. Technologists want to better understand how we think about SOA and where we are taking our products. They want to go to the next level of detail on how the ESB can fit into their projects. They want to "test for echo" in order to verify that the perceived benefits can be realized in their business and IT environments. In other words, because we're now often starting on the same page and with aligned visions, the conversations are much more potent. Yes, we've truly shifted into the next gear. Now the real fun starts.

Gordon Van Huizen
Chief Technology Officer
Sonic Software

¹What Is An Enterprise Service Bus?, Forrester Research, August 2004