Integration trends and development

We had yet another trend spotting get-together at the Dynabyte office, this time the subject at hand was systems integration. The group was comprised of everything from single platform developers with ~10 years experience, multi platform developers, to pure Java developers with nothing but theoretical knowledge regarding system integration.

The first major subject to be discussed was the incursion of open source alternatives on the integration market and the effects of this. Naturally this subject divides any group in to different camps and any common point of view is all but rendered impossible. The one subject where the group managed to agree upon is that open source does have a clear role to play for smaller companies who has a very limited need for a standardized integration solution.

As the discussion moved on the next big topic to be touched was SOA. Yes, the ever so hyped subject with some true potential at the core. The first discussion was whether integrations have any role to play in a true SOA architecture. My personal opinion is that they play the essential role of keeping the end systems simple and yet allow you to make complex services available, if you want to call this service orchestrations or composite applications I guess that is just a question of what platform lingo you adopt.

As the discussion progressed we drifted more to what incentives we have to justify the investment in a SOA architecture. Monitoring and proactive error handling, governance and reactive service level rules were all discussed. These were all countered with the problems like how do you determine a value on an event that does not occur, how do you estimate the cost of altering the corporate structure to match the need and utilize the potential of a different architecture and how do you solve the friction created when “the way things work” needs to change.

Also there was a discussion on whether or not companies would dare to adopt the cloud to their integration backbone or if that just is something that is best kept in house.

You could have hoped we could have come to a clearer vision of what is to come in the world of integration. But we got a most enjoyable discussions and the conclusion is that the big issues that has been affecting the general software market is now more extensively being applied on the integration segment. And that most of the major players are all more or less focusing on expanding from the central concept of directing information streams to the areas of orchestrating, monitoring and design of the information flow whilst open source alternatives in making headway in to the core segment.