A penchant for names

BEA, which is named after its founders Bill, Ed and Alfred, has extended its use of names with the introduction of TED (officially WebLogic Event Server), which is its “time and event driven” platform or, to put it more prosaically, its event processing platform.

There has been some speculation that TED is based around the open source Esper event processing engine in part, at least, because EPL (event processing language) is the name for the development language in both products. In fact, BEA has licensed Esper but has made significant changes to it.

From a technical perspective, TED is written entirely in Java (as one would expect from BEA) and integrates with, though it does not require, both BEA Real-Time and BEA AquaLogic. The latter is particularly useful because it provides a direct link with business process management capabilities so that you can build event processing (queries) into business processes and process-driven business intelligence applications. Throughput is around 50,000 events per second per server with sub-millisecond latency, which is lower than the likes of Coral8, Progress Apama or Streambase but, as I have noted before, the cost of any additional servers you might need is relatively small compared to the licence fees for the software, though it is obviously a less green solution if you need multiple servers rather than one. Of course, if you do not have such stringent latency requirements than this will be less of an issue.

Once events are processed these may be stored for playback but the events are not stored in a database but rather in a file system. The company’s Data Services Platform (DSP), previously known as Liquid Data, can be used to access external data sources such as data warehouses, in case you want to combine historic data with events for, say, pattern recognition.

In so far as development is concerned, TED lacks the sort of GUI development environment provided by the likes of Streambase, Progress et al. Indeed, in this first release you are more or less coding into a clever editor though BEA will be providing an Eclipse plug-in with its first release, even though this will not be as advanced as some of its competitors. In addition, BEA does not expect to add the sort of application functionality that Progress, for example, has recently announced for algorithmic trading. Nor will it be providing pre-built dashboards and the like, though it may offer dashboard components.

Most of the foregoing might suggest that TED is going to have a hard time competing with existing players within the event processing market. However, BEA believes that there is a market at the high-end where large organisations want the freedom to do their own development, don’t want packaged application capability, but do want the assurance of working with a major vendor and like the idea of integrating with existing products such as AquaLogic and BEA Real-Time. Certainly, indications from its beta sites (who reportedly like the flexibility offered) suggest that BEA is right in thinking that there is a market for such requirements that the company can develop and TED is likely to prove popular amongst existing BEA customers, especially if they already use these products.

On the other hand, I don’t expect the introduction of TED to have much impact on the existing players in the event processing space. The various companies mentioned, and others, are targeting smaller companies as well as large ones, are offering packages and are starting to leverage partners, all of which means that any successes that BEA may have will have little impact on their success. In other words, in the short term at least, I think that BEA’s entry into the market will be additive: expanding the market in general rather than taking much away from its competition.