Deep in the event processing jungle

Written By:
Content Copyright © 2009 Bloor. All Rights Reserved.

Lots of strange undiscovered animals lurk in the depths of the jungle. Most are prey. A few are predators. And two of the latter have recently been roaring within the jungle that is the event processing market. I refer to IBM and Sybase.

IBM has just made the first release of System S available after unveiling it last September, while Sybase has extended its Real-time Analytics Platform (RAP) which it released in a similar timeframe last year.

In my opinion, IBM will dominate the market for very high end event processing requirements. It will also dominate the market for processing streamed, unstructured data. On the other hand, the way it looks like right now, Sybase will dominate the market for environments where you need to combine high volume, low latency event processing with complex queries run against historic data. For a more detailed discussion on why I think this see articles on this subject.

The question is: where does this leave all the other event processing vendors? The short answer is that vendors like Progress, StreamBase, Aleri and others can only concentrate on structured, event processing with no back-end analytics. But how many customers won’t want to do analytics against their data at some time in the future? I can think of some sensor-based environments where that may be the case, and you can use products like this to support ETL processes (for example) but the problem is that there are other suppliers in those spaces. There are also a variety of use cases that these vendors have found but no real markets outside of financial services in general, and capital markets in particular.

I don’t think any of the established vendors (with the exception of Event Zero, which has its own heterogeneous event collection capabilities) will be able to compete with IBM at the high end. Their only leverage is price. That may be effective further down the market where structured data is concerned. Against Sybase they really have no leverage: they must partner or merge with a data warehouse supplier if they want to compete in the long term or else they should be looking for the niche markets they are going to run to.

Of course, an obvious move would be for StreamBase to cosy up even closer to Vertica (they were both founded by Mike Stonebraker). Perhaps less well known as a possibility is that Aleri used to be a data warehousing vendor with its own (quite cool) vector database, which perhaps it could revive? Of course, there’s Oracle but they can’t compete with Sybase for back-end analytics and they don’t seem to be doing anything very exciting at the CEP end either; still I don’t expect them to disappear into a niche (nice thought though!). My guess is that we’ll see a number of partnerships (some real, some just marketing) announced over the coming months between the CEP vendors and various data warehousing vendors, especially those, like Netezza, that have strong partnership ecosystems. The quality of those partnerships and the care with which they are lavished may determine who can survive and compete and who will die or just fade away.