We were in Stamford CT for the annual SWG Analyst event (see here for some good blogs/videocasts on the event), that Steve Mills hosts. I had the duties to (i) present a chat with the expert session on event processing (ii) schmooze with the analysts (iii) participate in the web 2.0 panels.
On (ii), we had a great version of speed dating -- we hung around tables and the analysts came by and peppered us with questions, and on (iii), David Boloker, Carol Jones, Arvind Krishna and I had a great dialog on the barriers and opportunities for web 2.0 in the enterprise, and it was very clear that this is a topic of great interest for the analysts attending the meeting.
Let me spend a few minutes on the chat with the expert sessions. The point we (Bob Madey, VP of Strategy in our WebSphere division, Lisa Amini, Sr. Mgr at IBM Research building System S, and I) were making were simple.
1. The market for event processing has several patterns, some around Business Activity Monitoring, some around IT monitoring, some around program trading etc.
2. These patterns have different architectural requirements -- along dimensions of event rate (say 200K/sec for program trading, and 1000/sec for RFID tags in a particular set of distribution centers) etc.; latency requirements (milliseconds to minutes), complexity of event correlation etc.
3. That IBM currently does a lot of event processing around its core capabilities -- be it WBM, our Netcool products, or DB2 Data Stream Engine.
4. That we are building to a common architecture
5. That we will continue to add these capabilities around our current customer pain points and leverage partners as appropriate
6. That research is working on the next sets of activities (more massive volume, unstructured information, semantics of events etc.)
We had some very active discussions, and look for more from me on how our architecture and market penetration progresses in the next few months.
Comments