|
|
09-19-2008
Registered User
26,240,
27
EPTS4: Underhyped and Over here
vincent
09-19-2008 05:00 AM
EPTS meeting number 4 started this week with a keynote by Ramin Marzabani on “The Event Processing market from a VC Perspective”. Ramin came across as someone who was “unnaturally happy” for someone who is investing in a startup to compete with not only the likes of established successful CEP players like TIBCO, but also both IBM and Oracle who have just bought their way into the CEP party for massive amounts of $! The VC’s approach, he explained, would be to count on the “big guys missing a gap”. He also added that he expected most CEP startups to “fail”. Brave man, for sure!
Next came a panel on whether Event Processing is Hype, moderated by Alan of TIBCO:
Alan held a vote afterwards to see how many in the (albeit self-selecting) audience were of the opinion that CEP was in any way hype.
Answer: none.
That might have been expected, but far more disconcerting was a tendancy by a few people to pronounce CEP as “sep” rather than C-E-P. Sep? What? Where is this going? “We need to sep the baggage arrival events with the already-departed flights for the connecting passengers”?* “I sepped the patient RFID locators with the medication-required events”? “I sepped my market trade data with some old news from google to try and drive a market panic-sell”? Yuk!
Notes:
[1] Visions of DBAs and SQL programmers suffering the shock of their Oracle rep telling them the world is switching to event-driven processing…
Source...
09-19-2008 05:00 AM
EPTS meeting number 4 started this week with a keynote by Ramin Marzabani on “The Event Processing market from a VC Perspective”. Ramin came across as someone who was “unnaturally happy” for someone who is investing in a startup to compete with not only the likes of established successful CEP players like TIBCO, but also both IBM and Oracle who have just bought their way into the CEP party for massive amounts of $! The VC’s approach, he explained, would be to count on the “big guys missing a gap”. He also added that he expected most CEP startups to “fail”. Brave man, for sure!
Next came a panel on whether Event Processing is Hype, moderated by Alan of TIBCO:
- Brenda Michelson, normally seen running the SOA Consortium (or live-blogging), set a high bar for the rest of the meeting with a great analysis which could be summed as “compared to SOA and Web 2.0, EP is totally UNDERhyped!”. As evidence, she showed Google hit plots for “event processing” vs SOA and web 2.0. Brenda’s analysis also included a study of what industries were downloading the EDA paper on her site, which spookily correlates what some of us vendors are seeing in terms of CEP take-up:
- in 2006 this was the investment banks
- in 2007 this was the telco / retail / logistics verticals
- in 2008 this is the transport / retail banking / insurance / government verticals.
- Vendor Streambase’s John Partridge commented that the existance of (presumably happy) customers disproves the hype theory, and wryly commented that the Big Boys wouldn’t be investing in new companies if their existing technologies could handle event processing.
- Oracle’s Stephanie McReynolds, representing one of these Big Boys, countered with the fact that “hype was her job” and was useful as it “added disruption to challenge the status quo” [*1]. Stephanie was doing “2-3 customer visits a week” to explain CEP (hopefully not all to the *same* database customer). She thought the market awareness was there but there was a need to more early adopter case studies - a bit of a dig at their database sales teams, no doubt!
- Larry Fulton of Forrester braved a Gartner-sponsored meeting to share some more interesting statistics. Larry covers Enterprise Architectures, which is an interesting dimension for event processing. Forrester’s data shows future CEP adoption plans at 33% for CEP (vs 16% for ESP and 53% for BPM). Looking at just the next 12mths (for adoption or POCs) the numbers reduce to 5% (CEP, 7% (ESP) and 10% (BPM). Forrester is developing a taxonomy for Event Processing, coveringThey seem to have mixed up techniques (rule-based) with application areas (BAM) with event source (IT, business) with CEP (???) in their classification scheme: for example, products like TIBCO BusinessEvents process IT events representing business events doing CEP using rule-based processing for both BAM and other types of applications. So back to drawing board for that classification scheme. On a good note, Larry made the fascinating insight that internal IT departments are often an obstacle to EP implementations - a comment also made later by Robert Almgren in his keynote.
- IT events,
- Business events,
- CEP,
- BAM,
- rules-based processing and
- general-purpose.
- Ramin also joined in this panel. He added that it was all about solutions.
Answer: none.
That might have been expected, but far more disconcerting was a tendancy by a few people to pronounce CEP as “sep” rather than C-E-P. Sep? What? Where is this going? “We need to sep the baggage arrival events with the already-departed flights for the connecting passengers”?* “I sepped the patient RFID locators with the medication-required events”? “I sepped my market trade data with some old news from google to try and drive a market panic-sell”? Yuk!
Notes:
[1] Visions of DBAs and SQL programmers suffering the shock of their Oracle rep telling them the world is switching to event-driven processing…
Source...