Here are some thoughts on how this should be structured from Giles
Nelson's blog
(http://apama.typepad.com/my_weblog/2008/10/cep-eda-and-soa.html):

<<I'd like to add my voice to the debate this week (here, here and
here) on how Complex Event Processing (CEP) fits into the wider
software architectural themes of Service Oriented Architectures (SOA)
and Event Driven Architectures (EDA). Although I think I know how
these three areas relate to one another fairly well, I was able to
further clarify my thinking this week by spending some time with Neil
Macehiter of Macehiter Ward-Dutton Advisors, a UK based software
analyst. I found our discussion enlightening as Neil had a slightly
different way of looking at these things than I had heard expressed
previously. So let me try and express my own view on this in as clear
a way as possible.

1. CEP is a technology. SOA and EDA are not technologies. SOA and EDA
are philosophies for the design and build of modern distributed
computing architectures.

2. A SOA is a loosely coupled set of services, the functionality of
which closely reflects an organisation's business functions and
processes. A SOA will typically use modern, Web services technology
and standards for implementation, but is not required to. Building a
SOA requires much thinking about the services that the SOA will use.

3. An EDA is a loosely coupled architecture, the endpoints of which
interact with one another in an event-driven fashion. Information
flows around the EDA as events. An EDA will have endpoints which
produce events and endpoints which consume events. An EDA works in a
"sense and respond" fashion. Building an EDA requires much thinking on
the event-types that the EDA will use.

4. An EDA may use business focussed services as endpoints. An EDA may
therefore also be a SOA but it does not have to be.

5. CEP is a capability within an EDA, providing analysis and matching
of multiple events being sent between endpoints. You can have an EDA
without CEP.

6. If you're building your architecture and focussing on defining
event-types, it's very likely you're building an EDA.

7. If you are using CEP then you have at least the beginnings of an
EDA because you will have been focussing on event-types. Your EDA may
a simple one, with one event producer and consumer, but it's still an
EDA.>>

Gervas

--- In [email protected], Michael Poulin
<[EMAIL PROTECTED]> wrote:
>
> I agree with Rob. However, coupling between sender and recipient in
EDA is shifted into more delicate area of intentions (sender) and
interests (receiver). How technically quantify these categories is a
question.
> - Michael
>
>
>
>
> ________________________________
> From: Rob Eamon <[EMAIL PROTECTED]>
> To: [email protected]
> Sent: Tuesday, December 2, 2008 8:42:39 PM
> Subject: [service-orientated-architecture] Re: Roch Clarifies Some
Key Terms
>
>
> Not a bad set of definitions. A couple of discussion items:
>
> EDA is not necessarily pub/sub.
>
> EDA does not result in "completely decoupled" systems. In a
> brokered/intermedia ted approach, they are more loosely coupled, but
are
> still coupled in two ways: 1) the messages they exchange; 2) the
> mechanism over which they exchange them. While the recipient doesn't
> care exactly where the message originated, it cares that *something*
> originates it and does so via the intermediary to which it is connected.
>
> -Rob
>


Reply via email to