A lot of posts to this group, and recent blogs by Joe
McKendrick among others, have brought up the debate
again about the Enterprise Service Bus.

For the most recent, see:
http://blogs.zdnet.com/service-oriented/index.php?p=560

Among the questions debated here is the lifetime of
the ESB product category.  Some suggest that it's a
temporary product category, soon to be subsumed by
something else.

I'm not so sure.  It takes a long time for a new
product category to get established.  Just ask our
friends at Sonic ;-).

And with IBM, BEA, Oracle, Tibco, and others recently
announcing they would ship an ESB the product category
has definitely been validated and, I believe,
established.  

But what is an ESB?  This question does indeed
continue to trouble the industry, since it still seems
as if every vendor has a different definition.

Several months ago I was invited to help deliver a 3
-hour tutorial on SOA and ESBs together with David
Chappell of Sonic.  He ended up injuring himself in a
water skiing accident (which he blogged about) shortly
before the tutorial date, so while the two of us
collaborated on the development of the presentation a
colleague of Dave's ended up physically joining me in
Washington for the event.

The way we split things up was:

-- I did about 45 minutes on generic SOA
-- Dave did about 45 minutes on generic ESB
-- Each of us did about 30 minutes on our
technologies, Artix and Sonic ESB, respectively

(With breaks and Q&A that filled the time.)

The interesting thing to me was how well the two of us
could agree on the generic SOA and ESB definitions.  I
would say it was like 90-95%.  Where we diverged was
in our respective approaches to addressing the
requirements of SOA infrastructure, aka our ESB
products.

I think up to this point Dave would agree with
everything I've said.  Now I will give my view.

The difference is that Sonic started with its
successful JMS product and added Web services support
and other elements of integration software in order to
create an ESB that meets SOA infrastructure
requirements.

We started with our patented configurable microkernel,
ART, and added a Web services "personality" aka
collection of plugins.

The result on the one hand is JMS centric while on the
other ours is a multi-communications protocol, multi
data format, brokerless, hubless distributed
architecture much better suited for SOA
infrastructure.

Unfortunately it seems like most vendors are adopting
the JMS centric approach, and that is what leads to
the controversy. 

I also wrote a bit about this in my blog:

http://www.iona.com/blogs/newcomer/archives/000269.html

I will try to post to this list now and then about why
the distributed approach to an ESB makes more sense.

Eric

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 




 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/service-orientated-architecture/

<*> To unsubscribe from this group, send an email to:
    [EMAIL PROTECTED]

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/
 



Reply via email to