RE: [soa-blueprints] Anti-Blueprints - Number of services

["Duane Nickull" <dnickull@adobe.com>]

I think a stronger metric for this would be to compare the number of
services to the number of functions you wish to expose.  A good practice
would be to make them as granular as needed for maximum re-purposing,
but no finer.  Since the quantitative measure of this varies with each
implementation, I do not think we or anyone else could prescribe a
single rule to govern this, just a cautionary note.

The other aspect is redundant services or orphaned services.  Having a
large number of services not being used probably indicates you need to
revisit some aspect of your IT strategy.  

I concur with Marc that this is a best practices conversation.

Duane

-----Original Message-----
From: Miko Matsumura [mailto:mmatsumura@infravio.com] 
Sent: Tuesday, November 01, 2005 10:49 AM
To: Davies Marc
Cc: soa-blueprints@lists.oasis-open.org
Subject: RE: [soa-blueprints] Anti-Blueprints - Number of services

 
+1 on this bit--

The pattern/antipattern bit is not intended to be universal or
descriptive, it's intended to be prescriptive and cautionary. In any
event, if it's a handy tool at any given time we will pull it from our
toolbox and use it...

--------------------

In a nutshell - do I think the number of services plays a part in the
pattern/anti-pattern debate? Yes I do, because this is not a debate
about service orientation in its purest sense - it's a debate about what
is 'good practice' versus 'bad practice' when assessing the fates of our
customers
(internal/external) and attempting to deliver an architecture that
reduces cost/complexity, not increases it. Put another way, if the RM TC
is where the Pure Mathematicians reside, then is not Blueprint TC where
the Applied Math is carried out?

Best,
Marc.
<ducks after typing.., :o) >


-----Original Message-----
From: Duane Nickull [mailto:dnickull@adobe.com]
Sent: 01 November 2005 15:25
To: Davies Marc; Miko Matsumura; Jones, Steve G; Beack, Theo; Ken Laskey
Cc: soa-blueprints@lists.oasis-open.org
Subject: RE: [soa-blueprints] Anti-Blueprints - Number of services


<post>
-----Original Message-----
From: Davies Marc [mailto:Marc.Davies@uk.fujitsu.com] 

I think Duane's example of the Internet perfectly underscores this
principle
- inasmuch as the Internet is a collection of millions of services - it
is
(IMHO) *not* an SOA - unless we're talking Service Oriented Anarchy :-)

... it doesn't conform to architectural disciplines, anyone can code how
they want, anyone can deploy what they want, there are no checks to
ensure
services deliver on their promised capability (I could go on). Sure, its
an
excellent example of how millions of services can be operating - but,
its
also an illustration of how millions can end up delivering very poor
service, to analogise - if you google the 'wrong' search string - you
end up
with 1 million 'hits' = meaningless.
</post>
Marc:

Hmm - I strongly disagree with this assertion.  The internet has the
same patterns as web services.  Request-response is the primary
mechanism, returning either a success state or a possible error code.
It is message oriented and event driven.  Each service may have specific
policies, metadata (<meta> tags along with the search engines synopsis),
a contract for use (in most cases it is freely available to everyone who
asks) and there are multiple mechanisms for advertising the availability
of services.

One of the core tenets of interface based design is that anyone can
implement whatever they want behind the service interface.  It is not
limited to just coding either - you could deploy chimpanzees with
abacuses who then serialize a response back into html.  The point is
that the interface hides the implementation and insulates the consumer
from those details.  This is another of the core tenets of SOA.

Your claim that the internet delivers poor services is also
unquantifiable.  From a pragmatic architectural standpoint, the value of
the content is moot.  It is the architecture that is SOA, not the
usefulness of the content.

Google roughly equates to the concept of the advertising/discovery
mechanism in the SOA Reference Model.  Yes - getting hundreds or
thousands of results is sub-optimal, however the patterns are the same.

The internet is the single largest SOA on the planet.

Duane