[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [soa-rm-ra] Please re-read Sect 4.3.4 of soa-ra-cd-02
|
Here is my attempt to slightly
re write the section. I send it out several month ago, but based on yesterday
it still seems relevant. I can do some more work, if we agree that it is a
proper direction From: Ken Laskey
[mailto:klaskey@mitre.org] Jeff, A useful analogy that comes to
mind is when we built an addition on our house. We hired an architect and
a general contractor, and the general contractor was responsible for hiring
plumbers, carpenters, carpet installers, etc. The general contractor had
the role of the composite service. I sometimes saw the plumber when
plumbing work was being done (often didn’t) but didn’t usually know
the details of what the plumber was doing or when. (The general
contractor did mention at some point that the plumber was two days late and
there would be some workarounds.) Same for the other
subcontractors. I knew there were dependencies but I knew nothing of the
details and assumed these would be handled adequately by the general contractor
as part of his service. If there were independent policies on the part of
the subcontractors, I was unaware of these and only dealt with the policies of
the general contractor. For example, there was a payment schedule and I
assume the general contractor took into account the need to pay the subcontractors,
but I only dealt with the policies as specified by the general contractor (my
composite service). I look at the provider of a
composite service as the consumer of all its components, and the general
contractor dealing with each of his subcontractors is no different in general
from me dealing with him. Also, the general contractor coordinating among
the subcontractors was similar to me coordinating with the general contractor
and the architect. One place where I may not
understand your perspective and have trouble relating to my home addition
example is lines 2338-2346 when you talk about a Service Management
Service. You say this service needs visibility into the composition so it
can manage dependencies, but I see managing dependencies as the job of the
composite service. Now the composite service may make use of tools to
monitor the components (or the ecosystem, in general) and may make use of
operation research tools to balance complex tradeoffs among requirements.
However, these are tools (possibly services) used by the composite to manage
its requirements and dependencies. I do not think of there being some
independent management entity unless the full management function was
subcontracted by the composite. Quick other points: -
Description information
is never complete but various subsets of “complete” description
should contain consistent information -
A given service may
have more than one endpoint but I would expect a single information model and,
possibly, a single process model. Imagine the difficulty in writing a
clear, machine-processible description if this is not the case. -
No service operates
in a vacuum, and trying to convince people that anything infrastructure is not
a service will be a very hard sell. Enough for now. Ken --------------------------------------------------------------------------- Dr. Kenneth Laskey MITRE Corporation, M/S
H305
phone: 703-983-7934 7515 Colshire
Drive
fax: 703-983-1379 McLean VA 22102-7508 From: Estefan, Jeff A (3100)
[mailto:jeffrey.a.estefan@jpl.nasa.gov] RAF Colleagues, A very interesting discussion ensured
today on the topic of atomic vs composite services. I’d like to
encourage everybody who participated in the discussion to re-read Sect 4.3.4 in
its entirety prior to next week’s meeting and proposed (via e-mail in the
interim if you like) some wording changes to the formal definitions captured in
this section. The reference document is soa-ra-cd-02 lines 2311 to 2350. Remember that the original issue
resolved around multiplicity of interfaces and descriptions and we did not even
discuss that point today and for the atomic service case, the issue around the
language of interacting with other services (which is accepted to change to use
or rely on other services). Indeed the importance of this
distinction depends on one’s perspective but I ask that you imagine the
challenges from a management point of view for a highly composed service in
terms of management of policies and contracts of each of the services that make
up the composite service. I think this has some real practical
implications and they are not trivial. Thanks! - Jeff The information contained in this communication may be CONFIDENTIAL and is intended only for the use of the recipient(s) named above. If you are not the intended recipient, you are hereby notified that any dissemination, distribution, or copying of this communication, or any of its contents, is strictly prohibited. If you have received this communication in error, please notify the sender and delete/destroy the original message and any copy of it from your computer or paper files. |
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]