Re: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

-----Original Message-----
From: Peter F Brown <peter@peterfbrown.com>
To: mpoulin@usa.com; klaskey@mitre.org; boris.lublinsky@navteq.com; jeffrey.a.estefan@jpl.nasa.gov
Cc: soa-rm-ra@lists.oasis-open.org
Sent: Fri, Jan 14, 2011 7:01 pm
Subject: RE: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Michael:

So can you please tell me what a service is? Can you formally model it for me (ideally in a way that shows that it is NOT a resource)? What are its parent classes?

I thought we had this settled already: We define service in the RM as

“a mechanism to enable access to one or more capabilities, where the access is provided using a prescribed interface and is exercised consistent with constraints and policies as specified by the service description.”

Mechanism is generally defined as the “the technical aspects of doing something” (my emphasis). I only said that “it is not an entity in the sense that ‘resource’ is an entity.”

Please show me precisely where I’m going wrong. And, if it is not a mechanism, we need to change the RM too….

Some further comments inline…

From: mpoulin@usa.com [mailto:mpoulin@usa.com]
Sent: Friday, 14 January, 2011 10:02
To: peter@peterfbrown.com; klaskey@mitre.org; boris.lublinsky@navteq.com; jeffrey.a.estefan@jpl.nasa.gov
Cc: soa-rm-ra@lists.oasis-open.org
Subject: Re: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Peter,

"we understand a service as ‘doing’ (mechanism) rather than ‘being’ (entity" - I CANNOT AGREE with this. [Peter:] but it’s what the RM states…

Service is an ENTITY , which is DOING. Service is an application [Peter:] No, an application is an application. If a service isn’t invoked, it remains a capability (and an application is-a capability)

or a business team [Peter:] ditto

that provides actions resulting in the execution of capabilities[Peter:] The service is precisely the provision of those actions – doing

. A half of SO Principles define the THING and another half defines the thing's BEHAVIOR.[Peter:] OK, so please point me to definitions of the THING.

I am afraid if we start going into direction that 'service is only doing', we have to re-write the RAF entirely.[Peter:] We can have another 5 years of academic debate or accept that ‘good enough is good enough’. I don’t intend to use the formulation exactly as I stated but try to give a general sense. I’m not saying service ‘is only’ doing…

As of 'communication activity' and 'service activity', this was my understanding of real world and our intent to reflect for the long time. I would also added a 'consumer activity' based on intent and willingness and resulted in 'communication activity'.[Peter:] Agreed, good point

The problem you face with service=doing is in that what service is, what service body is - isn't visible through the service interface.[Peter:] I agree, thus “service activity” and actions are ‘behind the curtain, opaque to consumer. What is a ‘service body’?

However, the consumer is interested in this interface at the last moment only. First of all, the consumer searches for the entity that can do certain thing. What the entity is, is it trustful, is is compliant, etc. are the first questions; only then the consumer concerns how interact with this service and trigger the 'doing' part.[Peter:] OK, but you are assuming that I mean service-as-doing as only including the ‘behind the scenes’ part. Doing SO involves all those steps, I agree

- Michael

-----Original Message-----
From: Peter F Brown <peter@peterfbrown.com>
To: 'Ken Laskey' <klaskey@mitre.org>; mpoulin@usa.com; boris.lublinsky@navteq.com; jeffrey.a.estefan@jpl.nasa.gov
Cc: soa-rm-ra@lists.oasis-open.org
Sent: Fri, Jan 14, 2011 4:16 pm
Subject: RE: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Ken:

When I stated that “service != resource” my concern was mainly with how we understand service in formal terms.

Given the modelling exercises to date, I did not want us to get into the trap of service *is-a* resource. Services use, create and manipulate resources. I would be OK with stating that a capability is-a resource (not that it adds a whole lot of understanding to an already complex situation) but I still maintain that a service is more the mechanism by which a capability is brought to bear to do something. It is not an ‘entity’ in the sense that we use in defining resource.

In Jeff and my discussion on action, following the TC call Wednesday, and taking on board the consensus of the TC to move “communicative action” > “communication” and “service action” > “action”, we realise that in a SOA ecosystem there are broadly two ‘types’ of ‘activity’:

– ‘communication’, which includes the process of semantic engagement and establishes an execution context; and

– ‘service activity’, which starts at the point of invocation by a participant and involves a ‘composition’ of one or more (service) actions, resulting in a shared state change that is delivered’ back to the participant and seen as a (series of) RWE. We sketched it out as an activity/behaviour diagram.

In this perspective, it seems clearer that everything that goes on behind the curtain of ‘service activity’ is what we understand a service as ‘doing’ (mechanism) rather than ‘being’ (entity). In other words, a service can be seen as a fulfilment (‘doing’) of the ‘promise’ offered by the capability (‘being’)…

Does that make sense?

Peter

From: Ken Laskey [mailto:klaskey@mitre.org]
Sent: Friday, 14 January, 2011 05:31
To: mpoulin@usa.com; peter@peterfbrown.com; boris.lublinsky@navteq.com; jeffrey.a.estefan@jpl.nasa.gov; soa-rm-ra@lists.oasis-open.org
Subject: RE: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Michael,

First, and without having read the text, I can certainly see our discussion of management focusing on services and not resources in general. What I didn’t understand was a blanket statement that service != resource. I can limit a discussion to services, but sometimes lumping services into the discussion of resources is useful.

As for single vs. multiple enterprises, I thought I sensed the assumption of a level of top-down control that typically is outside the realm of the organizations with which I work. Those organizations are committed to building a services enterprise and interacting, and I need to ensure that our work also supports them.

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: mpoulin@usa.com [mailto:mpoulin@usa.com]
Sent: Friday, January 14, 2011 8:16 AM
To: Laskey, Ken; peter@peterfbrown.com; boris.lublinsky@navteq.com; jeffrey.a.estefan@jpl.nasa.gov; soa-rm-ra@lists.oasis-open.org
Subject: Re: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

I have explained in my posts to Jeff, that, as Ken referred, resource includes service by the definition we use currently in the RAF. However, in several places (in management) it makes sense to talk about 'services' rather than about 'resources' and I made several changes with this regard in the draft sent to Jeff for the review.

- Michael

P.S. I'm still wondering what in my words caused Ken to believe that I limited my approach by the enterprise only. When I worked in Fidelity, we had the same multi-enterprise conglomerate and the SOA issues about cross-enterprise governance ans management are quite familiar to me. ( I do not work for Fidelity anymore, Ken, and my status in OASIS is an independent member)

-----Original Message-----
From: Ken Laskey <klaskey@mitre.org>
To: peter@peterfbrown.com; 'Lublinsky, Boris' <boris.lublinsky@navteq.com>; mpoulin@usa.com; jeffrey.a.estefan@jpl.nasa.gov; soa-rm-ra@lists.oasis-open.org
Sent: Fri, Jan 14, 2011 12:15 pm
Subject: RE: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

I’m not sure what the problem is in saying service = resource. Last I checked we said “A resource is any identifiable entity that has value to a stakeholder.” A service would certainly seem to fit there. Sorry I haven’t read the other material, but what is the harm?

One other thing for those working management. I sent the following to Michael while we were talking about testing and about composition. I think it is relevant to keep in mind for management too.

One other difference I think I see in our perspectives is that you tend to work within an “enterprise” and I tend to work in an enterprise that is explicitly a collection of enterprises. So I have 16 separate intelligence agencies, 5 separate military entities, and numerous other government organizations. They all expect to be service providers and service consumers. There is no single “top” to the hierarchy, i.e. no single CIO or similar figure. This ecosystem requires all the flexibility and consequent challenges we ever talk about with SOA. On top of that, lives (both participants and non-participants) are at risk in many outcomes. I think this also impacts our discussions of composition.

</perspective>

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: Peter F Brown [mailto:peter@peterfbrown.com]
Sent: Friday, January 14, 2011 1:57 AM
To: 'Lublinsky, Boris'; mpoulin@usa.com; jeffrey.a.estefan@jpl.nasa.gov; soa-rm-ra@lists.oasis-open.org
Subject: RE: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

+1 on service != resource. Jeff and I worked on the action and state parts of section 3 and relating them back to service, which we barely mention. I’m still, personally, in favour of thinking of service as a ‘mechanism’ but we have to find ways to express that more clearly still….

Peter

From: Lublinsky, Boris [mailto:boris.lublinsky@navteq.com]
Sent: Thursday, 13 January, 2011 18:39
To: mpoulin@usa.com; jeffrey.a.estefan@jpl.nasa.gov; soa-rm-ra@lists.oasis-open.org
Subject: RE: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

A lot of comments inside.

The big one services are not resources lets stop using these two terms intercheangeably.

From: mpoulin@usa.com [mailto:mpoulin@usa.com]
Sent: Thursday, January 13, 2011 2:16 PM
To: boris.lublinsky@navteq.com
Subject: Re: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Sure, here you are.

- Michael

-----Original Message-----
From: Lublinsky, Boris <boris.lublinsky@navteq.com>
To: mpoulin@usa.com <mpoulin@usa.com>
Sent: Thu, Jan 13, 2011 4:50 pm
Subject: RE: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Oops,

I deleted it in anticipation of the full thing. Can you resend? Please?

From: mpoulin@usa.com [mailto:mpoulin@usa.com]
Sent: Thursday, January 13, 2011 10:45 AM
To: boris.lublinsky@navteq.com
Subject: Re: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Oh! I cannot stand such argument. There are two things in my paper: several fragments from different parts of the doc (with related line numbers) - you can ignore all of them & the FULL section on Management Concept - as it was, commented, modified and extended. In this case, you can take it The Gospel from RAF & Michael; all context is in there.

Cheers,

- Michael

-----Original Message-----
From: Lublinsky, Boris <boris.lublinsky@navteq.com>
To: mpoulin@usa.com <mpoulin@usa.com>
Sent: Thu, Jan 13, 2011 2:09 pm
Subject: RE: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Understood,

But it still presumes knowing what was there. It’s like you can always grab enough quotes from Bible to prove that God does not exist

From: mpoulin@usa.com [mailto:mpoulin@usa.com]
Sent: Thursday, January 13, 2011 6:48 AM
To: boris.lublinsky@navteq.com
Subject: Re: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Boris, I have not merged the changed texts back to the full document yet. I'do it based on the Jeff's comments soon. Sorry about this. I tried to extract enough context for each change to justify it.

- Michael

-----Original Message-----
From: Lublinsky, Boris <boris.lublinsky@navteq.com>
To: mpoulin@usa.com <mpoulin@usa.com>
Sent: Thu, Jan 13, 2011 3:36 am
Subject: RE: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Michael,

This is really hard to read.

Can you send a doc with the full content

From: mpoulin@usa.com [mailto:mpoulin@usa.com]
Sent: Wednesday, January 12, 2011 10:53 AM
To: jeffrey.a.estefan@jpl.nasa.gov; soa-rm-ra@lists.oasis-open.org
Subject: [soa-rm-ra] Re: [SOA-RAF] Interested in helping with Mgmt Model?

Hi Jeff,

here are two docs attached - one is the commented version and another one is the version with the comments 'accepted'. What I did, I tried to merge the Dan's version with my comments. I think that the diagram to be updated again based on the Dan's additions.

Please, let me know if I am moving into the agreed direction.

Regards,

- Michael

-----Original Message-----
From: Estefan, Jeff A (3100) <jeffrey.a.estefan@jpl.nasa.gov>
To: mpoulin@usa.com <mpoulin@usa.com>
Sent: Sun, Jan 9, 2011 9:28 pm
Subject: RE: [SOA-RAF] Interested in helping with Mgmt Model?

Michael,

I really appreciate taking time to go through the Mgmt model as currently

articulated in the RAF but we knew going in that the focus there was by and

large too focused on systems management and so that is why we gave Dan Heston

the opportunity to write something up, which he did and I've attached to this

note.  This work was done back in Sept of last year and there are a few e-mail

threads on the topic plus some discussion on the calls, but then the topic was

unfortunately dropped.   Nevertheless, I will review your recommended updates as

they look quite well thought out.  In the meantime, if you could review Dan's

draft update, that would be much appreciated.  It is still mission visual (UML)

models.

Thanks again!

 - Jeff

________________________________________

From: mpoulin@usa.com [mpoulin@usa.com]

Sent: Sunday, January 09, 2011 7:48 AM

To: Estefan, Jeff A (3100)

Subject: Re: [SOA-RAF] Interested in helping with Mgmt Model?

Jeffrey,

as promised, I have reviewed the Management Model section and attached my notes.

Since we deal with Management Concept, it influences other Concepts in the

document. This required me to look through the entire document and review every

occasion where management is mentioned. As a result, I have several notes and

proposals for changes in other sections, e.g. in Governance sections, there are

references to TOGAF 8.1 which are outdate already due to TOGAF 9.0 available for

a year already.

Thus, I have extracted all related text fragments into separate document and

modified them as needed. The line numbers have been preserved and match the line

numbers in the original document. I’ve excluded the fragments that I suggest are

correct, do not require modifications and do not raise questions.

I have edited the text on two computers and the side notes marked OP... are mine

as well as MP.

In addition to the text changes, I put some side comments with explanations of

my changes and questions. Plus, I added a few small sections such as Service

Configuration Manageability, Contracts ( in the Management Contracts and

Policies) and new term 'Solution' in the glossary. I've put all manageability

items into the new diagram (embedded).

I am looking for your comments and an advice on how to proceed from this point:

will this work go under your supervision or independently, like Peter/Chris do.

Regards,

- Michael

-----Original Message-----

From: Estefan, Jeff A (3100) <jeffrey.a.estefan@jpl.nasa.gov>

To: mpoulin@usa.com <mpoulin@usa.com>

Sent: Fri, Jan 7, 2011 7:14 pm

Subject: [SOA-RAF] Interested in helping with Mgmt Model?

Michael,

Are you interested in helping with the Management Model?  I was planning on

reviewing what was originally written in the RAF, which is very sparse and

largely focused on system management vs. services as managed entities and then

look over what Dan H. put together.  That material is missing visual models.

I feel this is far too important a concept, particularly its relationship to

governance, to simply “punt” and say we are not going to include it in our RAF

work.  I see it as even more paramount than the Testing Model although I know

you and Ken are making headway there.  If we can tackle both, great, but we

simply cannot punt on Management and only include Testing.

Just let me know your thoughts.

Regards…

- 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.

soa-rm-ra message