OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

soa-rm message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [soa-rm] David Linthicum Says: "ESB versus Fabric.Stop It!"


I expect a single service description is sufficient to define any  
infrastructure or non-infrastructure service.  I think we can note some  
number will likely exist and interact, possibly stating sometimes a  
service may provide functionality and sometimes it may consume the  
functionality of another service.

However, there are certain fundamental capabilities (or at least  
accounting for concepts) that must exist for anyone to care to build or  
use some nonzero number of services.  Frank often brings up semantics.   
I've talked about security (and am still uncomfortable that this is  
easily subsumed by policy).  This gets back to the idea of structural  
integrity because indeed something analogous to structural integrity  
may need to be part of our RM.

Should the RM have one part that describes concepts that will have an  
identifiable instantiation and others that are a property, possibly  
distributed among the parts, of the while?

Ken

On May 25, 2005, at 12:35 PM, Duane Nickull wrote:

> It would be far to concrete for a reference model to dive into  
> infrastructure and services to support it.
>
> Duane
>
> Chiusano Joseph wrote:
>
>> <Quote>
>> Intuitively, I think that if I have some minimal level of  
>> infrastructure
>> (messaging, discovery, and mediation) and I expose one single
>> non-infrastructure service on this infrastructure, I have an SOA.
>> </Quote>
>>
>> Which implies that we may want to distinguish between "infrastructure"
>> and "application" services for our SOA RM (not necessarily advocating,
>> just pointing out the notion). That is, as long as this notion is not
>> too concrete for the RM - if so, it may be part of an RA.
>>
>> Joe
>>
>> Joseph Chiusano
>> Booz Allen Hamilton
>> Visit us online@ http://www.boozallen.com
>>
>>
>>> -----Original Message-----
>>> From: Christopher Bashioum [mailto:cbashioum@mitre.org] Sent:  
>>> Wednesday, May 25, 2005 9:40 AM
>>> To: soa-rm@lists.oasis-open.org
>>> Subject: RE: [soa-rm] David Linthicum Says: "ESB versus Fabric.Stop  
>>> It!"
>>>
>>> Joe,
>>>
>>> I'm beginning to think that the question you are asking (and have  
>>> been asking ; ) carries something more subtle that I don't believe  
>>> we have addressed yet.  It is the idea of intent.  I have been of  
>>> the impression that the intent of SOA is service opacity and  
>>> location opacity (i.e., you can't see behind the interface (allows  
>>> for replacement of parts) and you can't see where the service is on  
>>> the network (implies discovery mechanism).
>>> But - when it comes to the actual services, the intent there is to  
>>> create the interface in such a way as to allow for re-purposing.  In  
>>> other words, as I create a service, I include as an implied  
>>> requirement that it will be used by consumers I don't know in a way  
>>> that I can't foresee.
>>> It is this idea of intent that I think we are having a hard time  
>>> capturing in the RM.  I think your concern about multiple services  
>>> is another way of saying the same thing.  The problem with the  
>>> number of services is it really may not capture the intent.  For  
>>> example, if I have 4 services - is that really sufficient for an  
>>> SOA?  I'm not sure.  However, if I have at least the infrastructure  
>>> services that enable an SOA (yet to be defined, but conceptually  
>>> referred to as an ESB, or discovery, messaging, and mediation -
>>> whatever) do I have an SOA?  Or yet again, if I have the  
>>> infrastructure and one non-infrastructure service, do I then have an  
>>> SOA?
>>>
>>> Intuitively, I think that if I have some minimal level of  
>>> infrastructure (messaging, discovery, and mediation) and I expose  
>>> one single non-infrastructure service on this infrastructure, I have  
>>> an SOA.
>>>
>>> -----Original Message-----
>>> From: Chiusano Joseph [mailto:chiusano_joseph@bah.com]
>>> Sent: Wednesday, May 25, 2005 9:13 AM
>>> To: soa-rm@lists.oasis-open.org
>>> Subject: RE: [soa-rm] David Linthicum Says: "ESB versus Fabric.Stop  
>>> It!"
>>>
>>>
>>>> -----Original Message-----
>>>> From: Duane Nickull [mailto:dnickull@adobe.com]
>>>> Sent: Tuesday, May 24, 2005 2:56 PM
>>>> To: Michael Stiefel
>>>> Cc: soa-rm@lists.oasis-open.org
>>>> Subject: Re: [soa-rm] David Linthicum Says: "ESB versus Fabric.Stop  
>>>> It!"
>>>>
>>>> Endpoints are part of a service description IMO.  Orchestration of  
>>>> multiple services is out of the scope of
>>> the core RM,
>>>> much the same way as how multiple houses are positioned
>>> next to each
>>>> other in a grid layout is un-necessary in order to define a RM for  
>>>> house.
>>>>
>>>> A service or house do not have to exist amongst multiple houses in  
>>>> order to be services/houses.
>>>>
>>> Which brings us back to what I believe is the single most important  
>>> question for us to answer: Does one service constitute a SOA? Or are  
>>> 2 or more services required?
>>>
>>> If 2 or more services are required, then it seems to me that in  
>>> order to call something a *SOA* reference model, the notion of  
>>> multiple services must be incorporated - as that is the minimal  
>>> amount of information necessary to *effectively* represent/model the  
>>> "targeted entity" (which is SOA) for the intended audience.
>>>
>>> If one service constitutes a SOA, this implies that a SOA may have  
>>> more than one service. It then seems to me that one has a choice for  
>>> their
>>> RM: include only a single service in the model, or include multiple  
>>> services. The question then becomes which approach enables the most  
>>> effective representation for the intended audience.
>>>
>>> So as you see, I believe everything flows from this single most  
>>> important question.
>>>
>>> Joe
>>>
>>> Joseph Chiusano
>>> Booz Allen Hamilton
>>> Visit us online@ http://www.boozallen.com
>>>
>>>> Duane
>>>>
>>>> Michael Stiefel wrote:
>>>>
>>>>
>>>>> Could we then conceive of endpoints and orchestration in such a  
>>>>> fashion? Or is the critical point aspect or attribute in
>>> which case
>>>>> endpoint qualifies, but orchestration does not.
>>>>>
>>>>> To make a grammatical analogy, the RM defines a substantive, and  
>>>>> therefore adjectives (aspects and attributes) are part of
>>>>>
>>>> the RM, but
>>>>
>>>>> verbs (actions) are not.
>>>>>
>>>>> (side note: I know verbs have aspect, but we are not
>>> using the term
>>>>> that way).
>>>>>
>>>>> Michael
>>>>>
>>>>> At 02:34 PM 5/24/2005, Duane Nickull wrote:
>>>>>
>>>>>
>>>>>> Since Structural Integrity is an aspect of all houses,
>>> it could be
>>>>>> part of a RM as an abstract concept.  Even if you do not
>>>>>>
>>>> explicitly
>>>>
>>>>>> design a house to have a certain set of structural integrity  
>>>>>> parameters, it still does.  It is not a component
>>> itself, just an
>>>>>> aspect or attribute.
>>>>>>
>>>>>> Duane
>>>>>>
>>>>>>
>>>>>> Michael Stiefel wrote:
>>>>>>
>>>>>>
>>>>>>> I thought of structural integrity in terms of the entire
>>>>>>>
>>>> house, not
>>>>
>>>>>>> just a wall, but I think your point remains the same.
>>>>>>>
>>>>>>> Granted that each architecture needs to specify its structural  
>>>>>>> integrity, but shouldn't the RM have the concept of structural  
>>>>>>> integrity since it is an abstract concept shared by all RAs.
>>>>>>>
>>>>>>> Michael
>>>>>>>
>>>>>>> At 02:06 PM 5/24/2005, Duane Nickull wrote:
>>>>>>>
>>>>>>>
>>>>>>>> The RM does not necessarily have to get into cardinality
>>>>>>>>
>>>> rules IMO,
>>>>
>>>>>>>> unless they are very obvious.  In the case of a house,
>>>>>>>>
>>>> you may not
>>>>
>>>>>>>> make consistent rules stating that every house has to
>>>>>>>>
>>>> have at least
>>>>
>>>>>>>> three walls since a wall can be curved or any number of
>>>>>>>>
>>>> walls from
>>>>
>>>>>>>> 3 up.  You may be able to infer from the relationships
>>>>>>>>
>>>> that there
>>>>
>>>>>>>> is a certain cardinality if the RM for a house said that
>>>>>>>>
>>>> each room
>>>>
>>>>>>>> has one door.
>>>>>>>> That would declare an association between the number
>>> of rooms to
>>>>>>>> the number of doors.
>>>>>>>>
>>>>>>>> Structural integrity is an aspect of a wall, which must be  
>>>>>>>> specialized for each architecture based on a number
>>>>>>>>
>>>> criteria.  The
>>>>
>>>>>>>> RM declares what the wall is and its' purpose, the
>>> architect has
>>>>>>>> the job of specifying the actual walls to be used for each  
>>>>>>>> architecture and ensuring they map back to requirements.
>>>>>>>>
>>>>>>>> You are right - analogies are not definitions, however I
>>>>>>>>
>>>> have found
>>>>
>>>>>>>> them very useful in conveying the meaning.
>>>>>>>>
>>>>>>>> Duane
>>>>>>>>
>>>>>>>> Michael Stiefel wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>>> Does the RM understand that some of the concepts are
>>> unique and
>>>>>>>>> some multiple (without an exact number, you could have one  
>>>>>>>>> circular wall, 3 walls, 4 walls, etc.)?
>>>>>>>>>
>>>>>>>>> Using your analogy, how does the RM deal with
>>> concepts such as
>>>>>>>>> structural integrity. Structural integrity would apply to all  
>>>>>>>>> house RAs. In my way of thinking concepts such as
>>> endpoints or
>>>>>>>>> orchestration are analogous to this.
>>>>>>>>>
>>>>>>>>> In the analogy I would see the reference architecture
>>>>>>>>>
>>>> as Colonial
>>>>
>>>>>>>>> American Reference Architecture, or even more specifically  
>>>>>>>>> Colonial American Cape Ann, or Colonial American
>>> Greek Revival
>>>>>>>>> reference architectures.
>>>>>>>>>
>>>>>>>>> Analogies are useful, but they are not definitions.
>>>>>>>>>
>>>>>>>>> Michael
>>>>>>>>>
>>>>>>>>> At 12:56 PM 5/24/2005, Duane Nickull wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>> RA means Reference Architecture.  As per the previous
>>>>>>>>>>
>>>> emails on
>>>>
>>>>>>>>>> this subject, it is a generalized architecture.
>>>>>>>>>>
>>>>>>>>>> The relationship is that architects use a RM as a
>>>>>>>>>>
>>>> guiding model
>>>>
>>>>>>>>>> when building a RA.
>>>>>>>>>>
>>>>>>>>>> For example, if you are architecting a house, an RM
>>>>>>>>>>
>>>> may explain
>>>>
>>>>>>>>>> the concepts of gravity, a 3D environment, walls,
>>> foundations,
>>>>>>>>>> floors, roofs, ceilings etc.  It is abstract however.
>>>>>>>>>>
>>>> There is
>>>>
>>>>>>>>>> nothing specific like a wall with measurements such
>>> as 8 feet
>>>>>>>>>> high.  Note that the RM has only one each of these
>>> things - it
>>>>>>>>>> does not have 4, 16, 23 walls, just one as a concept.
>>>>>>>>>> The architect may uses this model to create a specific  
>>>>>>>>>> architecture for a specific house (accounting for such
>>>>>>>>>>
>>>> things as
>>>>
>>>>>>>>>> property, incline, climate etc) or an architect MAY
>>>>>>>>>>
>>>> elect to use
>>>>
>>>>>>>>>> it to build a more generalized reference architecture.  The  
>>>>>>>>>> latter is often done by architects who design houses.
>>>>>>>>>>
>>>> When they
>>>>
>>>>>>>>>> sell a house, they must often re-architect the RA
>>> for specific
>>>>>>>>>> implementation details such as incline of land,
>>>>>>>>>>
>>>> climate, facing
>>>>
>>>>>>>>>> the sun etc..
>>>>>>>>>>
>>>>>>>>>> So why do we need a RM?  Simple - we now have logical
>>>>>>>>>>
>>>> divisions
>>>>
>>>>>>>>>> amongst the components of a house and what they mean.
>>>> That way,
>>>>
>>>>>>>>>> when a company says " we are a flooring company..", that is  
>>>>>>>>>> meaningful since we all know what that means.  The
>>>>>>>>>>
>>>> same applies
>>>>
>>>>>>>>>> to a roofing company.  Without the basic consensus on
>>>>>>>>>>
>>>> the logical
>>>>
>>>>>>>>>> divisions, a roofing contractor may also try to include the  
>>>>>>>>>> ceiling and walls as part of his offerings.
>>>>>>>>>> That would not work and not allow the general
>>>>>>>>>>
>>>> contractor to build
>>>>
>>>>>>>>>> a house very easily since there may not be consensus
>>> upon the
>>>>>>>>>> division of labor and components to build the house.
>>>>>>>>>>
>>>>>>>>>> Do you guys think an explanation of this nature may
>>> be good to
>>>>>>>>>> include in the introduction section?
>>>>>>>>>>
>>>>>>>>>> Duane
>>>>>>>>>>
>>>>>>>>>> Chiusano Joseph wrote:
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>> What is an RA? What is the relationship between an RM
>>>>>>>>>>>
>>>> and an RA?
>>>>>>>>>>> What is
>>>>>>>>>>> the RM->RA path for SOA?
>>>>>>>>>>>
>>>>>>>>>>> Matt also submitted last week (I believe) that we may
>>>>>>>>>>>
>>>> not even
>>>>
>>>>>>>>>>> need an RA. How should that change our notion of RM,
>>>>>>>>>>>
>>>> if at all?
>>>>
>>>>>>>>>>> Joe
>>>>>>>>>>>
>>>>>>>>>>> Joseph Chiusano
>>>>>>>>>>> Booz Allen Hamilton
>>>>>>>>>>> Visit us online@ http://www.boozallen.com
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>
>>>
>>>
>>>
>>>
------------------------------------------------------------------------ 
------------------
Ken Laskey
MITRE Corporation, M/S H305     phone:  703-983-7934
7515 Colshire Drive                        fax:        703-983-1379
McLean VA 22102-7508




[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]