09/10/2007 6:11 PM
post5879
|
Comments received by e-mail
(1) Balazs Konya sent me the following comments.
Please find below my (nordugrid) comment on your document. Something is wrong
with the www.ogf.org server, therefore so far i was not able to upload my
comment (i'll keep on trying this afternoon).
regards,
Balazs
-----------------------------------------
The draft "Guidelines for Information Modeling for OGSA(TM) Entities"
is well-written and very valuable document providing recommendations on
grid entity modeling based on the CIM methodology.
One of the largest shortcomings of the Open Grid Services Architecture (GFD.80)
is the lack of consistent information model. This problem has already manifested
itself via the ad-hoc creation of implicit, temporary, limited-scope information
models to be found in e.g. the JSDL, BES specifications.
We (NorduGrid) fully support the statement on page 6 that the
"target for OGSA entities should be one information model, in order to unify the
concepts in the whole architecture and avoid translation of semantics"
This guideline document provides an important first step towards this goal,
proposes the use of a modeling process based on CIM methodology (Section 2).
Section 3 "Roles & Responsibilities" puts the Resource Management Design Team of
the OGSA Working Group in a central coordinator role for information modeling
within Open Grid Forum. In general, Nordugrid agrees that a "centralized work is
necessary", however we are not sure how much the coordinating role of the
RM Design Team is acknowledged by the other OGF working groups,
whether they are aware of this fact at all. To our understanding
Section 3 presents OGF organizational issues, does that belong to a technical
document?
Further details:
- I'd propose to change the title in order to include CIM since the presented
content is very much CIM specific (it even has two appendices on CIM)
"Guidelines for CIM-based Information Modeling for OGSA Entities"
- page 9, Section header: shouldn't it be RM Design Team instead of "Design
Teams"? To my understanding the section outlines the relation of OGF groups and
the RM Design Team.
- page 3, Introduction, first sentence: the "and services" at the end of
the long sentence is confusing, i'd remove them. The sentence currently reads as
"OGSA services span multiple areas ... and multiple activities .... and services"
- page 5, last sentence of the paragraph below the figure: "definitions of
information and data model above .. do not match RFC 3444". That RFC is about
"On the Difference between Information Models and Data Models". An explanatory
sentence on the deviation from the RFC 3444's definitions would be useful.
Balazs Konya
KnowARC Project
NorduGrid Collaboration
-------------------------------------------------------------------------------------
(2) The following comments were received from Laurence Field:
First of all I would like to reiterate the comments from Sergio and
Balazs and say that I think this is a good document.
What is not clear to me from reading the document is the different
between using the modelling methodology and the model itself. At the
start of the document it seems that we are defining a process and by the
end of the document it seems that we are taken a model as well. The
methodology to define a model and the model itself are two different
things. This document must make it clear that OGSA and hence the OGF has
decided to use CIM as the information model and extend it. The rest of
the document is the process of how the OGF can work with the DMTF to
extend the model.
I would suggest something like.
"The OGSA-WG has decided to leverage the modelling experience used in
CIM from the DMTF and in addition it has also been decided to build upon
the existing models. "
I agree that we should use a proven methodology for defining an
information model and building upon something so as not to reinvent...
View Full Message
|
|
|