[glue-wg] GLUE WG teleconference reminder, Wednesday, May 16, 2012

Thu May 17 11:41:50 EDT 2012

Thank you for your active participation in the last meeting.

As agreed, I have attached the meeting notes, the document will be uploaded later on the gridforge wiki page.

Best regards,
Shiraz

On Tue, May 15, 2012 at 5:28 PM, JP Navarro <navarro at mcs.anl.gov<mailto:navarro at mcs.anl.gov>> wrote:
Greetings Glue WG,

Coordinates and agenda for tomorrow's GLUE WG teleconference follow:

Coordinates:
: Wednesday, May 16, 2012
: 15:00-16:00 PM Central European Time (CET), 9:00-10:00 AM US Eastern Time (EST)

Teleconference and screen sharing:
: https://webconf.vc.dfn.de/ogf-glue-wg/ (Adobe Connect)
: Login as Guest using your real name

Agenda:
1) Proposal by David Meredith to introduce Service and Endpoint substitution groups (20 minutes)
  Discussion points:
  a) What are the benefits for using substitution groups?
  b) Should we apply this design to Service and Endpoint elements only, or also to other key elements?
  c) Next steps: Integrate this proposal in the XML rendering document? Any other next steps?

2) Finalizing an XML rendering agreement (20 minutes)
  Discussion points:
  a) What are the benefits of globalizing elements?
  b) Which elements do we want to globalize, should we make all significant elements root elements?
  c) How to represent cross-references/associations for globalized elements?
  Background material:
  - Globalized schema: https://github.com/OGF-GLUE/XSD/blob/master/schema/GLUE2.xsd
  - TeraGrid/XSEDE schema: https://github.com/OGF-GLUE/XSD/blob/master/schema/teragrid_glue_2.0_r01.xsd

3) Other proposed/outstanding GLUE2 changes (10 minutes)
  a) EGI service name and type values. What is the goal? Where do we record agreements on Enumerated values?
  b) Are there other attributes that we need a consistent enumeration for?

4) Future meetings (10 minutes)
  OGF 35 in Delft, Sunday, June 17, 13:30-15:00, room 1.180
  Other agenda items besides discussing each of the above 3 topics?

  Should we plan to hold more teleconferences?
  Teleconference to prepare for OGF 35 the week of May 29 thru June 1?

See you tomorrow,

Shiraz Memon
JP Navarro

_______________________________________________
glue-wg mailing list
glue-wg at ogf.org<mailto:glue-wg at ogf.org>
https://www.ogf.org/mailman/listinfo/glue-wg

--
Ahmed Shiraz Memon
Federated Systems and Data
Jülich Supercomputing Centre (JSC)

Phone: +49 2461 61 6899
Fax:     +49 2461 61 6656

------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich
Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher
Geschaeftsfuehrung: Prof. Dr. Achim Bachem (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt
------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------

Kennen Sie schon unseren neuen Film? http://www.fz-juelich.de/film
Kennen Sie schon unsere app? http://www.fz-juelich.de/app
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ogf.org/pipermail/glue-wg/attachments/20120517/eca6b2f1/attachment.html>
-------------- next part --------------
Date: 16.05.2012

Time: 15:00 CEST

Duration: 1 hr

Participants:

JP Navarro 
Shiraz Memon 
David Meredith 
Balazs Konya 
Etienne Urbah 
Florido Paganelli 
John Casson 
Stephen Burke 

JP: What are the benefits for using substitution groups?

David: Context and benefits of having substitution groups in GLUE XSD. The requirements of such are derived from GOCDB information system. Advantage of having abstract endpoint and service to define our own new specialized services and endpoints and take advantage of highly constrained schema. This can be a base profile for the glue2.The proliferation of new service and endpoint types would not raise concern as this can be defined in a new profile. Therefore a new profiled schema could extend the original glue2 schema inheriting the abstract service and endpoint types.

The new schema with examples demonstrating the aforesaid changes can be downloaded from:
http://tools.ngs.ac.uk/ngstools/glue2proposal/src.zip

JP: What other elements besides Service and Endpoint could be considered for the substitution groups?

Shiraz: Not that we require

Florido: Me neither, perhaps Manager or Computing Manager, might be the new elements may pop-up once the schema is in production

Stephen: People may want to extend any of the core classes

Balazs:  problem that everytime you want to extend an element we will need to introduce schema change, therefore every core element should be included

JP: possible candidate entities could be Administrative domains, as it is quite broad concept and could be used outside the glue context and people may want to extend it into other applications

JP: we continue to discuss this in our next call, whether all the core classes can be included eventually? We have three options: either we vote on these two elements that are proposed or we could adapt all the core classes?

Shiraz: it is reasonable to have all the major core classes

Florido: the idea is to avoid another schema change

Shiraz: the number of core classes should be limited in order to avoid introducing the unnecessary complexity

ACTION - Florido and David will give a proposal containing the core classes requiring substitution groups and that will be discussed in the next meeting

JP: The proposal will be discussed in next call - there will be document changes. Can those document changes be worked on while we are still working on the proposal or we have to wait until the proposal is finalised?

Shiraz: The rendering document should be worked on side by side

Balazs: there is NO official rendering document, that is the big issue

JP: suggestion would be to continue with the changes in it and ready by OGF, is this reasonable?

Shiraz: the xml rendering document should be ready for discussion by the next OGF

Balazs: i'd like to see the xml rendering document to be planned for public coment submission containing David's proposal

JP: does it require any changes in the spec. too?

David: No, in spec. it only suggests that you should derive sub-type specializations which is only relevant for specific rendering

ACTION - David will provide the xml rendering document in two weeks (by next phone meeting)

JP: by next meeting we should be ready with the draft to be circulated for the public comment and we close the public comment by OGF, is it too aggressive? or people see urgency of having this document?

Balazs: i'd like to avoid further delays with the renderings

JP: We can have draft by the meeting for internal comments

JP: There are two XML renderings out there: teragrid/xsede and the global one. What are the characterizations and differences

ACTION - JP: I will talk to Warren by the meeting in two weeks

Shiraz: The xsede project is using private and public schema, while using their own flavor of glue2 schema. However I have created two (public and private) example instances using the Globalized schema.

JP: I expect to adopt the globalized schema, one reason that we have used glue2 "global" element with everything under as sub.elements. This can be a non-issue as this is implementation specific and globalized schema provides entities that could be wrapped into another schema under some global element.

JP: What are the benefits of having global elements?

Shiraz: both public and private xml instances can be created by only re-using the specific global elements

David: One of the benefits of global elements is to compose and re-use in third party schemas, and doing the substitution groups make most of the elements global

JP: one could compose two valid public and private, however I will engage with Warren and discuss the changes

JP: EGI has been trying to standardise the service names and types

David: It was an informal agreement containing the suggestions defining the consistent enumeration

Florido: collect the names and create a consistent directory for someone referring to specific service

Balazs: the plan was that EGI-EMI-Globus prepares a proposal for enumerationsd and it gets submitted to OGF glue group then after agreement gets uploaded to an OGF-GLUE-wiki

Florido: It is not only limited to service name or type enumerations but all the open enumerations in the schema. And the the activity is independent from xml schema

JP: The people may enumerate same service with different names, having profiles can enable interoperability while classifying the services in a unified way

Shiraz, Stephen: All the enumerations should be collected on gridforge wiki

Florido: They are being collected on google docs, will be posted on wiki after agreements

JP: I would support glue2 names and types profile document

John, Shiraz: Having a document requires lot of effort, thus wiki page is more suitable for collecting the enumerations

JP: If wiki then what is the process of updating the wiki page?

Stephen: the main point is to record the types in use so everyone knows what they mean...

Florido: In the short term the wiki is sufficient, but in the long term I agree the services being published provide their profiles in a separate document. The collected enumerations will be forwarded to the ogf group. 

ACTION - Florido will send a link to the glue2 wiki page containing a link to the google docs

Balazs: in Lund we are reviwing the ldap rendering draft 

ACTION - Balazs plan to send a commented ldap document later this week

JP: OGF35 announcement, room information, and next phone meeting

Etienne: I will re-send an email concerning the proposals of extending the activity entity 