September 2010 - pgi-wg - lists.ogf.org

OGF PGI - Requirements for Job Description
by Etienne URBAH 19 Oct '10

19 Oct '10

Morris, Johannes and all, Concerning the OGF PGI requirements for Job Description : From the EDGI use case 'Marshal Activities from Service Grid to Desktop Grid', the most important requirements, sorted by decreasing priority, are : JD3 (100) The Job Description document MUST reference all grid entities in conformance to the GLUE specification JD11 (107) Job Description should be only used for Job Description not for any kind of feedback. Job status information should not be communicated with a 'changed JSDL' JD4 (101) The Job Description document specification MUST permit the Client to request automatic data stage-in and stage-out JD22 (116) Clearly separate file and directory in the data-stagings. e.g. within the job description the element should be clearly labeled as file or directory, e.g. different aspect of handling directory or a file The other requirements labeled 'yes' are welcome, EXCEPT the 2 following ones, which generate too much COMPLEXITY in the Execution Service : JD6 (103) The Job Description document specification MUST permit the Client to request that at specified Activity states, the Execution service sets the Activity on 'Hold' [NO] JD8 (105) Job Description supports, 1) service-directed data-staging (i.e. specify data-stagings),2) client-initiated data-staging (i.e. specifiy data-staging but service will do nothing only control whether files existing), 3) manual data-staging (i.e. no specification of data-staging elements) [NO] Best regards. ----------------------------------------------------- Etienne URBAH LAL, Univ Paris-Sud, IN2P3/CNRS Bat 200 91898 ORSAY France Tel: +33 1 64 46 84 87 Skype: etienne.urbah Mob: +33 6 22 30 53 27 mailto:urbah@lal.in2p3.fr -----------------------------------------------------

1 3

ARC Use Case updated
by Oxana Smirnova 03 Oct '10

03 Oct '10

Hi all, I uploaded a strongly revised version of the (now single) ARC Use Case to gridforge. With this revision, reviewers' requirements should be satisfied. Cheers, Oxana

2 1

PGI Glossary in 30-day Public Comment Period
by Morris Riedel 30 Sep '10

30 Sep '10

Dear PGI team, our document has entered the 30 day public comment period: http://www.ogf.org/gf/docs/?public_comment Please encourage your colleagues and/or collaborators to submit comments. Your co-chair, Morris ------------------------------------------------------------ Morris Riedel Deputy Division Leader Distributed Systems and Grid Computing Division Jülich Supercomputing Centre (JSC) Info: http://www.fz-juelich.de/jsc/JSCPeople/riedel

1 0

Reminder & Agenda: PGI Telcon Today 2010-09-30; 3:30 pm CET
by Morris Riedel 30 Sep '10

30 Sep '10

Dear PGI team, today will be a telcon starting work towards the specification again: (1) Telcon-System and Screen-Sharing http://webconf.vc.dfn.de/ogf-pgi-wg/ No login is required. Just enter as 'guest' with your real world name. (2) Proposed agenda: (a) Debrief notes from the last meeting Reminder for use case/requirements presenters at OGF30! (b) Status: PGI Glossary - in public comment (c) Status: Use case document - 95% ready, waiting for one use case to integrate... (d) Discuss the specification creation process initially agreed on at OGF28 (Munich)/Workshops. Clarify next steps taking requirements/use cases into account. (e) AOB Please feel free to add any other topic on the agenda that I may have missed. Your co-chair, Morris ------------------------------------------------------------ Morris Riedel Deputy Division Leader Distributed Systems and Grid Computing Division Jülich Supercomputing Centre (JSC) Info: http://www.fz-juelich.de/jsc/JSCPeople/riedel

1 0

PGI Glossary and Terms: OGF Editor Submission
by Morris Riedel 27 Sep '10

27 Sep '10

Dear PGI team, I just submitted the PGI glossary document to the OGF editor (Greg Newby). The version submitted can be found here: http://forge.gridforum.org/sf/docman/do/downloadDocument/projects.pgi-wg/do… Minor comments from Christoph Witzig and Steven Newhouse have been integrated. More comments can be performed later during the public comment period if necessary. Your co-chair, Morris P.S. The use case document is ready too - I miss one final revised use case before this document can have a final call for group comments and OGF editor submission too. Ciao. ------------------------------------------------------------ Morris Riedel Deputy Division Leader Distributed Systems and Grid Computing Division Jülich Supercomputing Centre (JSC) Info: http://www.fz-juelich.de/jsc/JSCPeople/riedel

1 0

GROMACS use case updated
by Oxana Smirnova 27 Sep '10

27 Sep '10

Hi all, I forgot to report that I updated the GROMACS use case on Monday, taking into account all reviewers' input. Cheers, Oxana

2 1

QoS Use Case document updated
by Mark Morgan 27 Sep '10

27 Sep '10

I have uploaded the "QoS Computing" usecase document with revisions to address comments. Summarized as follows: * Comments from Morris Riedel - Comments embedded in the original -- I tried to address these as they popped up in the document -- Most of the comments I felt were not related to the document, but rather the use case and so I simply left the comments there for future reference. * Comments from Etienne - The template is BAD -- The template is the exact one that we as a group decided on (more than once) to adopt. Comment is therefor irrelevant and unhelpful. * Oxana - Seemed to like the document pretty much as is.

2 1

Mid-range Computing Use case modified
by Mark Morgan 27 Sep '10

27 Sep '10

I have uploaded the "Mid-range Computing" usecase document with revisions to address comments. Summarized as follows: * Comments from Morris Riedel - Comments embedded in the original -- I tried to address these as they popped up in the document - PGI section could be filled in with more detailed "requirements" that are not covered by BES, JSDL, etc. -- I'm not really sure what you mean here. That section says that those specs cover what PGI needs, so there are no additional requirements not covered by them. I can't really list requirements not covered by BES and JSDL without knowing details of PGI that I don't have * Comments from Etienne - The template is BAD -- The template is the exact one that we as a group decided on (more than once) to adopt. Comment is therefor irrelevant and unhelpful. * Oxana - Does not address necessity of common interfaces and standards -- No part of the use case requires those things, why should the use case have them? - Too generic -- Possibly, however, changing it to be less generic would significantly change the use case itself making it a new use case and so I left it as it is. if it is unacceptable in its current state, then that is OK, but otherwise it is what it is.

2 1

High Capacity Use Case updated
by Mark Morgan 27 Sep '10

27 Sep '10

I have uploaded the "High Capacity Computing" usecase document with revisions to address comments. Summarized as follows: * Comments from Morris Riedel - Clarify difference between HTC and HPC -- Based off of your comments embedded in the document itself, I have added a small clarification statement in the introductory paragraph - Title is confusing -- clarify -- Wasn't entirely sure what to do with this, so I changed the title to get rid of the "Traditional HPC User" portion, hopefully disambiguating the title. * Comments from Etienne - The template is BAD -- The template is the exact one that we as a group decided on (more than once) to adopt. Comment is therefor irrelevant and unhelpful. * Oxana - Does not address necessity of common interfaces and standards -- No part of the use case requires those things, why should the use case have them? - Not sure how its different from the mid-range use case -- I believe that it is, but that is a subjective assertion and I am willing to accept that it isn't different. If it isn't, then we can as a group decide to pick one or the other - Implementation suggestions all over the place -- I can't agree with this comment. There are a number of places where implementations are given as examples, but they are not suggested as solutions -- I also describe some implementation details in the "How does it relate to PGI" which I agree shouldn't be in there, but we all agreed as a group to adopt that template, and the template asked for those items and so they are there. In my opinion, they are there because of the template, but not part of the use case in general.

2 1

Reminder: PGI Set of Requirements & Upcoming Specification Work
by Morris Riedel 24 Sep '10

24 Sep '10

Hi PGI team, I looked up the latest-greatest requirement table that is available in GridForge: http://forge.gridforum.org/sf/docman/do/downloadDocument/projects.pgi-wg/do… According to the prioritized use cases we can narrow them down to the urgent important ones in order to proceed with the specification work. We have done the long process of doing the glossary and use cases. Now it's time to finally create the specs while these documents are soon at the OGF editor. In this context we said that anybody can come up with a specification that take the requirements based on the prioritized use cases into account. Please make yourself familiar with the requirements again. This enables us to finally identify the urgent ones derived from the use cases and render them into specifications during October. The discussion about this process will start next week in our Telcon, but please prepare the telcon with getting yourself familiar again with initial collected requirements/use cases. Let's aim at a rough consensus-based specification at Supercomputing 2010 (mid-Nov). Take care, Morris ------------------------------------------------------------ Morris Riedel Deputy Division Leader Distributed Systems and Grid Computing Division Jülich Supercomputing Centre (JSC) Info: http://www.fz-juelich.de/jsc/JSCPeople/riedel

1 0