- ur-wg - lists.ogf.org

Test message
by Laura F McGinnis 05 Jul '06

05 Jul '06

I sent an email to the UR-WG list on Monday, but haven't received it back via the list yet. If you receive this, please let me know. Thx LM ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Laura F. McGinnis, Project Coordinator Data & Information Resource Services Pittsburgh Supercomputing Center email: lfm(a)psc.edu 300 South Craig St, #313 phone: 412-268-5642 Pittsburgh, PA 15213 fax: 412-268-5832 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

1 0

Concall on Thursday?
by Laura F McGinnis 22 Jun '06

22 Jun '06

I am not available for the concall this week. If you all want to continue, maybe with donal's version 2 work, that would be fine with me. If you do have the call, please post notes to the list. I'll keep working on comment responses and version 1 and post when I have material to be reviewed. Thx LM ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Laura F. McGinnis, Project Coordinator Data & Information Resource Services Pittsburgh Supercomputing Center email: lfm(a)psc.edu 300 South Craig St, #313 phone: 412-268-5642 Pittsburgh, PA 15213 fax: 412-268-5832 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

3 3

Updated Version 1 of document & changes to website
by Laura F McGinnis 20 Jun '06

20 Jun '06

I have added context and history to the first part of the usage record document (version 1). I have also updated the website to indicate that there are now two documents being worked on. Please see www.psc.edu/~lfm/PSC/Grid/UR-WG/ for the updates and links to the MSWord and PDF formats for the version 1 document. I'd rather not email to the group, since the document is long. As soon as I get a document from Donal for version 2, I'll post that onto the website. Thx LM ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Laura F. McGinnis, Project Coordinator Data & Information Resource Services Pittsburgh Supercomputing Center email: lfm(a)psc.edu 300 South Craig St, #313 phone: 412-268-5642 Pittsburgh, PA 15213 fax: 412-268-5832 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

1 0

Meeting Notes
by Donal K. Fellows 14 Jun '06

14 Jun '06

I'm not very good at this note taking business, but here's what I've got: ---------------------------------- Collaboration with OGSA good idea once things are solidified. Future version to extend to more complex things than batch system job accounting. Stick with using email for resolving issues; not worth changing technologies half way through. Short-term Timeline: Mailing list discussions to close: 30 June Consensus: 15 July Revised document: 15 August Submit to editor: 1 September Three-way joint meeting: OGSA RUS UR Call for use cases and reference implementations Telecons ======== Keep going. Biweekly. One hour. 1300 UTC. First on 25 May. ---------------------------------- Donal Fellows.

4 4

Concall tomorrow
by Laura F McGinnis 09 Jun '06

09 Jun '06

The concall is scheduled for tomorrow (Thursday), June 8. Call-in details are online at http://www.psc.edu/~lfm/PSC/Grid/UR-WG/ConCallInfo.html. We will discuss the draft that Donal posted to the mailing list last week. It's in the UR-WG email archives at http://www-unix.gridforum.org/mail_archive/ur-wg/2006/05/msg00082.html. Thx LM ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Laura F. McGinnis, Project Coordinator Data & Information Resource Services Pittsburgh Supercomputing Center email: lfm(a)psc.edu 300 South Craig St, #313 phone: 412-268-5642 Pittsburgh, PA 15213 fax: 412-268-5832 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

4 3

Concall meeting notes - 8 June 06
by Laura F McGinnis 08 Jun '06

08 Jun '06

On call: Laura, Sven, Stephen, Donal, Matt 1) Discussion of Donal's Strawman ++ All elements are documented, to make the schema a bit easier to comprehend ++ Compound Usage Record (CUR) - can be composed of CURs ++ Atomic Activity Usage Record - instead of contentiously-named "job" ++ Key fields: Record Identity End Time Start Time User Identity ++ This structure should allow atomic records to be more easily collected 2) Identity ?? Can a usage record have a non-human "owner" ++ Atomic record should have a user identity; CUR may or may not have user identity ++ Note that the Record Identity also has an additional ID field in it ?? What types of Identities are needed? ++ Record ++ User ++ Resource ?? Organization 3) How different is Donal's strawman from what we've already got? Very: probably a version 2.0 (given that the current version is 1.0) ++ Addressing the comments should give us a credible v1.0, but will it be usable? ++ There are already sites ("Early Adopters") using their interpretations of v1.0; can't alienate them ++ GFSG would also like to see v1.0 "done" ++ v2.0 is still a valid pursuit, tho ++ Will need to consider providing mappings between v1.0 and v2.0 4) Where do we go from here? ++ Wrap up v1.0, but include a promise of v2.0 ++ v2.0 development can occur concurrently with v1.0 wrap up; will help inform some of the responses ++ Need to engage Early Adopters in v2.0 development ++ v1.0 is primarily an editor task at this point; ACTION ITEM (Laura): ++ Respond to gridforge comments via mailing list ++ Mark changes in v1.0 document with footnotes for UR-WG members to note ++ Once mailing list grants consensus, post to gridforge comment and close ++ Donal will continue to work on v2.0 as time (and peer pressure :)) permit Thx LM ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Laura F. McGinnis, Project Coordinator Data & Information Resource Services Pittsburgh Supercomputing Center email: lfm(a)psc.edu 300 South Craig St, #313 phone: 412-268-5642 Pittsburgh, PA 15213 fax: 412-268-5832 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

1 0

UR-WG Concall meeting notes - May 25 2006
by Laura F McGinnis 08 Jun '06

08 Jun '06

(Sorry for the delay - too much holiday & life interfering with work) On the call: Stephen, Matt, Donal, Sven, Laura 1) Status on comments - how active is the mailing list? ++ Getting good response to comments posted to mailing list. Will keep encouraging use of mailing-list as key communication tool. 2) Review of GGF17 notes ++ Aggregation seems possible with current schema, but should be reviewed with more detail ++ "Aggregate" vs "Composite" vs "Summary" are not well-defined, but it's not worth dwelling too much on this right now, since the initial specification is for an atomic record and any combinations thereof have already been declared Out of Scope. ACTION ITEM (Donal): Make sure atomic schema is complete and sufficient 3) Discussion of comments that have been responded to - Do we have consensus on anything yet? ++ Comment 6c/9: Coordinate Status values with SAGA/BES model; whatever states they have, UR needs to support ++ ACTION ITEM (Laura): Remove wording that indicates "Job Completion"; references to end-date/time may indicate a snapshot rather than a true end-of-job ++ Comment 3.9 refers to job status; it should be address/covered in response to 6c/9 ++ Comment 5b: the discussion focused on start-time, but the question/comment is more about the definition of a "grid job". Discussion needs to be refocused. ++ Key issue is to get definitions for "atomic", "composite", and "aggregate" circulating. ACTION ITEM (Donal): develop working definition for "atomic" ACTION ITEM (Laura): draft context text for document and circulate to mailing list 4) Discussion of comments that have not been responded to - as time permits. <<time did not permit>> Thx LM ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Laura F. McGinnis, Project Coordinator Data & Information Resource Services Pittsburgh Supercomputing Center email: lfm(a)psc.edu 300 South Craig St, #313 phone: 412-268-5642 Pittsburgh, PA 15213 fax: 412-268-5832 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

1 0

Usage Record Schema, v1.0, candidate 1
by Donal K. Fellows 08 Jun '06

08 Jun '06

Hi everyone! As discussed on the concall (if you weren't there, where were you?) we're going to push out a version 1.0 of the usage record spec that is compatible with the existing specs so generators need minimal/no change to adapt, and then work on a version 2 of the specs that will allow us to properly answer the issues raised by our last public comment round. As part of that, please find attached a candidate for the version 1.0 schema. What follows is a total summary of *all* changes from the schema released for the second public comment period; everything else is the same. * The schema namespace has been changed to be GFD.58-compatible. New schema namespace is: http://schemas.ggf.org/ur/2006/06/usage.xsd * A copyright declaration has been added. * A documentation-annotation has been added to the Status element's declaration to note that BES state names SHOULD be legal State values. * A documentation-annotation has been added to the metric attribute's global declaration to explain what it is meant for. A matching doc-annotation has been added to the UR type declaration. That should help make it less misunderstood! Note that most of those are documentation alterations. Only the schema namespace is different, and that's to follow best-practice. What I sent around previously should be considered to be draft 1 of version 2, and so the next time I send it around, it should be draft 2. Donal.

1 0

Prototype of strawman
by Donal K. Fellows 31 May '06

31 May '06

Hi everyone! Here's a snapshot of where I've got to with doing an updated usage record schema. It's massively incomplete when it comes to listing all the elements, but it's got some of the flavour of what I intended. The main things to note are that it defines a CompoundUsageRecord and an AtomicActivityUsageRecord (which is incomplete) that are both in effect subtypes of UsageRecord (which has a few elements associated with it) and that I've split apart the concepts of Memory and NodeCount into PeakMemory and AverageMemory (and PeakNC and AverageNC respectively). There are probably more useful concepts to identify, and there's certainly more work to do copying things across. I'm also trying to put in sane documentation comments as I go. In any case, comments/brickbats would be welcome before I invest much more time and effort into this. :-) (I'm also using a somewhat-proper namespace URI now.) Donal.

1 0

Concall Agenda: 25-May
by Laura F McGinnis 24 May '06

24 May '06

Usage Records concalls will resume tomorrow, 25-May at 2pm BST/9am EDT (Pittsburgh). Call-in information is below. The agenda is: 1) Status on comments - how active is the mailing list? 2) Review of GGF17 notes 3) Discussion of comments that have been responded to - Do we have consensus on anything yet? 4) Discussion of comments that have not been responded to - as time permits. Thx LM Primary Dial-in Number: 0844.888.8888 (UK 5p National Rate) Alternate Dial-in Numbers: +44.870.088.5706 (For Callers Outside of the U.K.) 01805.004.102 (For Callers in Germany) The Participant Access Code is: 61652. ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Laura F. McGinnis, Project Coordinator Data & Information Resource Services Pittsburgh Supercomputing Center email: lfm(a)psc.edu 300 South Craig St, #313 phone: 412-268-5642 Pittsburgh, PA 15213 fax: 412-268-5832 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

1 0