February 2006 - jsdl-wg - lists.ogf.org

FW: comments on ideas that might be accomplished by extending JSDL
by Brisard, Fred D 27 Feb '06

27 Feb '06

I forwarded this to the JSDL-WG membership so everyone can be aware of our discussions. Fred Brisard fred.brisard(a)ca.com -----Original Message----- From: David A. Alexander [mailto:alexanda@txcorp.com] Sent: Monday, February 27, 2006 4:05 PM To: Ly, An V Cc: Brisard, Fred D; 'Johnson, Tony S.'; 'Jerome LAURET' Subject: RE: comments on ideas that might be accomplished by extending JSDL Hi An Ly, We have one reference implementation so far and that is our own Java Application client (and a command line program) to access the STAR Unified Meta Scheduler (see http://www.star.bnl.gov/STAR/comp/Grid/scheduler/) I have attached the RDL specification document that we have so far that also describes some use cases and our thoughts on the applicable situations and the possible limitations. I would be good if we could collaborate in some. For example, maybe we could propose to make our effort a special investigation for or under the guidance of the JSDL working group or something to that effect? I noticed in your last projection of what you would like to do with JSDL there was talk of work flows. We would be very interested to see how we could be of service in this area. Thanks, dave > -----Original Message----- > From: Ly, An V [mailto:An.Ly@ca.com] > Sent: Monday, February 27, 2006 1:45 PM > To: alexanda(a)txcorp.com > Cc: Brisard, Fred D > Subject: RE: comments on ideas that might be accomplished by extending > JSDL > > Hi David, > > I'm a member of JSDL and I reviewed the paper regarding RDL/TDL that you > sent to the JSDL group. The concept of a user-oriented/high-level way to > describe the (grid) job or its status is interesting. Would it be > possible to get examples/use cases of your applications that are > currently described/managed in the RDL/TDL-like manner? > > Thanks, > An Ly > > -----Original Message----- > From: Brisard, Fred D > Sent: Monday, February 27, 2006 2:44 PM > To: Ly, An V > Subject: RE: comments on ideas that might be accomplished by extending > JSDL > > > > Fred Brisard > fred.brisard(a)ca.com > > -----Original Message----- > From: David A. Alexander [mailto:alexanda@txcorp.com] > Sent: Monday, February 27, 2006 1:35 PM > To: ali(a)epcc.ed.ac.uk; andreas.savva(a)jp.fujitsu.com; Brisard, Fred D; > darren(a)pulsipher.org; scrumb(a)ggf.org; asm100(a)doc.ic.ac.uk > Subject: RE: comments on ideas that might be accomplished by extending > JSDL > > Gentlemen, > > I didn't hear back from anyone about my question. I realize that you > may be > busy and it may take more time and effort to respond to an email than it > would to talk in person. Therefore, would you mind if I called you to > talk > about the JSDL and how we our projects might be mutually beneficial? > > Thanks, > David Alexander > Cell 720.581.6737 > > _________________________________________________ > David Alexander desk: 303-448-7751 > Tech-X Corporation fax: 303-448-7756 > Boulder, CO email: alexanda(a)txcorp.com > (303)448-7751 http: www.techxhome.com > > > -----Original Message----- > > From: David A. Alexander [mailto:alexanda@txcorp.com] > > Sent: Thursday, February 23, 2006 10:30 AM > > To: 'ali(a)epcc.ed.ac.uk'; 'andreas.savva(a)jp.fujitsu.com'; > > 'fred.brisard(a)ca.com'; 'darren(a)pulsipher.org'; 'scrumb(a)ggf.org'; > > 'asm100(a)doc.ic.ac.uk' > > Cc: 'Jerome LAURET'; 'Johnson, Tony S.' > > Subject: comments on ideas that might be accomplished by extending > JSDL > > > > > > Dear JSDL Working Group Administrators & Chairs, > > > > We are formulating ideas on how to develop a "request language" (or > what > > in your community might be best termed a high-level request language > that > > a consuming submission system would turn into a workflow). I have > looked > > at the 1.0 specification of JSDL and we have a lot of overlapping > concerns > > except that it says that workflow issues are out of the scope of JSDL. > > Attached is a short white paper on our idea as it stands. We would > > appreciate any comments on how you think we can collaborate or where > our > > ideas fit into your views of JSDL or other specifications that already > > exist. We are not trying to re-invent the wheel. However, we do wish > to > > provide a way for Open Science Grid users to specify what they need at > a > > high-level beyond the individual job. > > > > Thanks, > > > > David Alexander, Tech-X Corporation > > > > On behalf of our team that includes: > > Jerome Lauret, Brookhaven National Lab > > Tony Johnson, Stanford Linear Accelerator Center > > > > > > _________________________________________________ > > David Alexander desk: 303-448-7751 > > Tech-X Corporation fax: 303-448-7756 > > Boulder, CO email: alexanda(a)txcorp.com > > (303)448-7751 http: www.techxhome.com

1 0

RE: [jsdl-wg] Future plans for JSDL
by Lee, William 23 Feb '06

23 Feb '06

+1 on the multiple <xsd:any/>, I don't really know what the precise semantic is. Do they collapse into one? How do you deal with multiplicity? ... Anyway, the toolings I've worked with seems to be able to cope with that (SOAPScape and Sun J2EE server). Another way we can do this is to define the new Application element and manually add all the posix:XYZ elements that we need in a sequence without using any of the XSD extension mechanism. William --- William Lee - Software Coordinator --- --- London e-Science Centre, Imperial College London --- A: Room 380, Department of Computing, Imperial College London, Huxley Building, South Kensington campus, London SW7 2AZ, UK E: wwhl at doc.ic.ac.uk | william at imageunion.com W: www.lesc.ic.ac.uk | www.imageunion.com P: +44(0)20 7594 8310 F: +44(0)20 7581 8024 -----Original Message----- From: Donal K. Fellows [mailto:donal.k.fellows@manchester.ac.uk] Sent: 23 February 2006 13:12 To: Lee, William Cc: JSDL WG Subject: Re: [jsdl-wg] Future plans for JSDL Lee, William wrote: > Roughly.... > <xsd:complexType name="MPIApplicationType"> > <xsd:complexContent> > <xsd:extension base="jsdl-posix:POSIXApplication_Type"> > <xsd:sequence> > <xsd:element ref="ProcessorCount"/> > </xsd:sequence> > </xsd:extension> > </xsd:complexContent> > </xsd:complexType> How does that work with <xsd:any>? When I've tried that sort of thing, the results have tended to be less than satisfactory when tooled up (multiple anys in the element, etc.) Donal.

2 1

RE: [jsdl-wg] Future plans for JSDL
by Lee, William 23 Feb '06

23 Feb '06

In the parallel extension we have created in GridSAM. We have a new element called <mpi:MPIApplication/> that extends the jsdl:posix:POSIXApplication_Type. Therefore the MPIApplication inherits the arguments, executable, etc. with the additional elements to specify parallelism. Roughly.... <xsd:complexType name="MPIApplicationType"> <xsd:complexContent> <xsd:extension base="jsdl-posix:POSIXApplication_Type"> <xsd:sequence> <xsd:element ref="ProcessorCount"/> </xsd:sequence> </xsd:extension> </xsd:complexContent> </xsd:complexType> <xsd:element name="ProcessorCount" type="xsd:positiveInteger"/> <xsd:element name="MPIApplication" type="MPIApplicationType"/> William --- William Lee - Software Coordinator --- --- London e-Science Centre, Imperial College London --- A: Room 380, Department of Computing, Imperial College London, Huxley Building, South Kensington campus, London SW7 2AZ, UK E: wwhl at doc.ic.ac.uk | william at imageunion.com W: www.lesc.ic.ac.uk | www.imageunion.com P: +44(0)20 7594 8310 F: +44(0)20 7581 8024 -----Original Message----- From: owner-jsdl-wg(a)ggf.org [mailto:owner-jsdl-wg@ggf.org] On Behalf Of Donal K. Fellows Sent: 23 February 2006 09:11 To: JSDL WG Subject: Re: [jsdl-wg] Future plans for JSDL A S McGough wrote: > 3) Prioritise the items. From the session parallel jobs seemed to be > the thing most people wanted next. Just a quick note to say that the main reason for doing the parallel jobs stuff next (probably starting from the base proposed by the people from Barcelona; they've identified the stuff that's needed after all) is that otherwise there'll be several different groups doing it themselves in an incompatible way. To me, that says it is an *excellent* idea to foster a standard extension for this and to expedite this. OK, I think other extensions/improvements are more important in the grand scheme of things, but the fact that people are doing it now forces our hand. Looking at the parallel proposal, I have to ask how it is supposed to be combined with the POSIX extension, since without that it's not going to work (still need command-line args, env-vars, etc.) Once we've nailed that down, I think it will be just A Simple Matter Of Writing. :-) Donal.

2 1

Future plans for JSDL
by A S McGough 23 Feb '06

23 Feb '06

All, We had a session here at GGF16 where we brain stormed some ideas for where JSDL should go next. This also included some material which had come up from the last OGSA f2f meeting. Attached are the edited slides from this session which lists some of the areas where people think that JSDL should look into. The things that need doing now include: 1) Request from the list for items that were missed out from the session. 2) Classify these items as: a) Out of Scope (TM) b) Extensions for JSDL 1.0 c) New functionality for JSDL 2.0 d) Separate to JSDL versions (eg items for a primer document) 3) Prioritise the items. From the session parallel jobs seemed to be the thing most people wanted next. steve.. -- ------------------------------------------------------------------------ Dr A. Stephen McGough http://www.doc.ic.ac.uk/~asm ------------------------------------------------------------------------ Technical Coordinator, London e-Science Centre, Imperial College London, Department of Computing, 180 Queen's Gate, London SW7 2BZ, UK tel: +44 (0)207-594-8409 fax: +44 (0)207-581-8024 ------------------------------------------------------------------------

3 2

RE: JSDL Spec (GFD-R-P.056) typo? - FileSystemType - tmp or temporary
by Lee, William 23 Feb '06

23 Feb '06

It's indeed a typo. The schema reads "temporary". The example is incorrect. It will go into the post-v1 tracker. William > Date: Wed, 22 Feb 2006 14:01:27 -0500 > From: Tom Seelbach <seelbach(a)us.ibm.com> > To: jsdl-wg(a)gridforum.org > Subject: JSDL Spec (GFD-R-P.056) typo? - FileSystemType - tmp or > temporary > > The FileSystemTypeEnumeration includes "temporary" but on page 23 and > 24 I see <jsdl:FileSystemType>tmp</jsdl:FileSystemType>. Should "tmp" > be "temporary" in these cases? > > Regards, > Tom Seelbach - Grid Development - seelbach(a)us.ibm.com > > --- William Lee - Software Coordinator --- --- London e-Science Centre, Imperial College London --- A: Room 380, Department of Computing, Imperial College London, Huxley Building, South Kensington campus, London SW7 2AZ, UK E: wwhl at doc.ic.ac.uk | william at imageunion.com W: www.lesc.ic.ac.uk | www.imageunion.com P: +44(0)20 7594 8310 F: +44(0)20 7581 8024

1 0

JSDL Spec (GFD-R-P.056) typo? - FileSystemType - tmp or temporary (fwd)
by Ali Anjomshoaa 23 Feb '06

23 Feb '06

Tom, please sign up to the JSDL-WG mailing list to post to the group. The traffic is light. Take care, Ali -- ---------------------------------------------------- |epcc| - Ali Anjomshoaa EPCC, University of Edinburgh James Clerk Maxwell Building Mayfield Road E-mail: ali(a)epcc.ed.ac.uk Edinburgh, EH9 3JZ Phone: + 44 (0) 131 650 5818 United Kingdom Fax: + 44 (0) 131 650 6555 ------------------------------------------------------------- ---------- Forwarded message ---------- Date: Wed, 22 Feb 2006 14:01:27 -0500 From: Tom Seelbach <seelbach(a)us.ibm.com> To: jsdl-wg(a)gridforum.org Subject: JSDL Spec (GFD-R-P.056) typo? - FileSystemType - tmp or temporary The FileSystemTypeEnumeration includes "temporary" but on page 23 and 24 I see <jsdl:FileSystemType>tmp</jsdl:FileSystemType>. Should "tmp" be "temporary" in these cases? Regards, Tom Seelbach - Grid Development - seelbach(a)us.ibm.com

1 0

RE: [jsdl-wg] Future plans for JSDL
by Brisard, Fred D 16 Feb '06

16 Feb '06

Hi Steve, My comments -- [Slide 3] Consider adding "Conditional data staging" - that is, allow some form of conditional processing of data staging. For example, on stage in, it would be useful to support the capability of the same data being staged from more than one source in case one of the sources is unavailable. On stage out, it would be useful to only stage out data if the program ran correctly - possibly even detecting a result value and staging data out based on that value. Regarding the item "Relationship to CDDLM" - The relationship regarding data staging with CDDLM is confusing to me. I believe that data staging is a "provisioning" issue yet we seem to accept that data staging is acceptable within JSDL and other forms of provisioning are not. Maybe we should be identifying required resources in JSDL that reference entities within CDDLM. These entities could be anything related to provisioning. Even the concept of the machine attributes could be handled this way. Maybe JSDL should get out of the business of identifying specific resource information and just refer to resources known to CDDLM. [Slide 4] Title should be Application types and adjust body of slide Consider "workflow" as a job type. Workflows can typically be invoked as a web service, but we delineate other job types such as "Scripting Language" that can be specified as command line jobs so why not specify workflow as a job type. [Slide 5] Regarding "workflow" - maybe a JSDL document is just a workflow activity. [Slide 6] Regarding "Scheduling to ..." - would data availability be covered under "events"? [Overall] There are lots of good thoughts here; and lots of opportunity for more hard work. I continue to be concerned with the large number of GGF specifications and their relationships. OGSA-BES is trying to integrate components created using these specs, but I believe that we still need a better understanding of the overall architecture that binds these components. Best regards, Fred Fred Brisard fred.brisard(a)ca.com -----Original Message----- From: owner-jsdl-wg(a)ggf.org [mailto:owner-jsdl-wg@ggf.org] On Behalf Of A S McGough Sent: Thursday, February 16, 2006 7:03 AM To: JSDL WG Subject: [jsdl-wg] Future plans for JSDL All, We had a session here at GGF16 where we brain stormed some ideas for where JSDL should go next. This also included some material which had come up from the last OGSA f2f meeting. Attached are the edited slides from this session which lists some of the areas where people think that JSDL should look into. The things that need doing now include: 1) Request from the list for items that were missed out from the session. 2) Classify these items as: a) Out of Scope (TM) b) Extensions for JSDL 1.0 c) New functionality for JSDL 2.0 d) Separate to JSDL versions (eg items for a primer document) 3) Prioritise the items. From the session parallel jobs seemed to be the thing most people wanted next. steve.. -- ------------------------------------------------------------------------ Dr A. Stephen McGough http://www.doc.ic.ac.uk/~asm ------------------------------------------------------------------------ Technical Coordinator, London e-Science Centre, Imperial College London, Department of Computing, 180 Queen's Gate, London SW7 2BZ, UK tel: +44 (0)207-594-8409 fax: +44 (0)207-581-8024 ------------------------------------------------------------------------

1 0

JSDL: GGF16 Workshop Slides
by Ali Anjomshoaa 15 Feb '06

15 Feb '06

Hi All, you can find the Workshop slides from this morning at: https://forge.gridforum.org/docman2/ViewCategory.php?group_id=122&category_… Enjoy and please send questions and discussion items to the list here. Take care, Ali -- ---------------------------------------------------- |epcc| - Ali Anjomshoaa EPCC, University of Edinburgh James Clerk Maxwell Building Mayfield Road E-mail: ali(a)epcc.ed.ac.uk Edinburgh, EH9 3JZ Phone: + 44 (0) 131 650 5818 United Kingdom Fax: + 44 (0) 131 650 6555 -------------------------------------------------------------

1 0

JSDL Mapping
by A S McGough 14 Feb '06

14 Feb '06

Folks, We're trying to make a list of those systems that JSDL can currently be mapped to. In time for the JSDL workshop on Wednesday. If you have a mapping can you please check below: Known to be map-able: ================ Globus 2 Condor Fork (Sun) Grid Engine Unicore Possible Mapping: - could someone confirm these? ============ LSF PBS No Mapping: - If you have a system not listed above please could you let us know! ========= -- ------------------------------------------------------------------------ Dr A. Stephen McGough http://www.doc.ic.ac.uk/~asm ------------------------------------------------------------------------ Technical Coordinator, London e-Science Centre, Imperial College London, Department of Computing, 180 Queen's Gate, London SW7 2BZ, UK tel: +44 (0)207-594-8409 fax: +44 (0)207-581-8024 ------------------------------------------------------------------------

4 5

JSDL Workshop at GGF16
by Ali Anjomshoaa 10 Feb '06

10 Feb '06

Dear colleagues, [My apologies for cross-posting] Just a quick announcement of the JSDL-WG Workshop at GGF16. Date: Wed. 15th Feb. Time: 08:30 - 12:00 Room: Mycenae The talks are in the schedule below. We all look forward to seeing you there and to your feedback and contributions towards our ongoing work in the JSDL-WG. Kind regards, Ali (on behalf of the JSDL-WG) JSDL-WG: https://forge.gridforum.org/projects/jsdl-wg/ JSDL-Spec: http://www.gridforum.org/documents/GFD.56.pdf -- ---------------------------------------------------- |epcc| - Ali Anjomshoaa EPCC, University of Edinburgh James Clerk Maxwell Building Mayfield Road E-mail: ali(a)epcc.ed.ac.uk Edinburgh, EH9 3JZ Phone: + 44 (0) 131 650 5818 United Kingdom Fax: + 44 (0) 131 650 6555 ------------------------------------------------------------- ---- GGF16, JSDL Workshop Schedule ---- 08:30 - Welcome and Introductions 08:35 - JSDL Tutorial Andreas Savva (Fujitsu Laboratories Ltd, Japan) Steve McGough (LeSC, Imperial College London, UK) William Lee (Lesc, Imperial College London, UK) 09:20 - GridSAM presentation William Lee, (LeSC, Imperial College London, UK) 09:40 - NAREGI presentation Kazushige Saga (National Institute of Informatics, Japan) 10:00 - *** BREAK *** 10:30 - HPCEuropa presentation Ariel Oleksiak (Poznan Supercomputing and Networking Centre, Poland) 10:50 - Japan Business Grid presentation Andreas Savva (Fujitsu Laboratories Ltd, Japan) 11:10 - UniGrids presentation Michel Drescher (Fujitsu Laboratories of Europe, UK) 11:30 - Platform Computing presentation Christopher Smith (Platform Computing, USA) 11:50 - Conclusions and closing remarks 12:00 - END ----------

1 0