February 2008 - ogsa-dmi-wg

HPC File Staging extension in public comment
by Steven Newhouse 27 Feb '08

27 Feb '08

The HPC File Staging Profile is now in public comment http://www.ogf.org/Public_Comment_Docs/Documents/2008-02/HPC%20File%20Stagi… Abstract: This document profiles the File staging capabilities of the Job Submission Description Language (JSDL) for use by HPC Basic Profile-compliance services. It includes clarifications, refinements, interpretations and amplifications of JSDL which promote interoperability. Please review and make comments (positive or negative): https://forge.gridforum.org/sf/discussion/do/listTopics/projects.ggf-editor… Thanks, Steven

1 0

OGSA DMI Functional Specification
by Steven Newhouse 27 Feb '08

27 Feb '08

The OGSA DMI Functional Specification is now in public comment http://www.ogf.org/Public_Comment_Docs/Documents/2008-02/ogsa-dmi-spec-edit… Abstract The Open Grid Services Architecture Data Movement Interface (OGSA-DMI) specification defines a standardized mechanism for moving data from its source to its destination. By abstracting this movement, the client complexity for moving data within a grid will be greatly reduced. OGSA-DMI defines two port types for initiating, scheduling and managing data transfers from a given source to a specified destination. The source and destination are described through Data End Point References (DEPRs), a specialized form of a WS-Address. An OGSA-DMI service will use the properties of the data to be transferred and the properties of the infrastructure to decide what the best underlying transfer protocol to undertake the actual data transfer is for the given combination of data source and sink. By this means OGSA-DMI provides a common interface, though its abstraction layer, that can be used to span the preferred transfer mechanisms used across many different types of Grid infrastructures. This version of the specification aims to ensure that the OGSA-DMI port types are composable with other Web Service (WS) specifications whilst being agnostic about any specific WS rendering. OGSA-DMI is only concerned with the moving of data, i.e. bytes. Any associated semantics are the responsibility of higher level services that leverage off OGSA-DMI. The OGSADMI framework is general enough to cater for the movement of single elements of data or composites. Please review and make comments (positive or negative): https://forge.gridforum.org/sf/discussion/do/listTopics/projects.ggf-editor… Thanks, Steven

1 0

OGF-22
by Michel Drescher 25 Feb '08

25 Feb '08

Folks, as I plan to dial in, can you please advise me of the phone bridge number? Cheers, Michel

3 2

DMI Telcon
by Mario Antonioletti 13 Feb '08

13 Feb '08

Hi, A short reminder that we will be having our usual DMI telcon today (if enough folks turn up). I will not also be able to make it - my apologies: Date: 13th February 2007 When: 9am (Pacific) 10am (Eastern) 5pm (UK) PIN: 2719140 USA: +1 866 500 6738 (Toll: +1-203-480-8000) UK: +44 118 909 2000 a plausible agenda for this meeting will be: Agenda: 1. IPR Notice 2. Agenda Bashing 3. Rreview of Steven's OGF slides 4. Planning 5. AOB Can change the above agenda on the call as required. Mario +-----------------------------------------------------------------------+ |Mario Antonioletti:EPCC,JCMB,The King's Buildings,Edinburgh EH9 3JZ. | |Tel:0131 650 5141|mario@epcc.ed.ac.uk|http://www.epcc.ed.ac.uk/~mario/ | +-----------------------------------------------------------------------+

3 2

Draft slides for OGF22
by Steven Newhouse 13 Feb '08

13 Feb '08

Draft slides for the DMI session at OGF22. Updated from the slideset used last time. Discussion on the list or on Wednesday's call. Steven

2 1

Apologies
by Ravi Madduri 12 Feb '08

12 Feb '08

Hi I cannot attend tomorrow's call as I have some personal business to take care of around the same time. Sorry about that. Regards -- Ravi K Madduri The Globus Alliance | Argonne National Laboratory | University of Chicago http://www-unix.mcs.anl.gov/~madduri

1 0

Proposed Specification Changes
by Steven Newhouse 07 Feb '08

07 Feb '08

All, I know we've just tried to close the document down... but I've been doing a bit of coding and have some suggested changes and documentation errors from the implementation to date. Hope these seem sensible. Steven Changes 4.2.1.2.1 - Change the names in the DEPR (same semantics but clearer meaning) AvailableProtocols --> DataLocations Protocol -> Data (Attributes: Name -> Protocol & url -> DataUrl 4.2.1 requestDataTransferInstance -> GetDataTransferInstance 4.2.2 getFactoryAttributesDocument -> GetFactoryAttributesDocument 4.3 Add a fault: NoDataLocationsInEpr Change NoTransferProtocolAggrementFault -> NoSourceSinkProtocolMatchFault 5.1.2.1 (& 5.2.7.1.1) Change Status element to State element 5.2 Capitalise first letter of each state change operation 5.2.4 In the text we have a restart operation. In the state diagram we have a resume operation and a resumed event. Something needs to be rationalized here! Personally I prefer changing to the resume operation - restart to me has a different meaning (going back to the beginning) while resume starts mid-transfer after coming out of the suspended state. 5.2.6 getStatus -> GetState Editorial Errors 4.1.1.1 - cross referencing errors 5.1.1.1 - ditto 5.1.4.1 5.2.6.1

5 5

Minutes
by Ravi Madduri 07 Feb '08

07 Feb '08

DMI Telcon - 02/06/08 ===================== Attendees: Mario Antonioletti, EPCC Steven Newhouse, Microsoft Allen Luniewski, IBM Ravi Madduri, ANL Agenda: 1. IPR Notice 2. Agenda Bashing 3. Comments on the Spec 4. Acknowledgements 5. Suggested Spec changes from Steven Newhouse 6. AOB 1. Comments on the Spec Ravi send the doc to James and Gavin (from CERN) and responses are due in a week. Nobody else had yet commented.. 2. Acknowledgements List of people on the acknowledgements looks fine. The following are the list of people: James Casey Bill Allcock Andrew Grimshaw Dave Snelling Gavin Mccance Alex Sim 3. Spec Changes from Steven Steven tried to implement the spec. Issues in naming were discussed. Separation of protocol and data location was discussed. There was a general consensus on what Steven suggested and sent mail to the DMI list. All other changes steven suggested also received general consensus. Steven will post his changes back to gridforge by end of the week along with list of acknowledgements. Ravi will then submit to the Editor project in gforge. Steven did not find any ambiguity in the spec while implementing it in C# other than the fact that Steven is minting DataEPRs in his implementation. HPCP has a file staging spec which is protocol agnostic. DMI will have a role when HPCP implementation ships the staging to a DMI implementation instead of implementing that functionality. -- Ravi K Madduri The Globus Alliance | Argonne National Laboratory | University of Chicago http://www-unix.mcs.anl.gov/~madduri

1 0

Updated specificaiton
by Steven Newhouse 06 Feb '08

06 Feb '08

Minor alterations in the specification change tracked from todays call... https://forge.gridforum.org/sf/docman/do/downloadDocument/projects.ogsa-dmi…

1 0

DMI Telcon
by Mario Antonioletti 05 Feb '08

05 Feb '08

Hi, A short reminder that we should be having a DMI telcon this Wednesday. Date: 6th February 2008 When: 9am (Pacific) 10am (Central) 5pm (UK) PIN: 2719140 USA: +1 866 500 6738 (Toll: +1-203-480-8000) UK: +44 118 909 2000 a plausible agenda for this meeting will be: Agenda: 1. IPR Notice 2. Previous Minutes and Action Review 3. Agenda Bashing 4. Spec Review - Addressing any comments received thus far - Acknowledgements - Some changes requested by Steven 5. Planning 6. AOB Outstanding actions: None as far as I know. Can change the above agenda on the call as required. Mario +-----------------------------------------------------------------------+ |Mario Antonioletti:EPCC,JCMB,The King's Buildings,Edinburgh EH9 3JZ. | |Tel:0131 650 5141|mario@epcc.ed.ac.uk|http://www.epcc.ed.ac.uk/~mario/ | +-----------------------------------------------------------------------+

1 0