OGSA F2F Meeting Wednesday Morning, 14 March 2007 Attendees --------------- Mark Morgan (minutes) Jay Unger Steve McGough Hiro Kishimoto Fred Maciel Alan Linewski Ellen Stokes Andrew Grimshaw Chris Kantarjiev ZKert Chris Jordan Michel Drescher Andreas Savva * Agenda Bashing * Group Dinner - Group Dinner on Thursday - Brazilian Restaurant -- Cleo's 451 El Camino Real, San Bruno 650-615-9120 - Action Item: Get head count tomorrow * Roadmap Discussion - Chris stepping through document comments and issues. - Need a picture to really describe the "archtecture" part of OGSA. - Should we write articles for places like ACM Transactions, or IEEE to = give motivations for the dry specs we currently have. -- Target the technically literate end user. -- General consensus is that this is a good idea --- Action Item: Andrew to write a first draft with Chris = Kantarjiev. -- Outside of roadmap. - Action Item: Andrew to take 1 or 2 pager that he and Hiro wrote a = year ago and send it out to the group. - OGSA-DMI Functional Specification status clarified -- Hiros comment is outdated by the document moving to the Normative = part of hte OGSA Roadmap document - Chris to change ByteIO public comment period to August 2006. - Action Item: Chris to complete these changes and some of these = updates on the HPC Profile and Usecase Documents. Will send update = version for review and further comment. =09 Attendee Changes --------------------------- Add Bahareh Rahmanzadeh Remove Chris Jordan * Workflow - In BPEL, how are events named? - Can't change the workflow once its started - OGSA should say, what are the architectural requirements that we = should have in the grid. - The lack of dynamiscism and the lack of workflows as first class = entities seems like a serious drawback. - Can we agree to not write our own workflow language until we look at = what's out there. - We have to figure out exactly how we fit under nominal work flow = structures. - We need to collect the use cases for work flow (Platform, OMII, = Altair) - Action Item: We should talk to Chris Smith about this (Hiro?) - Action Item: Andrew to contact Bill Nitzberg about Workflow - Action Item: Steve McGough to talk to OMII about Workflow - Action Item: Chris K. to be the "whip" for workflow use case study. - Most common use case is (I have a big dataset and algorithm that can = work on a subset of the data set). - Action Item: Chris K. to provide template for use case within a = couple of days.