Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-1123

requirements and specifications for initial version of inter-Task communication

    Details

    • Type: Epic
    • Status: Won't Fix
    • Resolution: Done
    • Fix Version/s: None
    • Component/s: ctrl_events, pipe_base
    • Labels:
      None
    • Epic Name:
      Inter-Task Communication
    • Story Points:
      50
    • WBS:
      02C.07.01.03.03
    • Team:
      Data Facility
    • Cycle:
      None

      Description

      gather requirements and specification for Implementation of a gather/scatter mechanism for a set of Tasks executing in parallel using the Event Services, providing for eventual extension to support MPI or other communication mechanisms.
      Start mid Jan 2015, finish Feb 2015.

      JK: In PMCS this would be Pietrowicz S

        Attachments

          Issue Links

          Stories in Epic (Custom Issue Matrix)

          Key Summary Story Points Assignee Status
           
          DM-1344

          Define gather/scatter mechanism for tasks with generic API

          8 Steve Pietrowicz Invalid

            Activity

            Hide
            price Paul Price added a comment -

            HSC uses a process pool that communicates over MPI: https://github.com/HyperSuprime-Cam/hscPipeBase/blob/master/python/hsc/pipe/base/pool.py

            One advantage is that the user doesn't need to know any MPI or care about the implementation (there are no if comm.rank == 0 embedded throughout the user code). I'd be happy to work with you on improving it if it seems like a suitable foundation for your work.

            Show
            price Paul Price added a comment - HSC uses a process pool that communicates over MPI: https://github.com/HyperSuprime-Cam/hscPipeBase/blob/master/python/hsc/pipe/base/pool.py One advantage is that the user doesn't need to know any MPI or care about the implementation (there are no if comm.rank == 0 embedded throughout the user code). I'd be happy to work with you on improving it if it seems like a suitable foundation for your work.
            Hide
            spietrowicz Steve Pietrowicz added a comment -

            Sounds good, Paul. Thanks! It'll be a while before I get to this task, but I'll drop you a line when I do.

            Show
            spietrowicz Steve Pietrowicz added a comment - Sounds good, Paul. Thanks! It'll be a while before I get to this task, but I'll drop you a line when I do.
            Hide
            mgelman2 Margaret Gelman added a comment -

            Unsetting the cycle for this epic to take it out of the W15 baseline. This needs to be discussed further and prioritized before being assigned to a cycle.

            Show
            mgelman2 Margaret Gelman added a comment - Unsetting the cycle for this epic to take it out of the W15 baseline. This needs to be discussed further and prioritized before being assigned to a cycle.
            Hide
            jbecla Jacek Becla added a comment -

            Has this been discussed? It is triggering a variance because it was originally scheduled for W15.

            Show
            jbecla Jacek Becla added a comment - Has this been discussed? It is triggering a variance because it was originally scheduled for W15.
            Hide
            petravick Donald Petravick added a comment -

            I understand that we have subsequently understood that requirements for this are needed. This relies in part on understanding the the new pipeline execution model for ARP which has recently beeb produced.

            Show
            petravick Donald Petravick added a comment - I understand that we have subsequently understood that requirements for this are needed. This relies in part on understanding the the new pipeline execution model for ARP which has recently beeb produced.
            Hide
            jbecla Jacek Becla added a comment -

            Margaret, based on Don's comment perhaps this can be marked "done"?

            Show
            jbecla Jacek Becla added a comment - Margaret, based on Don's comment perhaps this can be marked "done"?
            Hide
            plutchak Joel Plutchak (Inactive) added a comment -

            No longer relevant. Subsequent design of task execution was accomplished in other epics, and this work was reassigned to a different team.

            Show
            plutchak Joel Plutchak (Inactive) added a comment - No longer relevant. Subsequent design of task execution was accomplished in other epics, and this work was reassigned to a different team.

              People

              • Assignee:
                mgelman2 Margaret Gelman
                Reporter:
                ktl Kian-Tat Lim
                Watchers:
                Darko Jevremovic, Donald Petravick, Jacek Becla, Joel Plutchak (Inactive), Kian-Tat Lim, Margaret Gelman, Paul Price, Steve Pietrowicz
              • Votes:
                0 Vote for this issue
                Watchers:
                8 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved:

                  Summary Panel