Uploaded image for project: 'HPCC'
  1. HPCC
  2. HPCC-13559

Streaming Dataset to Python on Thor ends up processed on a single node

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Resolved
    • Priority: Not specified
    • Resolution: Duplicate
    • Affects Version/s: 5.0.14
    • Fix Version/s: None
    • Component/s: None
    • Labels:
      None

      Description

      When I use the streaming API with Python and submit a job that takes a well-distributed (randomly distributed) dataset as input, the actual Python call only happens on a single node, with all data redistributed to that node. The resulting dataset is completely skewed to a single node. It doesn't seem to me there's any reason to do this - the Python call should work fine when called with each node's piece of the dataset.

        Attachments

          Issue Links

            Activity

              People

              • Assignee:
                ghalliday Gavin Halliday
                Reporter:
                lneric lneric
              • Votes:
                0 Vote for this issue
                Watchers:
                3 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: