Uploaded image for project: 'Spark'
  1. Spark
  2. SPARK-42613

PythonRunner should set OMP_NUM_THREADS to task cpus instead of executor cores by default

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Resolved
    • Major
    • Resolution: Fixed
    • 3.3.0
    • 3.5.0
    • PySpark, YARN
    • None

    Description

      Follow up from https://github.com/apache/spark/pull/40199#discussion_r1119453996

      If OMP_NUM_THREADS is not set explicitly, we should set it to `spark.task.cpus` instead of `spark.executor.cores` as described in PR #38699.

      Attachments

        Issue Links

          is related to

          Bug - A problem which impairs or prevents the functions of the product. SPARK-41188 Set executorEnv OMP_NUM_THREADS to be spark.task.cpus by default for spark executor JVM processes

          • Major - Major loss of function.
          • Resolved

          Bug - A problem which impairs or prevents the functions of the product. SPARK-42596 [YARN] OMP_NUM_THREADS not set to number of executor cores by default

          • Major - Major loss of function.
          • Resolved

          Improvement - An improvement or enhancement to an existing feature or task. SPARK-28843 Set OMP_NUM_THREADS to executor cores reduce Python memory consumption

          • Major - Major loss of function.
          • Resolved
          links to

          Pull request #40212 [Github] Pull Request #40212 (jzhuge)

          Activity

            People

              jzhuge John Zhuge
              jzhuge John Zhuge
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: