Uploaded image for project: 'Spark'
  1. Spark
  2. SPARK-9437

SizeEstimator overflows for primitive arrays

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Resolved
    • Minor
    • Resolution: Fixed
    • 1.4.1
    • 1.5.0
    • Spark Core
    • None

    Description

      SizeEstimator can overflow when dealing w/ large primitive arrays eg if you have an Array[Double] of size 1 << 28. This means that when you try to broadcast a large primitive array, you get:

      java.lang.IllegalArgumentException: requirement failed: sizeInBytes was negative: -2147483608
         at scala.Predef$.require(Predef.scala:233)
         at org.apache.spark.storage.BlockInfo.markReady(BlockInfo.scala:55)
         at org.apache.spark.storage.BlockManager.doPut(BlockManager.scala:815)
         at org.apache.spark.storage.BlockManager.putIterator(BlockManager.scala:638)
      ...
      

      Attachments

        Activity

          People

            irashid Imran Rashid
            irashid Imran Rashid
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: