Uploaded image for project: 'Hadoop Common'
  1. Hadoop Common
  2. HADOOP-6148

Implement a pure Java CRC32 calculator

    XMLWordPrintableJSON

Details

    • Improvement
    • Status: Closed
    • Major
    • Resolution: Fixed
    • None
    • 0.21.0
    • performance, util
    • None
    • Reviewed

    Description

      We've seen a reducer writing 200MB to HDFS with replication = 1 spending a long time in crc calculation. In particular, it was spending 5 seconds in crc calculation out of a total of 6 for the write. I suspect that it is the java-jni border that is causing us grief.

      Attachments

        1. benchmarks20090714.txt
          9 kB
          Tsz-wo Sze
        2. benchmarks20090715.txt
          12 kB
          Tsz-wo Sze
        3. crc32-results.txt
          3 kB
          Todd Lipcon
        4. hadoop-5598.txt
          7 kB
          Todd Lipcon
        5. hadoop-5598.txt
          7 kB
          Todd Lipcon
        6. hadoop-5598-evil.txt
          8 kB
          Todd Lipcon
        7. hadoop-5598-hybrid.txt
          11 kB
          Todd Lipcon
        8. hadoop-6148.txt
          25 kB
          Todd Lipcon
        9. hadoop-6148.txt
          25 kB
          Todd Lipcon
        10. hadoop-6148.txt
          25 kB
          Todd Lipcon
        11. hdfs-297.txt
          19 kB
          Todd Lipcon
        12. PureJavaCrc32.java
          18 kB
          Tsz-wo Sze
        13. PureJavaCrc32.java
          17 kB
          Scott Carey
        14. PureJavaCrc32.java
          2 kB
          Scott Carey
        15. PureJavaCrc32.java
          2 kB
          Scott Carey
        16. PureJavaCrc32New.java
          16 kB
          Scott Carey
        17. PureJavaCrc32NewInner.java
          16 kB
          Scott Carey
        18. PureJavaCrc32NewLoop.java
          16 kB
          Scott Carey
        19. TestCrc32Performance.java
          3 kB
          Scott Carey
        20. TestCrc32Performance.java
          2 kB
          Scott Carey
        21. TestCrc32Performance.java
          2 kB
          Todd Lipcon
        22. TestCrc32Performance.java
          1 kB
          Todd Lipcon
        23. TestPureJavaCrc32.java
          3 kB
          Scott Carey

        Issue Links

          is related to

          Bug - A problem which impairs or prevents the functions of the product. HADOOP-8617 backport pure Java CRC32 calculator changes to branch-1

          • Major - Major loss of function.
          • Closed

          Improvement - An improvement or enhancement to an existing feature or task. MAPREDUCE-782 Use PureJavaCrc32 in mapreduce spills

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Closed
          relates to

          Bug - A problem which impairs or prevents the functions of the product. HADOOP-5318 Poor IO Performance due to AtomicLong operations

          • Major - Major loss of function.
          • Resolved

          Improvement - An improvement or enhancement to an existing feature or task. HADOOP-7401 Unit test TestPureJavaCRC32 warmup code warms up the not-so-important loop in PureJavaCRC32.update()

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Open

          Improvement - An improvement or enhancement to an existing feature or task. HADOOP-6166 Improve PureJavaCrc32

          • Major - Major loss of function.
          • Closed

          Improvement - An improvement or enhancement to an existing feature or task. HADOOP-7333 Performance improvement in PureJavaCrc32

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Improvement - An improvement or enhancement to an existing feature or task. HADOOP-7337 Annotate PureJavaCrc32 as a public API

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Improvement - An improvement or enhancement to an existing feature or task. HDFS-496 Use PureJavaCrc32 in HDFS

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Activity

            People

              scott_carey Scott Carey
              omalley Owen O'Malley
              Votes:
              0 Vote for this issue
              Watchers:
              20 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: