Uploaded image for project: 'Hadoop HDFS'
  1. Hadoop HDFS
  2. HDFS-7433

Optimize performance of DatanodeManager's node map

    XMLWordPrintableJSON

Details

    • Improvement
    • Status: Resolved
    • Critical
    • Resolution: Fixed
    • 2.0.0-alpha, 3.0.0-alpha1
    • 2.8.0, 3.0.0-alpha1
    • namenode
    • Reviewed

    Description

      The datanode map is currently a TreeMap. For many thousands of datanodes, tree lookups are ~10X more expensive than a HashMap. Insertions and removals are up to 100X more expensive.

      Attachments

        1. HDFS-7433.patch
          3 kB
          Daryn Sharp
        2. HDFS-7433.patch
          6 kB
          Daryn Sharp
        3. HDFS-7433.patch
          6 kB
          Daryn Sharp
        4. HDFS-7433.patch
          9 kB
          Daryn Sharp

        Issue Links

          relates to

          Bug - A problem which impairs or prevents the functions of the product. HDFS-7457 DatanodeID generates excessive garbage

          • Major - Major loss of function.
          • Closed

          Improvement - An improvement or enhancement to an existing feature or task. HDFS-7434 DatanodeID hashCode should not be mutable

          • Major - Major loss of function.
          • Closed

          Activity

            People

              daryn Daryn Sharp
              daryn Daryn Sharp
              Votes:
              0 Vote for this issue
              Watchers:
              11 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: