Uploaded image for project: 'Hadoop HDFS'
  1. Hadoop HDFS
  2. HDFS-5790

LeaseManager.findPath is very slow when many leases need recovery

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Closed
    • Major
    • Resolution: Fixed
    • 2.3.0
    • 2.4.0
    • namenode, performance
    • None
    • Reviewed
    • Committed to branch-2 and trunk.

    Description

      We recently saw an issue where the NN restarted while tens of thousands of files were open. The NN then ended up spending multiple seconds for each commitBlockSynchronization() call, spending most of its time inside LeaseManager.findPath(). findPath currently works by looping over all files held for a given writer, and traversing the filesystem for each one. This takes way too long when tens of thousands of files are open by a single writer.

      Attachments

        1. hdfs-5790.txt
          5 kB
          Todd Lipcon
        2. hdfs-5790.txt
          4 kB
          Todd Lipcon

        Issue Links

          duplicates

          Improvement - An improvement or enhancement to an existing feature or task. HDFS-4183 Throttle block recovery

          • Critical - Crashes, loss of data, severe memory leak.
          • Resolved
          is duplicated by

          Bug - A problem which impairs or prevents the functions of the product. HDFS-11326 FSNamesystem closeFileCommitBlocks block FSNamesystem Ops

          • Critical - Crashes, loss of data, severe memory leak.
          • Open

          Activity

            People

              tlipcon Todd Lipcon
              tlipcon Todd Lipcon
              Votes:
              0 Vote for this issue
              Watchers:
              8 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: