Uploaded image for project: 'Hadoop Common'
  1. Hadoop Common
  2. HADOOP-9898

Set SO_KEEPALIVE on all our sockets

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Closed
    • Minor
    • Resolution: Fixed
    • 3.0.0-alpha1
    • 2.3.0
    • ipc, net
    • None

    Description

      We recently saw an issue where network issues between slaves and the NN caused ESTABLISHED TCP connections to pile up and leak on the NN side. It looks like the RST packets were getting dropped, which meant that the client thought the connections were closed, while they hung open forever on the server.

      Setting the SO_KEEPALIVE option on our sockets would prevent this kind of leak from going unchecked.

      Attachments

        1. hadoop-9898.txt
          2 kB
          Todd Lipcon

        Issue Links

          Activity

            People

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

              Dates

                Created:
                Updated:
                Resolved: