Uploaded image for project: 'Hadoop Common'
  1. Hadoop Common
  2. HADOOP-2188

RPC should send a ping rather than use client timeouts

    XMLWordPrintableJSON

Details

    • Improvement
    • Status: Closed
    • Major
    • Resolution: Fixed
    • 0.16.1
    • 0.18.0
    • ipc
    • None
    • Incompatible change, Reviewed
    • Hide
      Replaced timeouts with pings to check that client connection is alive. Removed the property ipc.client.timeout from the default Hadoop configuration. Removed the metric RpcOpsDiscardedOPsNum.
      Show
      Replaced timeouts with pings to check that client connection is alive. Removed the property ipc.client.timeout from the default Hadoop configuration. Removed the metric RpcOpsDiscardedOPsNum.

    Description

      Current RPC (really IPC) relies on client side timeouts to find "dead" sockets. I propose that we have a thread that once a minute (if the connection has been idle) writes a "ping" message to the socket. The client can detect a dead socket by the resulting error on the write, so no client side timeout is required. Also note that the ipc server does not need to respond to the ping, just discard it.

      Attachments

        1. ipc-timeout.patch
          34 kB
          Hairong Kuang
        2. ipc-timeout1.patch
          48 kB
          Hairong Kuang
        3. ipc-timeout2.patch
          47 kB
          Hairong Kuang
        4. ipc-timeout3.patch
          48 kB
          Hairong Kuang
        5. ipc-timeout4.patch
          49 kB
          Hairong Kuang
        6. ipc-timeout5.patch
          50 kB
          Hairong Kuang
        7. ipc-timeout6.patch
          51 kB
          Hairong Kuang
        8. ipc-timeout7.patch
          53 kB
          Hairong Kuang
        9. ipc-timeout8.patch
          54 kB
          Hairong Kuang
        10. ipc-timeout9.patch
          54 kB
          Hairong Kuang
        11. rpc-to.patch
          27 kB
          Owen O'Malley

        Issue Links

          is part of

          Improvement - An improvement or enhancement to an existing feature or task. HADOOP-2864 Improve the Scalability and Robustness of IPC

          • Major - Major loss of function.
          • Resolved

          Activity

            People

              hairong Hairong Kuang
              omalley Owen O'Malley
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: