Uploaded image for project: 'Hadoop YARN'
  1. Hadoop YARN
  2. YARN-6889

The throughput of timeline server is too small

    XMLWordPrintableJSON

Details

    • Improvement
    • Status: Resolved
    • Critical
    • Resolution: Duplicate
    • None
    • None
    • timelineserver
    • None

    Description

      Recent large-scale pressure test single timeline server, and I found the throughput of timeline server is too small.
      I setup multiple servers, and each of them setup multiple processes. Each process was setup up multiple threads, I send different data size with each thread.

      Although I use different pressures and different scenarios. But the timeline server processing power is basically similar, and
      the ability to process messages is about 70 per second.
      It can't meet our requirements, we should improve it.

      Attachments

        Activity

          People

            Unassigned Unassigned
            daemon YunFan Zhou
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: