Uploaded image for project: 'Hadoop HDFS'
  1. Hadoop HDFS
  2. HDFS-5037

Active NN should trigger its own edit log rolls

    XMLWordPrintableJSON

Details

    • Improvement
    • Status: Closed
    • Critical
    • Resolution: Fixed
    • 2.1.0-beta, 3.0.0-alpha1
    • 2.3.0
    • ha, namenode
    • None

    Description

      We've seen cases where the SBN/2NN went down, and then users accumulated very very large edit log segments. This causes a slow startup time because the last edit log segment must be read fully to recover it before the NN can start up again. Additionally, in the case of QJM, it can trigger timeouts on recovery or edit log syncing because the very-large segment has to get processed within a certain time bound.

      We could easily improve this by having the NN trigger its own edit log rolls on a configurable size (eg every 256MB)

      Attachments

        1. hdfs-5037-1.patch
          8 kB
          Andrew Wang
        2. hdfs-5037-2.patch
          12 kB
          Andrew Wang
        3. hdfs-5037-3.patch
          13 kB
          Andrew Wang
        4. hdfs-5037-4.patch
          14 kB
          Andrew Wang
        5. hdfs-5037-5.patch
          14 kB
          Andrew Wang

        Activity

          People

            andrew.wang Andrew Wang
            tlipcon Todd Lipcon
            Votes:
            1 Vote for this issue
            Watchers:
            12 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: