Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-32124

Have WTCheckpointThread do raw checkpoints

    XMLWordPrintableJSON

Details

    • Icon: Improvement Improvement
    • Resolution: Fixed
    • Icon: Major - P3 Major - P3
    • 3.6.0-rc8, 3.7.1
    • None
    • Storage
    • None
    • Fully Compatible
    • v3.6
    • Storage 2017-12-04

    Description

      Calling waitUntilDurable has logic for concurrent callers to cooperate. This cooperation happens for concurrent callers to checkpoints as well as log flushes. Unfortunately a slow checkpoint can cause a faster log flush to block on the "cooperation" mutex. This can stall reads and writes, in particular, secondary reads of the oplog in an insert heavy workload.

      Attachments

        Activity

          People

            daniel.gottlieb@mongodb.com Daniel Gottlieb (Inactive)
            daniel.gottlieb@mongodb.com Daniel Gottlieb (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            7 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: