Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-21484

Race writing minvalid doc when becoming too stale and applying a batch

    XMLWordPrintableJSON

Details

    • Icon: Bug Bug
    • Resolution: Done
    • Icon: Major - P3 Major - P3
    • None
    • 2.0.0
    • Replication
    • None
    • Replication
    • ALL

    Description

      These two threads (bgsync+oplogreader/synctail) can both write to the minvalid document.

      We should have the helper error if the previous state isn't correct when writing the minvalid document when applying the batch. Then it should take the max of the minvalid.end and min of minvalid.start as the correct minvalid boundaries.

      In addition we should change oplogreader.connect to return a status so that bgsync can decide how to enter the too-stale state.

      Attachments

        Activity

          People

            backlog-server-repl Backlog - Replication Team
            scotthernandez Scott Hernandez (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: