Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-21765

Sharding balancer should check for shutdown more frequently

    XMLWordPrintableJSON

Details

    • Icon: Bug Bug
    • Resolution: Done
    • Icon: Major - P3 Major - P3
    • None
    • None
    • Sharding
    • None
    • Sharding
    • ALL

    Description

      The mongos clean shutdown sequence blocks waiting for all currently held distributed locks to be freed in order to proceed with the shutdown. This means that if the balancer takes long time to complete the round it would block shutdown.

      For this reason the balancer should be checking whether shutdown has been requested after each chunk has been migrated instead of once for the entire round.

      Attachments

        Activity

          People

            backlog-server-sharding [DO NOT USE] Backlog - Sharding Team
            kaloian.manassiev@mongodb.com Kaloian Manassiev
            Votes:
            0 Vote for this issue
            Watchers:
            6 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: