Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-37880

Add in backoff for retrying sending commit and abort messages

    XMLWordPrintable

    Details

    • Backwards Compatibility:
      Fully Compatible
    • Operating System:
      ALL
    • Sprint:
      Sharding 2019-01-14, Sharding 2019-01-28
    • Linked BF Score:
      0

      Description

      Right now, if a commit, abort, or prepare message to a transaction participant fails, we retry immediately. It might be worth adding some sort of backoff so that we slow down retries if they keep failing. Same goes for targeting a shard in the coordinator code.

        Attachments

          Issue Links

          is duplicated by

          Bug - A problem which impairs or prevents the functions of the product. SERVER-38798 shard registry busy-loops in commitTransaction

          • Major - P3 - Major loss of function.
          • Closed

          Task - A task that needs to be done. SERVER-38444 Coordinating a transaction should run fully in the ThreadPoolTaskExecutor for sharding tasks

          • Major - P3 - Major loss of function.
          • Closed

          Task - A task that needs to be done. SERVER-38795 Add test that if coordinator fails to target host in participant shard, coordinator schedules a retry attempt

          • Major - P3 - Major loss of function.
          • Closed

          Task - A task that needs to be done. SERVER-38881 Use ThreadClient rather than onCreateThread to create Clients for work done by TransactionCoordinatorService

          • Major - P3 - Major loss of function.
          • Closed

            Activity

              People

              Assignee:
              kaloian.manassiev Kaloian Manassiev
              Reporter:
              matthew.saltz Matthew Saltz
              Participants:
              Votes:
              0 Vote for this issue
              Watchers:
              4 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved: