Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-31679

Increase in disk i/o for writes to replica set

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Closed
    • Priority: Major - P3
    • Resolution: Fixed
    • Affects Version/s: 3.6.0-rc0
    • Fix Version/s: 3.6.6, 3.7.3
    • Component/s: Replication, Storage
    • Labels:
    • Backwards Compatibility:
      Fully Compatible
    • Operating System:
      ALL
    • Backport Requested:
      v3.6
    • Sprint:
      Storage 2018-02-12, Storage NYC 2018-03-12
    • Case:

      Description

      Simple insert loop running against 1-node replica set

          db.c.insert({_id: 0, i: 0})
          for (i = 0; i < 100000; i++)
              db.c.update({_id: 0}, {$inc: {i: 1}})
      

      Shows a large increase in journal syncs, disk write requests, bytes written. 3.4.9 left, 3.6.0-rc0 right:

      It appears we're doing a flush on every update, even though there's no write concern. In the simple repro above this doesn't seem to affect performance (for reasons tbd) but with more complex workloads the additional journal i/o will impact performance.

      Does not reproduce with standalone mongod.

        Attachments

        1. four.png
          73 kB
          Bruce Lucas
        2. io.png
          160 kB
          Bruce Lucas
        3. io.png
          188 kB
          Bruce Lucas

          Issue Links

            Activity

              People

              Assignee:
              geert.bosch Geert Bosch
              Reporter:
              bruce.lucas Bruce Lucas
              Participants:
              Votes:
              0 Vote for this issue
              Watchers:
              22 Start watching this issue

                Dates

                Created:
                Updated:
                Resolved: