Uploaded image for project: 'Drivers'
  1. Drivers
  2. DRIVERS-209

Change minHeartbeatFrequencyMS from 10 to 500

    XMLWordPrintableJSON

Details

    • Icon: Improvement Improvement
    • Resolution: Done
    • Icon: Major - P3 Major - P3
    • None
    • None
    • None
    • $i18n.getText("admin.common.words.hide")
      Key Status/Resolution FixVersion
      NODE-392 Done 2.0.23
      CXX-566 Won't Fix
      PYTHON-866 Done 3.0
      PERL-514 Done
      RUBY-879 Done 12_01_17
      CSHARP-1196 Done 2.0
      PHPC-184 Done 0.4.0
      CDRIVER-585 Done 1.2-beta0
      PHP-1415 Won't Fix
      CXX-568 Done
      $i18n.getText("admin.common.words.show")
      #scriptField, #scriptField *{ border: 1px solid black; } #scriptField{ border-collapse: collapse; } #scriptField td { text-align: center; /* Center-align text in table cells */ } #scriptField td.key { text-align: left; /* Left-align text in the Key column */ } #scriptField a { text-decoration: none; /* Remove underlines from links */ border: none; /* Remove border from links */ } /* Add green background color to cells with FixVersion */ #scriptField td.hasFixVersion { background-color: #00FF00; /* Green color code */ } /* Center-align the first row headers */ #scriptField th { text-align: center; } Key Status/Resolution FixVersion NODE-392 Done 2.0.23 CXX-566 Won't Fix PYTHON-866 Done 3.0 PERL-514 Done RUBY-879 Done 12_01_17 CSHARP-1196 Done 2.0 PHPC-184 Done 0.4.0 CDRIVER-585 Done 1.2-beta0 PHP-1415 Won't Fix CXX-568 Done

    Description

      The Server Discovery And Monitoring Spec has been updated: minHeartbeatFrequencyMS has changed from 10 ms to 500 ms. It is still not configurable (unless you're the Java Driver and you had a knob before the spec was written).

      Based on recent experiences with replica set reconfig in MMS, and experiments performed with the Java Driver and MongoDB 3.0, we've decided that 500 ms is a safer value than 10 ms.

      If a secondary is underpowered and collocated with other processes, then a huge number of clients calling ismaster every 10 ms might be able to make it less responsive, or delay the election, by overwhelming it while awaiting an election. The 500 ms number prevents a storm without unduly delaying the clients' notification when the election completes.

      https://github.com/mongodb/specifications/commit/32901d

      Future work may change this number again, or mandate exponential backoff, or rearchitect the method by which servers notify clients of topology changes.

      Attachments

        Activity

          People

            barrie Barrie Segal
            jesse@mongodb.com A. Jesse Jiryu Davis
            Votes:
            0 Vote for this issue
            Watchers:
            4 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: