Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-86819

jstest benchmark for $merge that can use Atlas target

    • Type: Icon: Task Task
    • Resolution: Fixed
    • Priority: Icon: Major - P3 Major - P3
    • 8.1.0-rc0
    • Affects Version/s: None
    • Component/s: None
    • Atlas Streams
    • Fully Compatible
    • Sprint 45, Sprint 48, Sprint 49

      There have two internal reports so far of slow $merge performance.

      1. Joe did some testing in https://mongodb.slack.com/archives/C04AH2TF7E1/p1708440672869589 of a pipeline that reads data from kafka $source into a mongoDB $merge.

      with an Atlas M30 unsharded target, the update rate we achieve is only 500 per second, for a 20kB doc... only 10MB per second of write throughput.

      2. Devika also saw slow performance for an M30->M30 [$source, $merge] pipeline https://mongodb.slack.com/archives/C065427D6F5/p1709912805521919?thread_ts=1709759727.899969&cid=C065427D6F5 

            Assignee:
            matthew.normyle@mongodb.com Matthew Normyle
            Reporter:
            matthew.normyle@mongodb.com Matthew Normyle
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Created:
              Updated:
              Resolved: