Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-70241

[CQF] Implement "auto mode" for cardinality estimation

    • Type: Icon: Improvement Improvement
    • Resolution: Unresolved
    • Priority: Icon: Major - P3 Major - P3
    • None
    • Affects Version/s: None
    • Component/s: None
    • Query Optimization

      This is regarding the flag "kInternalCascadesOptimizerDisableScanDefault".

       

      Currently we have 3 options: heuristic, sampling, and histograms. The sampling is currently ignored for small sizes while the histogram is not. The reasoning behind is that it may be too slow given the relative collection size.

       

      The goal of this ticket is to:

      1. Add "auto" CE estimator mode which will select the best estimator (contingent on current efforts to implement estimation based on histograms)
      2. Allow forcing of estimation strategies (independent on the metadata).

            Assignee:
            backlog-query-optimization [DO NOT USE] Backlog - Query Optimization
            Reporter:
            svilen.mihaylov@mongodb.com Svilen Mihaylov (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Created:
              Updated: