Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-55215

Handle small measurement counts in buckets for ARHASH

    XMLWordPrintable

Details

    • Task
    • Status: Closed
    • Major - P3
    • Resolution: Fixed
    • None
    • 5.0.0-rc0
    • Querying
    • Fully Compatible
    • Query Execution 2021-03-22, Query Execution 2021-04-05, Query Execution 2021-04-19
    • 28

    Description

      When testing SERVER-54221 we realized that if the gTimeseriesBucketMaxCount is significantly larger than the actual bucket counts we will exhaust the kMaxAttempt to get a non-duplicate document and fail the query.

      This behavior is not ideal. We should implement a fall-back mechanism when this is the case. Two ideas discussed is to figure out if we can cheaply compute the maximum bucket count, or use a trial stage to see if we are able to sample the collection, and if not fallback to topk sorting.

      Attachments

        Activity

          People

            david.storch@mongodb.com David Storch
            eric.cox@mongodb.com Eric Cox (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: