Uploaded image for project: 'Cassandra'
  1. Cassandra
  2. CASSANDRA-19497

ResultRetriever should batch clusterings/rows during SAI post-filtering reads

    XMLWordPrintableJSON

Details

    • Improvement
    • Status: In Progress
    • Normal
    • Resolution: Unresolved
    • 5.0.x, 5.x
    • Feature/SAI
    • None

    Description

      SAI currently creates and executes a SinglePartitionReadCommand for every PrimaryKey the index produces to read the corresponding row for post-filtering. Informed by the limits present in the read command itself, it should be possible to batch those reads w/ a ClusteringIndexNamesFilter in many fewer SinglePartitionReadCommands. When we have a handful of matches in a large partition, this seems like would involve many fewer seeks, less unnecessary object creation, etc.

      Attachments

        1. heap-flamegraph.html
          77 kB
          Caleb Rackliffe
        2. wall-no-parked-threads.html
          136 kB
          Caleb Rackliffe

        Activity

          People

            arkn98 Arun Ganesh
            maedhroz Caleb Rackliffe
            Arun Ganesh
            Caleb Rackliffe
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

              Created:
              Updated: