Joe McDonnell created IMPALA-13451:
--------------------------------------

             Summary: Calcite planner does not take advantage of the hash 
partition of previous aggregates
                 Key: IMPALA-13451
                 URL: https://issues.apache.org/jira/browse/IMPALA-13451
             Project: IMPALA
          Issue Type: Sub-task
          Components: Frontend
    Affects Versions: Impala 4.5.0
            Reporter: Joe McDonnell


In [IMPALA-13197|http://issues.apache.org/jira/browse/IMPALA-13197], we added 
support for analytic functions to the Calcite planner. The implementation 
passes empty lists to the second two parameters of AnalyticPlanner. That 
disables a small optimization which can avoid exchanges when the input is 
already hash partitioned in a compatible way:

[https://github.com/apache/impala/commit/0b3124ab35402f1ab8141e8ffcca28a5a481c81e]

This is a tracking Jira just so we don't forget about this difference.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to