[jira] [Updated] (CALCITE-4244) Minor improvements in EnumerableBatchNestedLoopJoin cost computation

["Ruben Q L (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/CALCITE-4244?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Ruben Q L updated CALCITE-4244:
-------------------------------
    Description: 
{{EnumerableBatchNestedLoopJoin#computeSelfCost}} requires some minor improvements:
- computeSelfCost: take care of "TODO" to handle the cost of the final loop that looks for the match between outer and inner result (currently ignored).
- estimateRowCount: override this method to have a more realistic row count (currently it inherits the computation from Join, which does not take into account the fact that the RHS of BatchNestedLoop gets re-evaluated)

  was:{{EnumerableBatchNestedLoopJoin#computeSelfCost}} requires some minor improvements, especially the "TODO" to handle the cost of the final loop that looks for the match between outer and inner result (currently ignored).


> Minor improvements in EnumerableBatchNestedLoopJoin cost computation
> --------------------------------------------------------------------
>
>                 Key: CALCITE-4244
>                 URL: https://issues.apache.org/jira/browse/CALCITE-4244
>             Project: Calcite
>          Issue Type: Improvement
>          Components: core
>            Reporter: Ruben Q L
>            Assignee: Ruben Q L
>            Priority: Minor
>
> {{EnumerableBatchNestedLoopJoin#computeSelfCost}} requires some minor improvements:
> - computeSelfCost: take care of "TODO" to handle the cost of the final loop that looks for the match between outer and inner result (currently ignored).
> - estimateRowCount: override this method to have a more realistic row count (currently it inherits the computation from Join, which does not take into account the fact that the RHS of BatchNestedLoop gets re-evaluated)



--
This message was sent by Atlassian Jira
(v8.3.4#803005)