[
https://issues.apache.org/jira/browse/DRILL-3219?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14612336#comment-14612336
]
Jinfeng Ni commented on DRILL-3219:
-----------------------------------
The root cause of the this problem is similar to DRILL-3412. Basically, we do
not have rules to push either Project or Filter through the Window operator.
That's was caused the timing of when Drill breaks a project with window
function into Project and Window operator ( currently, it just happened too
early in the planning logic).
> Filter is not pushed into subquery with window function
> --------------------------------------------------------
>
> Key: DRILL-3219
> URL: https://issues.apache.org/jira/browse/DRILL-3219
> Project: Apache Drill
> Issue Type: Bug
> Components: Query Planning & Optimization
> Affects Versions: 1.0.0
> Reporter: Victoria Markman
> Assignee: Jinfeng Ni
> Labels: window_function
> Fix For: 1.2.0
>
>
> {code}
> 0: jdbc:drill:schema=dfs> explain plan for select * from ( select a1, b1, c1,
> sum(a1) over(partition by b1) from t1 ) where c1 is not null;
> +------+------+
> | text | json |
> +------+------+
> | 00-00 Screen
> 00-01 Project(a1=[$0], b1=[$1], c1=[$2], EXPR$3=[$3])
> 00-02 Project(a1=[$0], b1=[$1], c1=[$2], EXPR$3=[CASE(>($3, 0),
> CAST($4):ANY, null)])
> 00-03 SelectionVectorRemover
> 00-04 Filter(condition=[IS NOT NULL($2)])
> 00-05 Window(window#0=[window(partition {1} order by [] range
> between UNBOUNDED PRECEDING and UNBOUNDED FOLLOWING aggs [COUNT($0),
> $SUM0($0)])])
> 00-06 SelectionVectorRemover
> 00-07 Sort(sort0=[$1], dir0=[ASC])
> 00-08 Project(a1=[$2], b1=[$1], c1=[$0])
> 00-09 Scan(groupscan=[ParquetGroupScan
> [entries=[ReadEntryWithPath [path=maprfs:///drill/testdata/subqueries/t1]],
> selectionRoot=/drill/testdata/subqueries/t1, numFiles=1, columns=[`a1`, `b1`,
> `c1`]]])
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
