[
https://issues.apache.org/jira/browse/IMPALA-10406?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17255378#comment-17255378
]
ASF subversion and git services commented on IMPALA-10406:
----------------------------------------------------------
Commit 5baadd1da7d554ea3446e2a025afe8e991339765 in impala's branch
refs/heads/master from xqhe
[ https://gitbox.apache.org/repos/asf?p=impala.git;h=5baadd1 ]
IMPALA-10406: Query with analytic functions doesn't need to materialize the
predicates bounded to kudu
Before when query with analytic functions will materialize the
unassigned conjuncts.
But for the predicates that can be evaluated by kudu don't need to
materialize.
This optimization can reduce the amount of data to exchange and sort.
Testing:
- Add planner test in analytic-fns.test
Change-Id: Iba8371eff6ae1bcffd51b44843175c52f2127e46
Reviewed-on: http://gerrit.cloudera.org:8080/16905
Tested-by: Impala Public Jenkins <[email protected]>
Reviewed-by: Tim Armstrong <[email protected]>
> Query with analytic function doesn't need to materialize the predicate pushed
> down to kudu
> ------------------------------------------------------------------------------------------
>
> Key: IMPALA-10406
> URL: https://issues.apache.org/jira/browse/IMPALA-10406
> Project: IMPALA
> Issue Type: Improvement
> Components: Frontend
> Affects Versions: Impala 4.0
> Reporter: Xianqing He
> Assignee: Xianqing He
> Priority: Minor
> Fix For: Impala 4.0
>
>
> The query with analytic function doesn't need to materialize the predicate
> pushed down to kudu.
> E.g.
> {code:java}
> select min(n_nationkey) over(partition by n_regionkey) from tpch_kudu.nation
> t1 where t1.n_name in ('ALGERIA', 'ARGENTINA');
> {code}
> The plan
> {code:java}
> PLAN-ROOT SINK
> |
> 02:ANALYTIC
> | functions: min(n_nationkey)
> | partition by: n_regionkey
> | row-size=25B cardinality=2
> |
> 01:SORT
> | order by: n_regionkey ASC NULLS LAST
> | row-size=23B cardinality=2
> |
> 00:SCAN KUDU [tpch_kudu.nation t1]
> kudu predicates: t1.n_name IN ('ALGERIA', 'ARGENTINA')
> row-size=27B cardinality=2
> {code}
> We don't need to materialize the slot 'n_name'.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
