[jira] [Commented] (HIVE-8920) SplitSparkWorkResolver doesn't work with UnionWork [Spark Branch]

["Xuefu Zhang (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HIVE-8920?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14236924#comment-14236924 ] 

Xuefu Zhang commented on HIVE-8920:
-----------------------------------

With latest Spark branch, I got the following plan
{code}
hive> explain from (select * from dec2 union all select * from dec3) s 
> insert overwrite table dec4 select name, avg(value) group by name
> insert overwrite table dec5 select name, sum(value) group by name;
OK
STAGE DEPENDENCIES:
  Stage-2 is a root stage
  Stage-0 depends on stages: Stage-2
  Stage-3 depends on stages: Stage-0
  Stage-1 depends on stages: Stage-2
  Stage-4 depends on stages: Stage-1

STAGE PLANS:
  Stage: Stage-2
    Spark
      Edges:
        Union 2 <- Map 1 (NONE, 0), Map 4 (NONE, 0)
        Reducer 3 <- Union 2 (SORT, 1)
{code}
The plan seems correct to me. [~csun], what's the real issue here?

> SplitSparkWorkResolver doesn't work with UnionWork [Spark Branch]
> -----------------------------------------------------------------
>
>                 Key: HIVE-8920
>                 URL: https://issues.apache.org/jira/browse/HIVE-8920
>             Project: Hive
>          Issue Type: Sub-task
>          Components: Spark
>    Affects Versions: spark-branch
>            Reporter: Chao
>
> The following query will not work:
> {code}
> from (select * from table0 union all select * from table1) s
> insert overwrite table table3 select s.x, count(1) group by s.x
> insert overwrite table table4 select s.y, count(1) group by s.y;
> {code}
> Currently, the plan for this query, before SplitSparkWorkResolver, looks like below:
> {noformat}
>    M1    M2
>      \  / \
>       U3   R5
>       |
>       R4
> {noformat}
> In {{SplitSparkWorkResolver#splitBaseWork}}, it assumes that the {{childWork}} is a ReduceWork, but for this case, you can see that for M2 the childWork could be UnionWork U3. Thus, the code will fail.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)