hive-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Matt McCline (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HIVE-19275) Vectorization: Wrong Results / Execution Failures when Vectorization turned on in Spark
Date Mon, 23 Apr 2018 18:38:02 GMT

    [ https://issues.apache.org/jira/browse/HIVE-19275?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16448678#comment-16448678
] 

Matt McCline commented on HIVE-19275:
-------------------------------------

It would when HIVE-19269 goes in turning hive.vectorized.execution.enabled on.  However, this
change should hopefully prevent any EXPLAIN plan output changes.  And, leave the Wrong Results
/ Execution Failures to investigated later.

> Vectorization: Wrong Results / Execution Failures when Vectorization turned on in Spark
> ---------------------------------------------------------------------------------------
>
>                 Key: HIVE-19275
>                 URL: https://issues.apache.org/jira/browse/HIVE-19275
>             Project: Hive
>          Issue Type: Bug
>          Components: Hive
>            Reporter: Matt McCline
>            Assignee: Matt McCline
>            Priority: Critical
>             Fix For: 3.0.0, 3.1.0
>
>         Attachments: HIVE-19275.01.patch
>
>
> Quite a number of the bucket* tests had Wrong Results or Execution Failures.
> And others like semijoin, skewjoin, avro_decimal_native, mapjoin_addjar, mapjoin_decimal, nullgroup, decimal_join, mapjoin1.
> Some of the problems might be as simple as "-- SORT_QUERY_RESULTS" is missing.
> The bucket* problems looked more serious.
> This change sets "hive.vectorized.execution.enabled" to false at the top of those Q files.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message