[jira] [Commented] (DRILL-4321) Difference in results count distinct with min max query on JDK8

["Khurram Faraaz (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/DRILL-4321?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15130178#comment-15130178 ] 

Khurram Faraaz commented on DRILL-4321:
---------------------------------------

I can repro on mapr Drill 1.5.0 git.commit.id=6a36a704 and JDK8 and MapR FS 5.0.0 on 4 node cluster on CentOS

> Difference in results count distinct with min max query on JDK8
> ---------------------------------------------------------------
>
>                 Key: DRILL-4321
>                 URL: https://issues.apache.org/jira/browse/DRILL-4321
>             Project: Apache Drill
>          Issue Type: Bug
>          Components: Query Planning & Optimization
>    Affects Versions: 1.4.0
>         Environment: 4 node cluster
>            Reporter: Khurram Faraaz
>         Attachments: expected_results.res
>
>
> count distinct query with min max and group by and order by returns incorrect results on MapR Drill 1.4.0, MapR FS 5.0.0 GA and JDK8
> The difference is in the way we round off values after the decimal when using JDK8.
> Expected results file can be found here
> https://github.com/mapr/drill-test-framework/blob/master/framework/resources/Functional/aggregates/aggregation/count_distinct/with_min_max_c_float_group_by_1_cols.res
> Failing query is Functional/aggregates/aggregation/count_distinct/
> with_min_max_c_float_group_by_1_cols.sql
> {noformat}
> select count(distinct c_float), max(c_float), min(c_float) from alltypes_with_nulls group by  c_date order by  c_date;
> {noformat}



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