You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hive.apache.org by "Zhizhen Hou (Jira)" <ji...@apache.org> on 2022/09/14 06:50:00 UTC
[jira] [Commented] (HIVE-26499) 启用向量化（hive.vectorized.execution.enabled=true）后，case when的计算结果出现预设外的值
[ https://issues.apache.org/jira/browse/HIVE-26499?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17603908#comment-17603908 ]
Zhizhen Hou commented on HIVE-26499:
------------------------------------
Master branch has solved this problem. https://issues.apache.org/jira/browse/HIVE-26408
> 启用向量化（hive.vectorized.execution.enabled=true）后，case when的计算结果出现预设外的值
> --------------------------------------------------------------------
>
> Key: HIVE-26499
> URL: https://issues.apache.org/jira/browse/HIVE-26499
> Project: Hive
> Issue Type: Bug
> Components: Hive
> Affects Versions: 3.1.0
> Environment: hdfs(3.1.1)
> yarn(3.1.1)
> zookeeper(3.4.6)
> hive(3.1.0)
> tez(0.9.1)
> Reporter: Ricco-Chan
> Priority: Major
> Attachments: image-2022-08-29-11-04-21-921.png
>
>
> -- case when预设值只有1、2、3，计算结果中出现5和6。发现此bug时，对应表使用的是parquet格式 + snappy 压缩
>
> select distinct(traveller_type) from
> (
> select pri_acct_no,
> case
> when (t1.consume_flag = '1' and substr(t1.areacode, 1, 2) <> '65') then '2'
> when (substr(t1.areacode, 1, 2) = substr(t1.country_id_new, 1, 2) and t1.consume_flag = '1') then '1'
> else '3'
> end as traveller_type
> from my_table t1 where consume_flag = '1'
> ) t2;
> -----------------------------------------------------------------
> !image-2022-08-29-11-04-21-921.png!
--
This message was sent by Atlassian Jira
(v8.20.10#820010)