[
https://issues.apache.org/jira/browse/FLINK-30567?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jing Zhang updated FLINK-30567:
-------------------------------
Description:
If use hive dialect, the data would be wrong if we rerun the following job
multiple tables.
{code:java}
INSERT overwrite TABLE db_name.target_tB_name partition(p_date = '20230101')
select
author_id
from db_name.source_tb_name
WHERE p_date = '20230101'
and author_id <> '0'{code}
The framework has a minor bug when determining whether an override is required
if the target name contains uppercase character.
> Wrong insert overwrite behavior when the table contains uppercase character
> with Hive dialect
> ----------------------------------------------------------------------------------------------
>
> Key: FLINK-30567
> URL: https://issues.apache.org/jira/browse/FLINK-30567
> Project: Flink
> Issue Type: Sub-task
> Affects Versions: 1.16.0, 1.16.1
> Reporter: luoyuxia
> Assignee: luoyuxia
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.17
>
>
> If use hive dialect, the data would be wrong if we rerun the following job
> multiple tables.
> {code:java}
> INSERT overwrite TABLE db_name.target_tB_name partition(p_date = '20230101')
> select
> author_id
> from db_name.source_tb_name
> WHERE p_date = '20230101'
> and author_id <> '0'{code}
> The framework has a minor bug when determining whether an override is
> required if the target name contains uppercase character.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
