[
https://issues.apache.org/jira/browse/HIVE-13251?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15189444#comment-15189444
]
Aihua Xu commented on HIVE-13251:
---------------------------------
Attached the patch-1: when we read the file achema, we will try to read and
convert the string to int for decimal scale, but we disallow setting precision
and scale to string when hive creates the avro file. Added the positive and
negative unit tests.
> hive can't read the decimal in AVRO file generated from previous version
> ------------------------------------------------------------------------
>
> Key: HIVE-13251
> URL: https://issues.apache.org/jira/browse/HIVE-13251
> Project: Hive
> Issue Type: Bug
> Components: Hive
> Affects Versions: 2.1.0
> Reporter: Aihua Xu
> Assignee: Aihua Xu
> Attachments: HIVE-13251.1.patch
>
>
> HIVE-7174 makes the avro schema change to match avro definition, while it
> breaks the compatibility if the file is generated from the previous Hive
> although the file schema from the file for such decimal is not correct based
> on avro definition. We should allow to read old file format "precision" :
> "4", "scale": "8", but when we write, we should write in the new format.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
