[
https://issues.apache.org/jira/browse/HIVE-17357?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16794046#comment-16794046
]
t oo commented on HIVE-17357:
-----------------------------
can you clarify this issue? does it mean before the fix there was no way for
HiveonSpark (with spark.master in non local mode) to use custom
serdes/udfs?[https://cwiki.apache.org//confluence/display/Hive/Hive+on+Spark:+Getting+Started]
does not mention how to register custom serde/udf jars/classes. For example if
i want to query a {{'com.uber.hoodie.hadoop.HoodieInputFormat' (this class
relies on parquet) table the docs dont say where to place the jar.}}
> Plugin jars are not properly added for LocalHiveSparkClient
> -----------------------------------------------------------
>
> Key: HIVE-17357
> URL: https://issues.apache.org/jira/browse/HIVE-17357
> Project: Hive
> Issue Type: Bug
> Components: Spark
> Affects Versions: 3.0.0
> Reporter: Aihua Xu
> Assignee: Aihua Xu
> Priority: Major
> Fix For: 3.0.0
>
> Attachments: HIVE-17357.1.patch
>
>
> I forgot to include the same change for LocalHiveSparkClient.java in
> HIVE-17336. We need to make the same change as HIVE-17336 in
> LocalHiveSparkClient class to include plugin jars. Maybe we should have a
> common base class for both LocalHiveSparkClient and RemoteHiveSparkClient to
> have some common functions.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
