[
https://issues.apache.org/jira/browse/SPARK-27595?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16830165#comment-16830165
]
Liang-Chi Hsieh commented on SPARK-27595:
-----------------------------------------
Is turning off {{spark.sql.sources.partitionColumnTypeInference.enabled}}
helpful?
> Spark couldn't read partitioned(string type) Orc column correctly if the
> value contains Float/Double value
> ----------------------------------------------------------------------------------------------------------
>
> Key: SPARK-27595
> URL: https://issues.apache.org/jira/browse/SPARK-27595
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Affects Versions: 1.6.0
> Reporter: Ameer Basha Pattan
> Priority: Critical
>
> create external table unique_keys (
> key string
> ,locator_id string
> , create_date string
> , sequence int
> )
> partitioned by (source string)
> stored as orc location '/user/hive/warehouse/reservation.db/unique_keys';
> /user/hive/warehouse/reservation.db/unique_keys contains data like below:
> /user/hive/warehouse/reservation.db/unique_keys/source=6S
> /user/hive/warehouse/reservation.db/unique_keys/source=7F
> /user/hive/warehouse/reservation.db/unique_keys/source=7H
> /user/hive/warehouse/reservation.db/unique_keys/source=8D
>
> If I try to read orc files through Spark,
> val masterDF =
> hiveContext.read.orc("/user/hive/warehouse/reservation.db/unique_keys")
> source value getting changed to *7.0 and 8.0* for 7F and 8D respectively.
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
