[
https://issues.apache.org/jira/browse/SPARK-21435?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16089442#comment-16089442
]
Li Yuanjian commented on SPARK-21435:
-------------------------------------
[~sowen] I tested the patch in our scenario and add a UT, I think it can skip
empty files in Parquet.
> Empty files should be skipped while write to file
> -------------------------------------------------
>
> Key: SPARK-21435
> URL: https://issues.apache.org/jira/browse/SPARK-21435
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 2.2.0
> Reporter: Li Yuanjian
> Priority: Minor
>
> Consider of this scenario, source table has many partitions and data files,
> after the query filter, only a few data write to the destination dir.
> In this case the destination dir or table will have many empty files or files
> only have schema meta(parquet format), I know we can use coalesce but skip
> the empty file may be more better in this case.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
