[
https://issues.apache.org/jira/browse/SPARK-35279?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17339227#comment-17339227
]
Apache Spark commented on SPARK-35279:
--------------------------------------
User 'SaurabhChawla100' has created a pull request for this issue:
https://github.com/apache/spark/pull/32437
> _SUCCESS file not written when using partitionOverwriteMode=dynamic
> -------------------------------------------------------------------
>
> Key: SPARK-35279
> URL: https://issues.apache.org/jira/browse/SPARK-35279
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Affects Versions: 3.1.1
> Reporter: Goran Kljajic
> Priority: Minor
>
> Steps to reproduce:
>
> {code:java}
> case class A(a: String, b:String)
> def df = List(A("a", "b")).toDF
> spark.conf.set("spark.sql.sources.partitionOverwriteMode", "dynamic")
> val writer = df.write.mode(SaveMode.Overwrite).partitionBy("a")
> writer.parquet("s3a://some_bucket/test/")
> {code}
> when spark.sql.sources.partitionOverwriteMode is set to dynamic, the output
> written doesn't have _SUCCESS file updated.
> (I have checked different versions of hadoop from 3.1.4 to 3.22 and they all
> behave the same, so the issue is with spark)
> This is working in spark 3.0.2
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
