steveloughran commented on code in PR #29000:
URL: https://github.com/apache/spark/pull/29000#discussion_r1053638820
##########
core/src/main/scala/org/apache/spark/internal/io/FileCommitProtocol.scala:
##########
@@ -169,4 +169,8 @@ object FileCommitProtocol extends Logging {
ctor.newInstance(jobId, outputPath)
}
}
+
+ def getStagingDir(path: String, jobId: String): Path = {
+ new Path(path, ".spark-staging-" + jobId)
+ }
Review Comment:
HadoopMapReduceCommitProtocol is using the RDD.id as the job ID here, not
uuid or hadoop job ID (which has had uniqueness issues elsewhere). if different
jobs on different spark clusters try to write to same table using dynamic
output partitioning, risk of output corruption -especially if hadoop "broken"
v2 commit algorithm is used.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
