dongjoon-hyun commented on a change in pull request #31499:
URL: https://github.com/apache/spark/pull/31499#discussion_r580765104
##########
File path:
sql/core/src/main/scala/org/apache/spark/sql/execution/command/ddl.scala
##########
@@ -791,8 +799,26 @@ case class AlterTableRecoverPartitionsCommand(
logDebug(s"Recovered ${parts.length} partitions ($done/$total so far)")
}
}
-}
+ // Drops the partitions that do not exist in the file system
+ private def dropPartitions(catalog: SessionCatalog, fs: FileSystem): Int = {
+ val dropPartSpecs = ThreadUtils.parmap(
+ catalog.listPartitions(tableName),
+ "AlterTableRecoverPartitionsCommand: non-existing partitions",
+ maxThreads = 8) { partition =>
+ partition.storage.locationUri.flatMap { uri =>
+ if (fs.exists(new Path(uri))) None else Some(partition.spec)
+ }
+ }.flatten
+ catalog.dropPartitions(
+ tableName,
+ dropPartSpecs,
+ ignoreIfNotExists = true,
+ purge = false,
+ retainData = true)
Review comment:
Could you add a comment about the reason why we use `retainData=true`? I
guess the reason is that `fs.exists(..)` is already `false` and we don't want
addition file system calls. Did I understand correctly?
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
