[GitHub] [spark] dhruve commented on a change in pull request #21048: [SPARK-23966][SS] Refactoring all checkpoint file writing logic in a common CheckpointFileManager interface

[GitBox <gi...@apache.org>]

dhruve commented on a change in pull request #21048:
URL: https://github.com/apache/spark/pull/21048#discussion_r617654509



##########
File path: sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/CheckpointFileManager.scala
##########
@@ -303,9 +303,11 @@ class FileContextBasedCheckpointFileManager(path: Path, hadoopConf: Configuratio
     fc.mkdir(path, FsPermission.getDirDefault, true)
   }
 
-  override def create(path: Path): FSDataOutputStream = {
+  override def createTempFile(path: Path): FSDataOutputStream = {
     import CreateFlag._
-    fc.create(path, EnumSet.of(CREATE, OVERWRITE))
+    import Options._
+    fc.create(
+      path, EnumSet.of(CREATE, OVERWRITE), CreateOpts.checksumParam(ChecksumOpt.createDisabled()))

Review comment:
       Thanks for getting back on this. We hit a nasty bug in hadoop with files with Null checksum which wouldn't allow the NodeManager to start and wanted to disable this.




-- 
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:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org