[SPARK-53791][CORE][SQL] Make the rename operations multi-threaded. #52507
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Xtpacz
changed the title
Xtpacz
force-pushed
the
multi-threaded-rename
branch
from
1fd74df to
14614f9
Compare
|
cc @huaxingao @cloud-fan could you review this PR? This is my first PR in the Spark community. Thank you ：） |
cloud-fan
reviewed
Oct 9, 2025
| .booleanConf | ||
| .createWithDefault(false) | ||
|
|
||
| private[spark] val FILES_RENAME_NUM_THREADS = ConfigBuilder("spark.files.rename.numThreads") |
There was a problem hiding this comment.
can this be a dynamic session config in SQLConf?
There was a problem hiding this comment.
can this be a dynamic session config in
SQLConf?
My understanding is to add a session-scoped SQLConf (e.g., spark.sql.files.rename.numThreads) and have SQL paths read it first, while retaining the global key (spark.files.rename.numThreads) as a fallback since core (e.g., HadoopMapReduceCommitProtocol) cannot depend on SQL. Please let me know if I’ve misunderstood. Thank you!
cloud-fan
reviewed
Oct 9, 2025
| case Some(sc) => sc.conf.get(FILES_RENAME_NUM_THREADS) | ||
| case None => FILES_RENAME_NUM_THREADS.defaultValue.get | ||
| } | ||
| val pool = ThreadUtils.newDaemonFixedThreadPool(numThreads, "file-rename") |
There was a problem hiding this comment.
shall we have a global long-standing thread pool to do this work?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What changes were proposed in this pull request?
Improved two classes:
core/src/main/scala/org/apache/spark/internal/io/HadoopMapReduceCommitProtocol.scala.sql/hive/src/main/scala/org/apache/spark/sql/hive/execution/InsertIntoHiveDirCommand.scala.Added a configuration item:
spark.files.rename.numThreadsincore/src/main/scala/org/apache/spark/internal/config/package.scala.Why are the changes needed?
For example, during the
insert overwrite directoryoperation, each rename operation triggers an RPC request. Therefore, when there are too many files, it can be time-consuming.Converting the serial rename operations to multi-threaded operations can save job execution time.
Does this PR introduce any user-facing change?
Yes. A spark configuration item has been added:
spark.files.rename.numThreads.How was this patch tested?
Verified that the affected tests are passing successfully.
Was this patch authored or co-authored using generative AI tooling?
No.