[SPARK-42353][SS] Cleanup orphan sst and log files in RocksDB checkpoint directory #39897
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
HeartSaVioR
reviewed
Feb 9, 2023
Contributor
HeartSaVioR
left a comment
HeartSaVioR
left a comment
There was a problem hiding this comment.
Only nit. Thanks! Shall we remove WIP in the PR title as it seems to be ready to be reviewed and merged?
sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/state/RocksDBFileManager.scala
Outdated
Show resolved
Hide resolved
Contributor
Author
chaoqin-li1123
changed the title
|
+1. |
Contributor
|
Thanks! Merging to master. |
Kimahriman
pushed a commit
to Kimahriman/spark
that referenced
this pull request
Aug 8, 2023
…int directory ### What changes were proposed in this pull request? When RocksDB version.zip file get overwritten (e.g. concurrent task execution, task/stage/batch reattempts) or the zip file don't get uploaded successfully, the associated sst and log files don't get garbage collected. We can clean up these SST files during periodic state store maintenance. The major concern is that sst files for ongoing version also appear to be "orphan" because they are uploaded before zip file, we have to be careful not to delete them. To be conservative, we only delete orphan files older than all files tracked by current metadata when there are at least 2 versions. ### Why are the changes needed? When RocksDB version.zip file get overwritten (e.g. concurrent task execution, task/stage/batch reattempts) or the zip file don't get uploaded successfully, the associated sst and log files don't get garbage collected.([https://github.com/databricks/runtime/blob/master/sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/state/RocksDBFileManager.scala](https://github.com/databricks/runtime/blob/master/sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/state/RocksDBFileManager.scala#L305-L309)) These files consume storage. ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? Add unit test. Closes apache#39897 from chaoqin-li1123/orphan_cleanup. Authored-by: Chaoqin Li <chaoqin.li@databricks.com> Signed-off-by: Jungtaek Lim <kabhwan.opensource@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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?
When RocksDB version.zip file get overwritten (e.g. concurrent task execution, task/stage/batch reattempts) or the zip file don't get uploaded successfully, the associated sst and log files don't get garbage collected. We can clean up these SST files during periodic state store maintenance. The major concern is that sst files for ongoing version also appear to be "orphan" because they are uploaded before zip file, we have to be careful not to delete them. To be conservative, we only delete orphan files older than all files tracked by current metadata when there are at least 2 versions.
Why are the changes needed?
When RocksDB version.zip file get overwritten (e.g. concurrent task execution, task/stage/batch reattempts) or the zip file don't get uploaded successfully, the associated sst and log files don't get garbage collected.(https://github.com/databricks/runtime/blob/master/sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/state/RocksDBFileManager.scala) These files consume storage.
Does this PR introduce any user-facing change?
No
How was this patch tested?
Add unit test.