[SPARK-51724][SS] RocksDB StateStore's lineage manager should be synchronized #50520
Conversation
This file contains 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
anishshri-db
reviewed
Apr 5, 2025
| lineage = lineage :+ item | ||
| } | ||
|
|
||
| def resetLineage(newLineage: Array[LineageItem]): Unit = { | ||
| def truncateFromVersion(versionToKeep: Long): Unit = synchronized { | ||
| resetLineage(getLineageForCurrVersion().filter(i => i.version >= versionToKeep)) |
There was a problem hiding this comment.
Are there any other places where these 2 operations were split ?
anishshri-db
reviewed
Apr 5, 2025
| @@ -1467,8 +1467,7 @@ class RocksDB( | |||
| // This is relative aggressive because that even if the uploading succeeds, | |||
| // it is not necessarily the one written to the commit log. But we can always load lineage | |||
| // from commit log so it is fine. | |||
| lineageManager.resetLineage(lineageManager.getLineageForCurrVersion() | |||
| .filter(i => i.version >= snapshot.version)) | |||
| lineageManager.truncateFromVersion(snapshot.version) | |||
There was a problem hiding this comment.
I guess its not easily possible. But any way to add a test to verify correctness across threads here ?
anishshri-db
approved these changes
Apr 5, 2025
HyukjinKwon
changed the title
|
It seems OK with handling the above comments as post review if these are addressable. Thanks! Merging to master/4.0. |
HeartSaVioR
pushed a commit
that referenced
this pull request
Apr 8, 2025
…hronized ### Why are the changes needed? RocksDB State Store's Lineage Manager currently isn't synchronized, but it can be accessed by both DB loading and maintenance thread. In theory, it can cause wrong lineage: 1. maintenance thread get current lineage 2. task commit() adds a lineage from the lienage 3. maintenance thread does the truncation and store it back In this case, the new lineage added by 2. is lost. It should be fixed by simply synchronizing those operations. ### Does this PR introduce _any_ user-facing change? No. ### How was this patch tested? Run existing unit tests ### Was this patch authored or co-authored using generative AI tooling? No. Closes #50520 from siying/lineagemanagerrace. Authored-by: Siying Dong <[email protected]> Signed-off-by: Jungtaek Lim <[email protected]> (cherry picked from commit 187adb8) Signed-off-by: Jungtaek Lim <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Why are the changes needed?
RocksDB State Store's Lineage Manager currently isn't synchronized, but it can be accessed by both DB loading and maintenance thread. In theory, it can cause wrong lineage:
In this case, the new lineage added by 2. is lost.
It should be fixed by simply synchronizing those operations.
Does this PR introduce any user-facing change?
No.
How was this patch tested?
Run existing unit tests
Was this patch authored or co-authored using generative AI tooling?
No.