[GLUTEN-12298][FLINK] Support nexmark source multi-parallelism via ParallelSplit#12304
Open
ggjh-159 wants to merge 5 commits into
Open
[GLUTEN-12298][FLINK] Support nexmark source multi-parallelism via ParallelSplit#12304ggjh-159 wants to merge 5 commits into
ggjh-159 wants to merge 5 commits into
Conversation
lgbo-ustc
approved these changes
Jun 17, 2026
Contributor
|
Is there any test to cover this case? |
Contributor
|
you need to update github workflow config to refer the right velox4j branch/commit |
lgbo-ustc
reviewed
Jun 17, 2026
Comment on lines
+92
to
+94
| subtaskSplits.add( | ||
| new NexmarkConnectorSplit( | ||
| "connector-nexmark", toVeloxGeneratorConfig(generatorConfig), null)); |
Contributor
There was a problem hiding this comment.
for safty， subtask splits shoud use a non-ParallelSplit class
Author
There was a problem hiding this comment.
Addressed. Split into two types:
NexmarkConnectorSplit— leaf split,extends ConnectorSplit, only carries
NexmarkGeneratorConfig config. Serialized across the Java/C++ boundary as before.NexmarkParallelSplit— container,extends ParallelSplit, holds
List<NexmarkConnectorSplit> subtaskSplits. Only this type implements
getSubtaskSplit, so a leaf split can no longer be mistaken for a parallel
container.
NexmarkSourceFactory now builds a NexmarkParallelSplit wrapping the
per-subtask leaf splits, and the instanceof ParallelSplit check in
GlutenSourceFunction cleanly distinguishes container vs leaf.
Author
|
cc @lgbo-ustc UTs were added and the velox4j branch was set. |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 are proposed in this pull request?
fix: #12298
depends: bigo-sg/velox4j#36、bigo-sg/velox#45
This PR consumes the
ParallelSplitabstraction introduced by the companion velox4j PR:NexmarkSourceFactory.buildVeloxSourceiterates over all per-subtask splits fromgetSplits(parallelism)and packs them into a singleNexmarkConnectorSplitwhosesubtaskSplitscarries one entry per subtask.GlutenSourceFunction.initSessiondetectsParallelSplitwithparallelism > 1and selects the per-subtask split viagetSubtaskSplit(subtaskIndex, totalParallelism); otherwise it behaves as before.How was this patch tested?
Manual run on a standalone Flink cluster with
parallelism.default = 2, nexmarkevents.num = 10000, tps = 2000, query q0.dateTimespanfirstEventId=1, maxEvents=5000; subtask 1firstEventId=5001, maxEvents=5000