This repository was archived by the owner on Jul 1, 2025. It is now read-only.
[WIP] Work stealing in Nemo (alternative implementation) #321
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
…-nemo into follow_hurricane
|
Kudos, SonarCloud Quality Gate passed!
|
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
1 participant
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Outdated suggestions cannot be applied.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.
JIRA: NEMO-466 Dynamic Work Stealing on Nemo for handling skews
Major changes:
WorkStealingStatePassandWorkStealingSubSplitPassto enable annotations related to work stealing.WorkStealingStatePassannotates three phases of work stealing process:DEFAULT,SPLIT, andMERGE. These are tagged to the vertex based on the name assigned to beam transform by the user.SPLITphase and merged inMERGEphase.WorkStealingSubSplitPassannotates the number to split the tasks inSPLITphase.WorkStealingPolicy, which executesWorkStealingPassandWorkStealingSubSplitPassadditionally toDefaultCompositePass.task idandblock idto differentiate between work stealing tasks and default tasks.task id[stage id] - [task index] - [task attempt][stage id] - [task index] - [* or sub index for work stealing tasks] - [task attempt]block id[runtime edge id] - [producer task index] - [producer task attempt][runtime edge id] - [producer task index] - [producer task sub index] - [producer task attempt]for blocks generated by work stealing tasks.Minor changes to note:
PipelineTranslatorto read the user tags given to the transform.PlanStateManagerandBlockInputReaderto manage work stealing tasks and non-split tasks together.Tests for the changes:
WorkStealingPolicy. Note that in order to execute WorkStealing policy, the application pipeline should provide appropriate splitting and merging strategy.WorkStealingCompositePasswhich encapsulatesWorkStealingStatePassandWorkStealingSubSplitPassParentTaskDataFetcherTestandTaskExecutorTest, reflecting the changes inBlockInputReaderOther comments:
Closes #321