MSQ: Use full parallelism in localSort. #18765
Open
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
Since localSort was introduced for sort-merge join in apache#13506, it has used 1 processor at a time with a max of 2 channels per processor. This is inefficient, because at the time the sorter runs, it is the only sorter running. The patch contains a comment describing the situation in more detail. This patch has two benefits. First, the sorter should run faster if multiple processing threads are available. Second, the sorter, due to having a larger max channels per merger, will make more efficient use of intermediate channels.
github-actions
bot
added
Area - Batch Ingestion
Area - MSQ
cryptoe
reviewed
Nov 24, 2025
| // concurrently with the sorter, but in that case it will only have one output channel so won't | ||
| // have many frames buffered. | ||
| memoryParameters.getSuperSorterConcurrentProcessors(), | ||
| memoryParameters.getSuperSorterMaxChannelsPerMerger(), |
Contributor
There was a problem hiding this comment.
There is an implicit assumption that getChannelsPerMerger >=2.
I am not sure if that assumption is honored in the memory parameters anywhere.
Contributor
Author
There was a problem hiding this comment.
Yes, it'll always be at least 2.
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Outdated suggestions cannot be applied.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.
Since localSort was introduced for sort-merge join in #13506, it has used 1 processor at a time with a max of 2 channels per processor. This is inefficient, because at the time the sorter runs, it is the only sorter running. The patch contains a comment describing the situation in more detail.
This patch has two benefits. First, the sorter should run faster if multiple processing threads are available. Second, the sorter, due to having a larger max channels per merger, will make more efficient use of intermediate channels.