refactor(inkless): pipelining committer threading #300
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
Instead of running this as a separate thread, run directly on the same thread as it is not an IO task. Reduces thread allocation.
Cached thread polls are unbounded. As concurrent uploads wait for commits to complete, the amound of uploads can grow without limits, affecting cpu utilization and increasing latencies. To bound the uploads to the expected commit latencies, this PR proposes to use fixed thread pool (8 by default) to be able to upload up to 8 files concurrently, these uploads have a P99 of 400ms on S3. As sequential commits can have a P99 of 50ms on PG, having 8 commits queued would already account for doubling the upload time. This is an initial estimation and the values could be updated and the defaults changed as this setup is tested; but having configurable thread pools should benefit in the long term.
At the moment, the threads are managed by the jobs. Specially, the commit job blocks on upload future. Instead we can use CompletableFuture to compose the thread coordination. This should help with better resource utilization.
|
WriteIntegrationTest is getting flaky which may indicate the proposed approach is breaking ordering on the response. Have to look a bit deeper into this. |
Base automatically changed from
jeqo/fixed-thread-pools
to
jeqo/refactor-append-completer
May 22, 2025 08:53
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Outdated suggestions cannot be applied.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.
At the moment, the threads are managed by the jobs. Specially, the commit job blocks on upload future.
Instead we can use CompletableFuture to compose the thread coordination. This should help with better resource utilization.