PQ: avoid deserialization cost (POC) #17820
Draft
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
Avoids deserialization cost by holding a list of soft-references on each PQ page to the actual event objects from the page, allowing GC to collect the referenced objects in the event of memory pressure. While forming batches during reads, elements are returned from the cache until one is not present, at which point we fall through to reading the bytes from the memory-mapped file, returning an intermediate `BoxedQueueable` in either case. The elements are then unboxed outside of the queue locking mechanism.
🤖 GitHub commentsExpand to view the GitHub comments
Just comment with:
|
|
This pull request does not have a backport label. Could you fix it @yaauie? 🙏
|
|
💚 Build Succeeded
|
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Outdated suggestions cannot be applied.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 does this PR do?
Avoids deserialization cost of events passing through the PQ when memory pressure is absent, by maintaining a list of soft-references to the events added to each PQ page. When constructing batches during reads, we use a new internal intermediate type
BoxedQueueable, each of which holds either a strong-reference to aQueueableor thebyte[]and deserializer necessary to construct it. Deserialization, if any, still occurs outside of the Queue locking mechanism.Why is it important/What is the impact to the user?
Reduces the overhead of reading from a PQ when a pipeline is staying caught up with the workload.
Status
This is a proof-of-concept; actual implementation may change significantly. As a proof-of-concept spike, I did not take the time to arrange it into bite-size changes.