Fix metrics recorder task leak - #5602
Closed
joshua-spacetime wants to merge 1 commit into
Closed
Conversation
joshua-spacetime
changed the title
Contributor
Author
|
Note, we should go with #5601 if this change regresses the Keynote Benchmark. |
Contributor
Author
|
It does regress the benchmark, so closing this in favor of #5601. |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description of Changes
Removes the 5ms sleep in the metrics recording task, as well as the explicit abort handle.
Previously we were dropping this task on database init failure. Hence every such failure would leak/accumulate these tasks that would wake every 5ms. This could result in meaningful and significant context switching given enough of these orphaned tasks.
This change removes the sleep and the abort handle. The task is dropped when the sender is dropped/closed and we have drained the receiver.
Note, the 5ms sleep was added as a micro-optimization for a particular benchmark, however requests are now fully pipelined through all layers of the database, and so my hunch is that this optimization no longer makes a difference and therefore can be removed.
API and ABI breaking changes
None
Expected complexity level and risk
2
Testing
TBD