[MM-54335] Improve write concurrency for screen tracks #138
Conversation
This file contains 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
streamer45
added
2: Dev Review
streamer45
commented
May 8, 2024
streamer45
commented
May 8, 2024
cpoile
approved these changes
May 8, 2024
streamer45
added
3: Reviews Complete
streamer45
removed
the
Do Not Merge
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Outdated suggestions cannot be applied.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.
Summary
Initial ceiling tests identified a bottleneck when writing screen tracks in large calls (> 250 participants). This is because the write loop runs on a single goroutine since we use a single output track. This results in the CPU not being utilized to the fullest and packets getting significantly delayed or even dropped.
To overcome the current limitation, this PR proposes to create N output tracks and concurrently write to them.
These changes are showing to improve overall performance and CPU utilization on our target instance (
c7i.xlarge) by a factor of ~2.5, getting us from around 300 participants to 800 per call with near-zero packet loss.Ticket Link
https://mattermost.atlassian.net/browse/MM-54335