feat(replay): Use requestIdleCallback for events
#7191
Closed
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
mydea
added
the
CI-Overhead-Measurements
size-limit report 📦
|
Replay SDK metrics 🚀
develop |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Revision | LCP | CLS | CPU | JS heap avg | JS heap max | netTx | netRx | netCount | netTime |
|---|---|---|---|---|---|---|---|---|---|
| 1c9ed4f | +50.07 ms | -0.00 ms | +6.39 pp | +928.84 kB | +1.05 MB | +2.2 kB | +41 B | +1 | +70.97 ms |
| 1921888 | +55.79 ms | +0.00 ms | +5.50 pp | +926.71 kB | +1.06 MB | +2.23 kB | +41 B | +1 | +79.78 ms |
| e9eec27 | +61.38 ms | -0.00 ms | +6.04 pp | +927.84 kB | +1.05 MB | +2.21 kB | +41 B | +1 | +88.58 ms |
| d604022 | +58.83 ms | -0.00 ms | +7.65 pp | +930.16 kB | +1.05 MB | +2.21 kB | +41 B | +1 | +109.63 ms |
| a961e57 | +54.75 ms | -0.00 ms | +6.50 pp | +929.18 kB | +1.07 MB | +2.21 kB | +41 B | +1 | +92.73 ms |
| f7c0a2f | +46.14 ms | +0.00 ms | +6.37 pp | +921.47 kB | +1.06 MB | +2.23 kB | +41 B | +1 | +207.30 ms |
| cb19818 | +57.16 ms | +0.00 ms | +11.95 pp | +1.07 MB | +2.21 MB | +2.52 kB | +41 B | +1 | +111.50 ms |
| ee301c3 | +71.07 ms | -0.00 ms | +12.64 pp | +1.07 MB | +2.22 MB | +2.55 kB | +41 B | +1 | +94.67 ms |
| 93c4759 | +61.10 ms | -0.00 ms | +12.72 pp | +1.08 MB | +2.19 MB | +2.57 kB | +41 B | +1 | +116.75 ms |
*) pp - percentage points - an absolute difference between two percentages.
Last updated: Wed, 15 Feb 2023 11:25:05 GMT
Last updated: Wed, 15 Feb 2023 11:25:05 GMT
AbhiPrasad
approved these changes
Feb 15, 2023
billyvg
requested changes
Feb 16, 2023
Yes, definitely! |
|
Closing this for now, as we're going to explore a different approach - see getsentry/rrweb#58 |
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.
This PR updates the event handling of replay to use
window.requestIdleCallback, if available.This API is supported everywhere except for Safari, and should schedule a callback to be run when the browser has "resources" for it.
Hopefully, this should improve performance in scenarios where a lot of events are emitted.
I defined a max. timeout of
50ms, meaning that after max. 50ms each callback will be executed anyhow. This could be tweaked, but feels like a reasonably defensive value here. If the API is not available, we just run the callbacks immediately (like now), so it should degrade gracefully.Note that there would also be an approach to only have a single callback that keeps an internal queue, but that would require more work to ensure that events cannot be postponed for too long, so IMHO this is the easier solution.
I tried this in my super simple test app, where it works as expected. it would be nice though to test this in a more complex app...
Note: I also refactored the
addUpdatemethod to split this into dedicated methods, as this was a bit mixed up right now. Now, we have a dedicated (private) method for the event emitting, and one for adding a "generic" event from other handlers etc.See https://w3c.github.io/requestidlecallback/ for some more info.
ref #6946
ref #7085