perf(udp): multi-message receive on apple #2302
Merged
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
Failed Interop TestsQUIC Interop Runner, client vs. server, differences relative to ef9e4dc. neqo-latest as client
neqo-latest as server
All resultsSucceeded Interop TestsQUIC Interop Runner, client vs. server neqo-latest as client
neqo-latest as server
Unsupported Interop TestsQUIC Interop Runner, client vs. server neqo-latest as client
neqo-latest as server
|
mxinden
changed the title
mxinden
force-pushed
the
apple-mmsg
branch
7 times, most recently
from
aeb6fff to
6f98357
Compare
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #2302 +/- ##
==========================================
- Coverage 93.34% 93.33% -0.02%
==========================================
Files 114 114
Lines 36908 36889 -19
Branches 36908 36889 -19
==========================================
- Hits 34451 34429 -22
- Misses 1675 1679 +4
+ Partials 782 781 -1 ☔ View full report in Codecov by Sentry. |
larseggert
approved these changes
Jan 7, 2025
larseggert
reviewed
Jan 9, 2025
Benchmark resultsPerformance differences relative to ef9e4dc. decode 4096 bytes, mask ff: No change in performance detected. time: [11.168 µs 11.205 µs 11.248 µs]
change: [-0.3788% +0.1613% +0.7455%] (p = 0.58 > 0.05)
decode 1048576 bytes, mask ff: 💔 Performance has regressed. time: [3.0172 ms 3.0265 ms 3.0375 ms]
change: [+3.9724% +4.5153% +5.0587%] (p = 0.00 < 0.05)
decode 4096 bytes, mask 7f: No change in performance detected. time: [19.533 µs 19.582 µs 19.638 µs]
change: [-0.2929% +0.0864% +0.4468%] (p = 0.66 > 0.05)
decode 1048576 bytes, mask 7f: 💔 Performance has regressed. time: [5.1594 ms 5.1706 ms 5.1834 ms]
change: [+1.2562% +1.6462% +2.0193%] (p = 0.00 < 0.05)
decode 4096 bytes, mask 3f: No change in performance detected. time: [5.5262 µs 5.5440 µs 5.5706 µs]
change: [-1.0699% -0.4711% +0.0274%] (p = 0.10 > 0.05)
decode 1048576 bytes, mask 3f: 💔 Performance has regressed. time: [1.7625 ms 1.7694 ms 1.7776 ms]
change: [+23.916% +24.654% +25.381%] (p = 0.00 < 0.05)
coalesce_acked_from_zero 1+1 entries: Change within noise threshold. time: [99.199 ns 99.471 ns 99.750 ns]
change: [+0.0355% +0.4283% +0.7920%] (p = 0.02 < 0.05)
coalesce_acked_from_zero 3+1 entries: Change within noise threshold. time: [117.56 ns 117.92 ns 118.31 ns]
change: [+0.5061% +0.8668% +1.2566%] (p = 0.00 < 0.05)
coalesce_acked_from_zero 10+1 entries: Change within noise threshold. time: [117.16 ns 117.69 ns 118.31 ns]
change: [-0.0775% +0.6481% +1.2617%] (p = 0.04 < 0.05)
coalesce_acked_from_zero 1000+1 entries: Change within noise threshold. time: [98.535 ns 98.674 ns 98.837 ns]
change: [+0.8541% +1.8132% +2.7847%] (p = 0.00 < 0.05)
RxStreamOrderer::inbound_frame(): Change within noise threshold. time: [111.19 ms 111.23 ms 111.26 ms]
change: [-0.2096% -0.1446% -0.0814%] (p = 0.00 < 0.05)
SentPackets::take_ranges: No change in performance detected. time: [5.4984 µs 5.6614 µs 5.8311 µs]
change: [-15.570% -4.1665% +4.2182%] (p = 0.64 > 0.05)
transfer/pacing-false/varying-seeds: Change within noise threshold. time: [41.883 ms 41.958 ms 42.037 ms]
change: [-2.7910% -2.5481% -2.2906%] (p = 0.00 < 0.05)
transfer/pacing-true/varying-seeds: Change within noise threshold. time: [42.297 ms 42.372 ms 42.446 ms]
change: [-2.3559% -2.1151% -1.8833%] (p = 0.00 < 0.05)
transfer/pacing-false/same-seed: Change within noise threshold. time: [41.615 ms 41.674 ms 41.732 ms]
change: [-2.8392% -2.6525% -2.4468%] (p = 0.00 < 0.05)
transfer/pacing-true/same-seed: Change within noise threshold. time: [42.313 ms 42.372 ms 42.430 ms]
change: [-2.6838% -2.4963% -2.3095%] (p = 0.00 < 0.05)
1-conn/1-100mb-resp/mtu-1504 (aka. Download)/client: No change in performance detected. time: [883.30 ms 893.00 ms 902.85 ms]
thrpt: [110.76 MiB/s 111.98 MiB/s 113.21 MiB/s]
change:
time: [-1.0177% +0.5464% +2.0802%] (p = 0.49 > 0.05)
thrpt: [-2.0379% -0.5434% +1.0282%]
1-conn/10_000-parallel-1b-resp/mtu-1504 (aka. RPS)/client: No change in performance detected. time: [302.24 ms 304.42 ms 306.68 ms]
thrpt: [32.607 Kelem/s 32.849 Kelem/s 33.086 Kelem/s]
change:
time: [-0.3781% +0.5948% +1.6414%] (p = 0.25 > 0.05)
thrpt: [-1.6149% -0.5913% +0.3795%]
1-conn/1-1b-resp/mtu-1504 (aka. HPS)/client: Change within noise threshold. time: [34.065 ms 34.258 ms 34.469 ms]
thrpt: [29.012 elem/s 29.191 elem/s 29.355 elem/s]
change:
time: [-2.2643% -1.3232% -0.4249%] (p = 0.00 < 0.05)
thrpt: [+0.4267% +1.3410% +2.3168%]
1-conn/1-100mb-resp/mtu-1504 (aka. Upload)/client: 💚 Performance has improved. time: [1.5954 s 1.6102 s 1.6251 s]
thrpt: [61.534 MiB/s 62.105 MiB/s 62.682 MiB/s]
change:
time: [-6.1738% -4.9823% -3.8219%] (p = 0.00 < 0.05)
thrpt: [+3.9738% +5.2435% +6.5801%]
Client/server transfer resultsTransfer of 33554432 bytes over loopback.
|
mxinden
added a commit
to mxinden/neqo
that referenced
this pull request
Jan 13, 2025
This reverts commit 3c48567.
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.
Do multi-message receive via quinn-rs/quinn#1993 on apple platforms.
Localhost benchmark (Download) with PMTUD disabled shows ~11% performance improvement.
Fixes #2279.