-
Notifications
You must be signed in to change notification settings - Fork 126
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
chore(qlog): Add meaningful qlog titles and descriptions. #2281
chore(qlog): Add meaningful qlog titles and descriptions. #2281
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you!
Apart from the one suggestion, this looks good to me.
Let's give others a bit of time to review as well before merging.
Failed Interop TestsQUIC Interop Runner, client vs. server, differences relative to fd42bed. 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
|
@mansf-osk unit tests are failing. It should be as simple as updating the constant below: Lines 419 to 423 in fd42bed
You can run them locally via |
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #2281 +/- ##
=======================================
Coverage 93.42% 93.43%
=======================================
Files 113 113
Lines 36743 36743
Branches 1632 1632
=======================================
+ Hits 34327 34329 +2
Misses 1689 1689
+ Partials 727 725 -2 ☔ View full report in Codecov by Sentry. |
Benchmark resultsPerformance differences relative to fd42bed. decode 4096 bytes, mask ff: No change in performance detected.time: [11.844 µs 11.875 µs 11.913 µs] change: [-2.3064% -0.6455% +0.8317%] (p = 0.47 > 0.05) decode 1048576 bytes, mask ff: No change in performance detected.time: [3.1110 ms 3.1344 ms 3.1687 ms] change: [-0.0578% +0.8095% +1.9023%] (p = 0.12 > 0.05) decode 4096 bytes, mask 7f: No change in performance detected.time: [19.804 µs 19.862 µs 19.923 µs] change: [-0.3626% +0.0859% +0.6318%] (p = 0.73 > 0.05) decode 1048576 bytes, mask 7f: No change in performance detected.time: [5.1721 ms 5.1882 ms 5.2080 ms] change: [-0.0793% +0.3291% +0.7621%] (p = 0.12 > 0.05) decode 4096 bytes, mask 3f: No change in performance detected.time: [6.9029 µs 6.9337 µs 6.9706 µs] change: [-0.2766% +0.1656% +0.6230%] (p = 0.49 > 0.05) decode 1048576 bytes, mask 3f: No change in performance detected.time: [1.7597 ms 1.7640 ms 1.7695 ms] change: [-0.5255% -0.0601% +0.3404%] (p = 0.85 > 0.05) coalesce_acked_from_zero 1+1 entries: 💚 Performance has improved.time: [99.350 ns 99.599 ns 99.866 ns] change: [-5.2651% -4.8418% -4.3782%] (p = 0.00 < 0.05) coalesce_acked_from_zero 3+1 entries: 💚 Performance has improved.time: [117.55 ns 117.94 ns 118.36 ns] change: [-2.9611% -2.3882% -1.8502%] (p = 0.00 < 0.05) coalesce_acked_from_zero 10+1 entries: 💚 Performance has improved.time: [117.31 ns 118.13 ns 119.23 ns] change: [-2.5171% -2.0768% -1.6167%] (p = 0.00 < 0.05) coalesce_acked_from_zero 1000+1 entries: 💚 Performance has improved.time: [97.472 ns 97.611 ns 97.763 ns] change: [-4.4914% -3.8239% -3.2083%] (p = 0.00 < 0.05) RxStreamOrderer::inbound_frame(): Change within noise threshold.time: [111.26 ms 111.40 ms 111.62 ms] change: [+0.2249% +0.3587% +0.5745%] (p = 0.00 < 0.05) SentPackets::take_ranges: No change in performance detected.time: [5.5429 µs 5.6911 µs 5.8490 µs] change: [-3.7978% -0.8247% +2.1471%] (p = 0.59 > 0.05) transfer/pacing-false/varying-seeds: No change in performance detected.time: [28.253 ms 28.773 ms 29.288 ms] change: [-0.0748% +2.7210% +5.4849%] (p = 0.06 > 0.05) transfer/pacing-true/varying-seeds: Change within noise threshold.time: [29.696 ms 30.213 ms 30.720 ms] change: [-5.0001% -2.7205% -0.2172%] (p = 0.03 < 0.05) transfer/pacing-false/same-seed: No change in performance detected.time: [24.270 ms 24.806 ms 25.336 ms] change: [-2.3270% +0.5814% +3.6418%] (p = 0.70 > 0.05) transfer/pacing-true/same-seed: No change in performance detected.time: [27.766 ms 28.642 ms 29.511 ms] change: [-1.8316% +2.3713% +6.7333%] (p = 0.27 > 0.05) 1-conn/1-100mb-resp/mtu-1504 (aka. Download)/client: No change in performance detected.time: [860.60 ms 870.08 ms 879.78 ms] thrpt: [113.66 MiB/s 114.93 MiB/s 116.20 MiB/s] change: time: [-2.4346% -0.9967% +0.5069%] (p = 0.18 > 0.05) thrpt: [-0.5044% +1.0067% +2.4953%] 1-conn/10_000-parallel-1b-resp/mtu-1504 (aka. RPS)/client: No change in performance detected.time: [315.69 ms 318.35 ms 321.03 ms] thrpt: [31.149 Kelem/s 31.412 Kelem/s 31.676 Kelem/s] change: time: [-0.3350% +0.9085% +2.0989%] (p = 0.15 > 0.05) thrpt: [-2.0558% -0.9003% +0.3362%] 1-conn/1-1b-resp/mtu-1504 (aka. HPS)/client: No change in performance detected.time: [33.620 ms 33.809 ms 34.011 ms] thrpt: [29.403 elem/s 29.578 elem/s 29.744 elem/s] change: time: [-0.9962% -0.1077% +0.7489%] (p = 0.81 > 0.05) thrpt: [-0.7434% +0.1078% +1.0062%] 1-conn/1-100mb-resp/mtu-1504 (aka. Upload)/client: 💔 Performance has regressed.time: [1.6859 s 1.7055 s 1.7255 s] thrpt: [57.954 MiB/s 58.635 MiB/s 59.317 MiB/s] change: time: [+1.6004% +3.2026% +4.9465%] (p = 0.00 < 0.05) thrpt: [-4.7134% -3.1032% -1.5752%] Client/server transfer resultsTransfer of 33554432 bytes over loopback.
|
Failure on Windows installing Rust seems unrelated. Retriggered. |
Hey Hey! 👋
First time neqo contributor here!
I'm Oskar, working student on the necko team, for those that don't already know me 😄
@mxinden was showing me qvis the other day and we noticed that description and title were just showing something the likes of
Example qlog [...]
. With this change they now showNeqo [client|server] qlog
.I also added a
server-
prefix to the filename of the server qlogs as I found it hard to understand which was which at first glance, but happy to take that out again if it's not wanted.I hope these small change are somewhat useful and would appreciate a review and maybe a pointer as to what I could work on next. I'm definitely very interested to work more on neqo! 🙂
Cheers
Oskar