Skip to content
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

fix: Correctly manage bytes_in_flight during a rebinding event #1812

Merged
merged 2 commits into from
Apr 11, 2024

Conversation

larseggert
Copy link
Collaborator

Fixes #1289

Copy link

codecov bot commented Apr 10, 2024

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.13%. Comparing base (c44e536) to head (412dc22).

Additional details and impacted files
@@            Coverage Diff             @@
##             main    #1812      +/-   ##
==========================================
- Coverage   93.13%   93.13%   -0.01%     
==========================================
  Files         117      117              
  Lines       36339    36337       -2     
==========================================
- Hits        33843    33841       -2     
  Misses       2496     2496              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@larseggert larseggert changed the title fix: Correctly manage bytes_in_flight during a rebinding event fix: Correctly manage bytes_in_flight during a rebinding event Apr 10, 2024
Copy link

Benchmark results

Performance differences relative to c44e536.

  • drain a timer quickly time: [309.01 ns 318.07 ns 326.63 ns]
    change: [-2.3861% -0.1270% +2.1980%] (p = 0.91 > 0.05)
    No change in performance detected.

  • coalesce_acked_from_zero 1+1 entries
    time: [196.83 ns 197.21 ns 197.64 ns]
    change: [-2.6406% -1.1766% -0.2719%] (p = 0.04 < 0.05)
    Change within noise threshold.

  • coalesce_acked_from_zero 3+1 entries
    time: [242.24 ns 242.80 ns 243.41 ns]
    change: [-0.5633% -0.2732% +0.0465%] (p = 0.08 > 0.05)
    No change in performance detected.

  • coalesce_acked_from_zero 10+1 entries
    time: [241.66 ns 242.52 ns 243.57 ns]
    change: [-0.2138% +0.4353% +1.1415%] (p = 0.23 > 0.05)
    No change in performance detected.

  • coalesce_acked_from_zero 1000+1 entries
    time: [218.67 ns 218.85 ns 219.06 ns]
    change: [-1.1135% -0.3454% +0.4797%] (p = 0.40 > 0.05)
    No change in performance detected.

  • RxStreamOrderer::inbound_frame()
    time: [118.20 ms 118.27 ms 118.34 ms]
    change: [-0.2655% -0.0567% +0.0840%] (p = 0.62 > 0.05)
    No change in performance detected.

  • transfer/Run multiple transfers with varying seeds
    time: [119.59 ms 119.90 ms 120.21 ms]
    thrpt: [33.275 MiB/s 33.360 MiB/s 33.446 MiB/s]
    change:
    time: [-1.5563% -1.2182% -0.8583%] (p = 0.00 < 0.05)
    thrpt: [+0.8658% +1.2332% +1.5809%]
    Change within noise threshold.

  • transfer/Run multiple transfers with the same seed
    time: [121.04 ms 121.25 ms 121.47 ms]
    thrpt: [32.929 MiB/s 32.989 MiB/s 33.046 MiB/s]
    change:
    time: [-1.0089% -0.7663% -0.5152%] (p = 0.00 < 0.05)
    thrpt: [+0.5179% +0.7722% +1.0192%]
    Change within noise threshold.

  • 1-conn/1-100mb-resp (aka. Download)/client
    time: [1.0960 s 1.1084 s 1.1239 s]
    thrpt: [88.979 MiB/s 90.223 MiB/s 91.245 MiB/s]
    change:
    time: [+0.1919% +1.5142% +3.0071%] (p = 0.05 < 0.05)
    thrpt: [-2.9193% -1.4916% -0.1916%]
    Change within noise threshold.

  • 1-conn/10_000-parallel-1b-resp (aka. RPS)/client
    time: [421.61 ms 424.06 ms 426.56 ms]
    thrpt: [23.444 Kelem/s 23.582 Kelem/s 23.718 Kelem/s]
    change:
    time: [-1.9746% -1.2541% -0.5067%] (p = 0.00 < 0.05)
    thrpt: [+0.5093% +1.2701% +2.0144%]
    Change within noise threshold.

  • 1-conn/1-1b-resp (aka. HPS)/client
    time: [51.936 ms 52.343 ms 52.756 ms]
    thrpt: [18.955 elem/s 19.105 elem/s 19.255 elem/s]
    change:
    time: [-1.4258% -0.3196% +0.8280%] (p = 0.59 > 0.05)
    thrpt: [-0.8212% +0.3206% +1.4464%]
    No change in performance detected.

Client/server transfer results

Transfer of 134217728 bytes over loopback.

Client Server CC Pacing Mean [ms] Min [ms] Max [ms] Relative
msquic msquic 836.9 ± 284.3 547.7 1486.0 1.00
neqo msquic reno on 896.7 ± 185.6 769.6 1322.7 1.00
neqo msquic reno 980.9 ± 225.2 802.1 1386.5 1.00
neqo msquic cubic on 915.9 ± 218.8 749.9 1370.8 1.00
neqo msquic cubic 899.7 ± 178.2 751.8 1303.6 1.00
msquic neqo reno on 4409.0 ± 272.3 4157.0 5087.9 1.00
msquic neqo reno 4448.2 ± 315.6 4058.5 5020.1 1.00
msquic neqo cubic on 4421.6 ± 286.0 4153.3 5155.9 1.00
msquic neqo cubic 4428.6 ± 163.7 4152.4 4742.7 1.00
neqo neqo reno on 3728.4 ± 237.6 3384.3 4200.5 1.00
neqo neqo reno 3681.4 ± 169.2 3388.5 3862.7 1.00
neqo neqo cubic on 4357.0 ± 192.0 4080.7 4758.4 1.00
neqo neqo cubic 4299.4 ± 297.8 4012.9 4974.2 1.00

⬇️ Download logs

@larseggert larseggert added this pull request to the merge queue Apr 11, 2024
Merged via the queue into mozilla:main with commit c0ca26d Apr 11, 2024
15 checks passed
@larseggert larseggert deleted the fix-1289 branch April 11, 2024 05:35
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Panic when connection IP address changes
2 participants