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.

Sign up for GitHub

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

Jump to bottom

ci: Experiment with GCP #1892

Closed
wants to merge 1 commit into from
Closed

ci: Experiment with GCP #1892

wants to merge 1 commit into from

Conversation

larseggert
Copy link
Collaborator

WIP

@github-actions GitHub Actions
Copy link

github-actions bot commented May 8, 2024

Failed Interop Tests

QUIC Interop Runner, client vs. server

All results

Succeeded Interop Tests

QUIC Interop Runner, client vs. server

Unsupported Interop Tests

QUIC Interop Runner, client vs. server

@codecov Codecov
Copy link

codecov bot commented May 8, 2024

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.20%. Comparing base (beffcc6) to head (adf5380).

Additional details and impacted files 
@@           Coverage Diff           @@
##             main    #1892   +/-   ##
=======================================
  Coverage   93.20%   93.20%           
=======================================
  Files         111      111           
  Lines       36027    36027           
=======================================
  Hits        33580    33580           
  Misses       2447     2447

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

@larseggert larseggert closed this May 8, 2024
@larseggert larseggert deleted the ci-gcp branch May 8, 2024 07:18
@github-actions GitHub Actions
Copy link

github-actions bot commented May 8, 2024

Benchmark results

Performance differences relative to beffcc6.

  • drain a timer quickly time: [309.40 ns 316.44 ns 322.91 ns]
    change: [-2.0801% +0.0186% +2.1595%] (p = 0.99 > 0.05)
    No change in performance detected.

  • coalesce_acked_from_zero 1+1 entries
    time: [194.40 ns 194.82 ns 195.27 ns]
    change: [-2.2637% -1.0985% -0.3531%] (p = 0.01 < 0.05)
    Change within noise threshold.

  • coalesce_acked_from_zero 3+1 entries
    time: [239.51 ns 242.14 ns 247.04 ns]
    change: [+0.7750% +1.5070% +2.3547%] (p = 0.00 < 0.05)
    Change within noise threshold.

  • coalesce_acked_from_zero 10+1 entries
    time: [239.43 ns 240.26 ns 241.24 ns]
    change: [+0.7320% +1.4052% +2.2637%] (p = 0.00 < 0.05)
    Change within noise threshold.

  • coalesce_acked_from_zero 1000+1 entries
    time: [217.43 ns 217.69 ns 218.01 ns]
    change: [-0.1801% +0.4628% +1.0576%] (p = 0.15 > 0.05)
    No change in performance detected.

  • RxStreamOrderer::inbound_frame()
    time: [119.96 ms 120.01 ms 120.06 ms]
    change: [+0.7294% +0.7979% +0.8659%] (p = 0.00 < 0.05)
    Change within noise threshold.

  • transfer/Run multiple transfers with varying seeds
    time: [118.91 ms 119.15 ms 119.39 ms]
    thrpt: [33.503 MiB/s 33.571 MiB/s 33.638 MiB/s]
    change:
    time: [-1.2521% -0.9493% -0.6630%] (p = 0.00 < 0.05)
    thrpt: [+0.6674% +0.9584% +1.2680%]
    Change within noise threshold.

  • transfer/Run multiple transfers with the same seed
    time: [118.56 ms 118.86 ms 119.16 ms]
    thrpt: [33.569 MiB/s 33.653 MiB/s 33.738 MiB/s]
    change:
    time: [-1.5038% -1.1855% -0.8968%] (p = 0.00 < 0.05)
    thrpt: [+0.9050% +1.1997% +1.5268%]
    Change within noise threshold.

  • 1-conn/1-100mb-resp (aka. Download)/client
    time: [1.1405 s 1.1592 s 1.1922 s]
    thrpt: [83.877 MiB/s 86.265 MiB/s 87.685 MiB/s]
    change:
    time: [+1.2426% +3.5733% +7.1097%] (p = 0.02 < 0.05)
    thrpt: [-6.6378% -3.4500% -1.2274%]
    💔 Performance has regressed.

  • 1-conn/10_000-parallel-1b-resp (aka. RPS)/client
    time: [424.40 ms 426.40 ms 428.45 ms]
    thrpt: [23.340 Kelem/s 23.452 Kelem/s 23.563 Kelem/s]
    change:
    time: [-0.8743% -0.2151% +0.4987%] (p = 0.54 > 0.05)
    thrpt: [-0.4962% +0.2155% +0.8820%]
    No change in performance detected.

  • 1-conn/1-1b-resp (aka. HPS)/client
    time: [49.823 ms 49.948 ms 50.116 ms]
    thrpt: [19.954 elem/s 20.021 elem/s 20.071 elem/s]
    change:
    time: [-1.2313% -0.2312% +0.4991%] (p = 0.68 > 0.05)
    thrpt: [-0.4966% +0.2317% +1.2467%]
    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 703.0 ± 293.6 403.3 1182.5 1.00
neqo msquic reno on 921.2 ± 178.9 803.8 1319.5 1.00
neqo msquic reno 898.8 ± 226.0 720.0 1373.2 1.00
neqo msquic cubic on 979.6 ± 228.0 792.8 1375.1 1.00
neqo msquic cubic 911.2 ± 159.2 776.3 1259.0 1.00
msquic neqo reno on 5403.2 ± 307.2 4989.0 5980.4 1.00
msquic neqo reno 5428.0 ± 368.7 5010.1 6363.0 1.00
msquic neqo cubic on 5107.1 ± 269.8 4750.2 5618.4 1.00
msquic neqo cubic 5057.8 ± 239.3 4765.7 5488.9 1.00
neqo neqo reno on 4368.8 ± 510.7 3029.3 4838.9 1.00
neqo neqo reno 4254.0 ± 165.2 3963.2 4498.6 1.00
neqo neqo cubic on 4927.3 ± 163.7 4570.8 5081.1 1.00
neqo neqo cubic 4789.8 ± 271.2 4122.6 5099.0 1.00

⬇️ Download logs

@github-actions GitHub Actions
Copy link

github-actions bot commented May 8, 2024

Firefox builds for this PR

The following builds are available for testing. Crossed-out builds did not succeed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@KershawChang KershawChang Awaiting requested review from KershawChang KershawChang will be requested when the pull request is marked ready for review KershawChang is a code owner

@martinthomson martinthomson Awaiting requested review from martinthomson martinthomson will be requested when the pull request is marked ready for review martinthomson is a code owner

@mxinden mxinden Awaiting requested review from mxinden mxinden will be requested when the pull request is marked ready for review mxinden is a code owner

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
1 participant
@larseggert