Skip to content

[C#] GenHTTP: Enable SSL and configure compression globally#239

Merged
MDA2AV merged 3 commits intoMDA2AV:mainfrom
Kaliumhexacyanoferrat:genhttp-global-compression
Mar 29, 2026
Merged

[C#] GenHTTP: Enable SSL and configure compression globally#239
MDA2AV merged 3 commits intoMDA2AV:mainfrom
Kaliumhexacyanoferrat:genhttp-global-compression

Conversation

@Kaliumhexacyanoferrat
Copy link
Copy Markdown
Collaborator

@Kaliumhexacyanoferrat Kaliumhexacyanoferrat commented Mar 29, 2026

No description provided.

@Kaliumhexacyanoferrat Kaliumhexacyanoferrat self-assigned this Mar 29, 2026
@Kaliumhexacyanoferrat
Copy link
Copy Markdown
Collaborator Author

Kaliumhexacyanoferrat commented Mar 29, 2026

/benchmark

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Mar 29, 2026

🚀 Benchmark run triggered for genhttp (all profiles). Results will be posted here when done.

@github-actions
Copy link
Copy Markdown
Contributor

github-actions bot commented Mar 29, 2026

Benchmark Results

Framework: genhttp | Profile: all profiles

genhttp / baseline / 512c (p=1, r=0, cpu=64)
  Best: 46553 req/s (CPU: 4815.8%, Mem: 286.9MiB) ===

genhttp / baseline / 4096c (p=1, r=0, cpu=64)
  Best: 40969 req/s (CPU: 5078.8%, Mem: 259.5MiB) ===

genhttp / baseline / 16384c (p=1, r=0, cpu=64)
  Best: 36277 req/s (CPU: 4741.5%, Mem: 381.3MiB) ===

genhttp / pipelined / 512c (p=16, r=0, cpu=unlimited)
  Best: 12112307 req/s (CPU: 7322.6%, Mem: 311.2MiB) ===

genhttp / pipelined / 4096c (p=16, r=0, cpu=unlimited)
  Best: 12944387 req/s (CPU: 8121.4%, Mem: 1.4GiB) ===

genhttp / pipelined / 16384c (p=16, r=0, cpu=unlimited)
  Best: 11694105 req/s (CPU: 8471.3%, Mem: 3.3GiB) ===

genhttp / limited-conn / 512c (p=1, r=10, cpu=unlimited)
  Best: 50599 req/s (CPU: 7483.7%, Mem: 434.7MiB) ===

genhttp / limited-conn / 4096c (p=1, r=10, cpu=unlimited)
  Best: 49648 req/s (CPU: 7610.2%, Mem: 648.5MiB) ===

genhttp / json / 4096c (p=1, r=0, cpu=unlimited)
  Best: 643092 req/s (CPU: 8900.9%, Mem: 1.1GiB) ===

genhttp / json / 16384c (p=1, r=0, cpu=unlimited)
  Best: 549495 req/s (CPU: 8820.6%, Mem: 4.4GiB) ===

genhttp / upload / 64c (p=1, r=0, cpu=unlimited)
  Best: 589 req/s (CPU: 6829.3%, Mem: 467.2MiB) ===

genhttp / upload / 256c (p=1, r=0, cpu=unlimited)
  Best: 607 req/s (CPU: 11424.8%, Mem: 702.8MiB) ===

genhttp / upload / 512c (p=1, r=0, cpu=unlimited)
  Best: 558 req/s (CPU: 9950.4%, Mem: 801.0MiB) ===

genhttp / compression / 4096c (p=1, r=0, cpu=unlimited)
  Best: 8386 req/s (CPU: 3341.5%, Mem: 1.2GiB) ===

genhttp / compression / 16384c (p=1, r=0, cpu=unlimited)
  Best: 7673 req/s (CPU: 3163.0%, Mem: 3.1GiB) ===

genhttp / noisy / 512c (p=1, r=0, cpu=unlimited)
  Best: 1058915 req/s (CPU: 6378.1%, Mem: 333.9MiB) ===

genhttp / noisy / 4096c (p=1, r=0, cpu=unlimited)
  Best: 1216943 req/s (CPU: 7312.9%, Mem: 1.2GiB) ===

genhttp / noisy / 16384c (p=1, r=0, cpu=unlimited)
  Best: 1071182 req/s (CPU: 7174.2%, Mem: 4.4GiB) ===

genhttp / mixed / 4096c (p=1, r=5, cpu=unlimited)
  Best: 7934 req/s (CPU: 9894.0%, Mem: 1.9GiB) ===

genhttp / mixed / 16384c (p=1, r=5, cpu=unlimited)
  Best: 9307 req/s (CPU: 9889.6%, Mem: 6.1GiB) ===

genhttp / static / 4096c (p=1, r=0, cpu=unlimited)
  Best: 535283 req/s (CPU: 9137.4%, Mem: 1.4GiB) ===

genhttp / static / 16384c (p=1, r=0, cpu=unlimited)
  Best: 402083 req/s (CPU: 9023.9%, Mem: 3.5GiB) ===

genhttp / async-db / 512c (p=1, r=0, cpu=unlimited)
  Best: 138240 req/s (CPU: 3825.3%, Mem: 420.1MiB) ===

genhttp / async-db / 1024c (p=1, r=0, cpu=unlimited)
  Best: 160564 req/s (CPU: 4392.3%, Mem: 637.9MiB) ===
Full log 
    Latency   32.50ms   22.30ms   67.30ms   143.00ms    1.62s

  2026546 requests in 5.00s, 2010418 responses
  Throughput: 401.84K req/s
  Bandwidth:  5.90GB/s
  Status codes: 2xx=2010418, 3xx=0, 4xx=0, 5xx=0
  Latency samples: 2010418 / 2010418 responses (100.0%)
  Per-template: 105509,103980,106531,100444,101083,100111,102873,97793,98860,99242,103233,104993,96476,97363,101280,99988,90542,99223,99054,101840
  Per-template-ok: 105509,103980,106531,100444,101083,100111,102873,97793,98860,99242,103233,104993,96476,97363,101280,99988,90542,99223,99054,101840
  CPU: 9023.9% | Mem: 3.5GiB

[run 3/3]
gcannon — io_uring HTTP load generator
  Target:    localhost:8080/
  Threads:   64
  Conns:     16384 (256/thread)
  Pipeline:  1
  Req/conn:  unlimited (keep-alive)
  Templates: 20
  Expected:  200
  Duration:  5s


  Thread Stats   Avg      p50      p90      p99    p99.9
    Latency   35.73ms   24.50ms   74.50ms   155.20ms    1.71s

  1941611 requests in 5.00s, 1925483 responses
  Throughput: 384.94K req/s
  Bandwidth:  5.68GB/s
  Status codes: 2xx=1925483, 3xx=0, 4xx=0, 5xx=0
  Latency samples: 1925483 / 1925483 responses (100.0%)
  Per-template: 98977,99411,98743,97094,96370,97238,95535,94959,93909,94988,100644,100169,93224,91715,97113,96928,88150,95178,95795,99343
  Per-template-ok: 98977,99411,98743,97094,96370,97238,95535,94959,93909,94988,100644,100169,93224,91715,97113,96928,88150,95178,95795,99343
  CPU: 8427.3% | Mem: 4.6GiB

=== Best: 402083 req/s (CPU: 9023.9%, Mem: 3.5GiB) ===
  Input BW: 21.86MB/s (avg template: 57 bytes)
[dry-run] Results not saved (use --save to persist)
httparena-bench-genhttp
httparena-bench-genhttp

==============================================
=== genhttp / async-db / 512c (p=1, r=0, cpu=unlimited) ===
==============================================
26fcd84fb49535739b621ebf84f0e24a24b9b04340ca3f236de6e4dfa21c4854
[wait] Waiting for server...
[ready] Server is up

[run 1/3]
gcannon — io_uring HTTP load generator
  Target:    localhost:8080/async-db?min=10&max=50
  Threads:   64
  Conns:     512 (8/thread)
  Pipeline:  1
  Req/conn:  unlimited (keep-alive)
  Expected:  200
  Duration:  5s


  Thread Stats   Avg      p50      p90      p99    p99.9
    Latency   17.04ms   7.71ms   23.00ms   90.70ms    1.64s

  150275 requests in 5.00s, 150275 responses
  Throughput: 30.04K req/s
  Bandwidth:  237.77MB/s
  Status codes: 2xx=150275, 3xx=0, 4xx=0, 5xx=0
  Latency samples: 150275 / 150275 responses (100.0%)
  CPU: 5905.2% | Mem: 922.7MiB

[run 2/3]
gcannon — io_uring HTTP load generator
  Target:    localhost:8080/async-db?min=10&max=50
  Threads:   64
  Conns:     512 (8/thread)
  Pipeline:  1
  Req/conn:  unlimited (keep-alive)
  Expected:  200
  Duration:  5s


  Thread Stats   Avg      p50      p90      p99    p99.9
    Latency   2.82ms   2.25ms   5.47ms   9.43ms   13.30ms

  678753 requests in 5.00s, 678753 responses
  Throughput: 135.69K req/s
  Bandwidth:  1.05GB/s
  Status codes: 2xx=678753, 3xx=0, 4xx=0, 5xx=0
  Latency samples: 678751 / 678753 responses (100.0%)
  CPU: 3784.4% | Mem: 932.7MiB

[run 3/3]
gcannon — io_uring HTTP load generator
  Target:    localhost:8080/async-db?min=10&max=50
  Threads:   64
  Conns:     512 (8/thread)
  Pipeline:  1
  Req/conn:  unlimited (keep-alive)
  Expected:  200
  Duration:  5s


  Thread Stats   Avg      p50      p90      p99    p99.9
    Latency   2.70ms   2.18ms   5.19ms   9.26ms   13.20ms

  691204 requests in 5.00s, 691204 responses
  Throughput: 138.19K req/s
  Bandwidth:  1.07GB/s
  Status codes: 2xx=691204, 3xx=0, 4xx=0, 5xx=0
  Latency samples: 691201 / 691204 responses (100.0%)
  CPU: 3825.3% | Mem: 420.1MiB

=== Best: 138240 req/s (CPU: 3825.3%, Mem: 420.1MiB) ===
[dry-run] Results not saved (use --save to persist)
httparena-bench-genhttp
httparena-bench-genhttp

==============================================
=== genhttp / async-db / 1024c (p=1, r=0, cpu=unlimited) ===
==============================================
de0cdfc11588df7634dcd1e31cf16162d230bd60f44578a5d6437ce45830ab51
[wait] Waiting for server...
[ready] Server is up

[run 1/3]
gcannon — io_uring HTTP load generator
  Target:    localhost:8080/async-db?min=10&max=50
  Threads:   64
  Conns:     1024 (16/thread)
  Pipeline:  1
  Req/conn:  unlimited (keep-alive)
  Expected:  200
  Duration:  5s


  Thread Stats   Avg      p50      p90      p99    p99.9
    Latency   36.01ms   17.40ms   36.40ms   112.70ms    2.27s

  143685 requests in 5.00s, 143685 responses
  Throughput: 28.72K req/s
  Bandwidth:  227.35MB/s
  Status codes: 2xx=143685, 3xx=0, 4xx=0, 5xx=0
  Latency samples: 143685 / 143685 responses (100.0%)
  CPU: 6756.3% | Mem: 858.8MiB

[run 2/3]
gcannon — io_uring HTTP load generator
  Target:    localhost:8080/async-db?min=10&max=50
  Threads:   64
  Conns:     1024 (16/thread)
  Pipeline:  1
  Req/conn:  unlimited (keep-alive)
  Expected:  200
  Duration:  5s


  Thread Stats   Avg      p50      p90      p99    p99.9
    Latency   6.07ms   5.02ms   11.10ms   19.80ms   41.00ms

  731883 requests in 5.00s, 731883 responses
  Throughput: 146.31K req/s
  Bandwidth:  1.13GB/s
  Status codes: 2xx=731883, 3xx=0, 4xx=0, 5xx=0
  Latency samples: 731883 / 731883 responses (100.0%)
  CPU: 4972.6% | Mem: 638.4MiB

[run 3/3]
gcannon — io_uring HTTP load generator
  Target:    localhost:8080/async-db?min=10&max=50
  Threads:   64
  Conns:     1024 (16/thread)
  Pipeline:  1
  Req/conn:  unlimited (keep-alive)
  Expected:  200
  Duration:  5s


  Thread Stats   Avg      p50      p90      p99    p99.9
    Latency   5.39ms   4.74ms   9.84ms   13.80ms   16.40ms

  802821 requests in 5.00s, 802821 responses
  Throughput: 160.50K req/s
  Bandwidth:  1.24GB/s
  Status codes: 2xx=802821, 3xx=0, 4xx=0, 5xx=0
  Latency samples: 802820 / 802821 responses (100.0%)
  CPU: 4392.3% | Mem: 637.9MiB

=== Best: 160564 req/s (CPU: 4392.3%, Mem: 637.9MiB) ===
[dry-run] Results not saved (use --save to persist)
httparena-bench-genhttp
httparena-bench-genhttp
[skip] genhttp does not subscribe to baseline-h2
[skip] genhttp does not subscribe to static-h2
[skip] genhttp does not subscribe to baseline-h3
[skip] genhttp does not subscribe to static-h3
[skip] genhttp does not subscribe to unary-grpc
[skip] genhttp does not subscribe to unary-grpc-tls
[skip] genhttp does not subscribe to echo-ws
httparena-postgres
httparena-postgres
[restore] Restoring CPU governor to performance...

@Kaliumhexacyanoferrat Kaliumhexacyanoferrat marked this pull request as ready for review March 29, 2026 08:14
@BennyFranciscus
Copy link
Copy Markdown
Collaborator

BennyFranciscus commented Mar 29, 2026

@Kaliumhexacyanoferrat Honestly not sure on this one — the upload numbers did come back (589 req/s at 64c) so it's not completely broken, but that CPU usage is wild for that throughput (6829% at 64 connections). Something's definitely off.

I don't have direct visibility into what validation the bench harness checks beyond status codes though. @MDA2AV would know better if there were recent changes to the upload test validation or if something in the SSL/compression changes is interacting badly with the upload path.

The rest of the benchmark looks solid btw — 12.9M pipelined, 643K JSON, 1.2M noisy. Those are really strong numbers for a managed C# framework.

@MDA2AV
Copy link
Copy Markdown
Owner

MDA2AV commented Mar 29, 2026 •
edited
Loading

@BennyFranciscus there is no regression with the current values for the upload test, check the main website https://www.http-arena.com/leaderboard/

@BennyFranciscus
Copy link
Copy Markdown
Collaborator

BennyFranciscus commented Mar 29, 2026

Ah, my bad — checked the leaderboard and yeah, those upload numbers are in line with other .NET entries. No regression. Thanks for the correction.

@MDA2AV MDA2AV merged commit 0132c02 into MDA2AV:main Mar 29, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@MDA2AV MDA2AV MDA2AV approved these changes

@BennyFranciscus BennyFranciscus Awaiting requested review from BennyFranciscus

Assignees

@Kaliumhexacyanoferrat Kaliumhexacyanoferrat

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@Kaliumhexacyanoferrat @BennyFranciscus @MDA2AV