Sync HttpClient Send by ManickaP · Pull Request #34948 · dotnet/runtime (original) (raw)
Results from our micro benchamarks, pr vs master:
BenchmarkDotNet=v0.12.1, OS=manjaro Intel Core i7-4771 CPU 3.50GHz (Haswell), 1 CPU, 8 logical and 4 physical cores .NET Core SDK=5.0.100-preview.6.20310.4 [Host] : .NET Core 5.0.0 (CoreCLR 5.0.20.30506, CoreFX 5.0.20.30506), X64 RyuJIT Job-DJGXQP : .NET Core 5.0 (CoreCLR 42.42.42.42424, CoreFX 42.42.42.42424), X64 RyuJIT Job-LKRKYA : .NET Core 5.0 (CoreCLR 42.42.42.42424, CoreFX 42.42.42.42424), X64 RyuJIT
PowerPlanMode=00000000-0000-0000-0000-000000000000 Arguments=/p:DebugType=portable IterationTime=250.0000 ms
MaxIterationCount=20 MinIterationCount=15 WarmupCount=1
| Method | Job | Toolchain | ssl | chunkedResponse | responseLength | Mean | Error | StdDev | Median | Min | Max | Ratio | RatioSD | Gen 0 | Gen 1 | Gen 2 | Allocated |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Get | Job-DJGXQP | /master/corerun/corerun | False | False | 1 | 51.68 μs | 1.355 μs | 1.506 μs | 51.45 μs | 49.07 μs | 54.62 μs | 1.00 | 0.00 | 0.4211 | - | - | 2.02 KB |
| Get | Job-LKRKYA | /sync_pr/corerun/corerun | False | False | 1 | 51.58 μs | 1.601 μs | 1.844 μs | 51.41 μs | 48.07 μs | 54.85 μs | 1.00 | 0.05 | 0.4006 | - | - | 2.05 KB |
| Get | Job-DJGXQP | /master/corerun/corerun | False | False | 100000 | 84.33 μs | 1.019 μs | 0.954 μs | 84.42 μs | 82.25 μs | 85.68 μs | 1.00 | 0.00 | 0.3359 | - | - | 2.04 KB |
| Get | Job-LKRKYA | /sync_pr/corerun/corerun | False | False | 100000 | 85.54 μs | 1.084 μs | 0.906 μs | 85.39 μs | 84.35 μs | 87.48 μs | 1.01 | 0.01 | 0.4390 | - | - | 2.07 KB |
| Get | Job-DJGXQP | /master/corerun/corerun | False | True | 1 | 51.74 μs | 1.130 μs | 1.256 μs | 51.67 μs | 48.91 μs | 54.08 μs | 1.00 | 0.00 | 0.4246 | - | - | 2.01 KB |
| Get | Job-LKRKYA | /sync_pr/corerun/corerun | False | True | 1 | 52.02 μs | 0.992 μs | 0.974 μs | 52.22 μs | 50.49 μs | 53.54 μs | 1.01 | 0.02 | 0.4214 | - | - | 2.04 KB |
| Get | Job-DJGXQP | /master/corerun/corerun | False | True | 100000 | 113.25 μs | 0.796 μs | 0.744 μs | 113.25 μs | 112.06 μs | 114.50 μs | 1.00 | 0.00 | 0.4619 | - | - | 2.02 KB |
| Get | Job-LKRKYA | /sync_pr/corerun/corerun | False | True | 100000 | 114.99 μs | 0.997 μs | 0.833 μs | 115.06 μs | 113.38 μs | 116.15 μs | 1.01 | 0.01 | 0.4615 | - | - | 2.05 KB |
| Get | Job-DJGXQP | /master/corerun/corerun | True | False | 1 | 242.44 μs | 34.959 μs | 38.857 μs | 237.31 μs | 177.12 μs | 336.79 μs | 1.00 | 0.00 | - | - | - | 5.23 KB |
| Get | Job-LKRKYA | /sync_pr/corerun/corerun | True | False | 1 | 243.98 μs | 29.132 μs | 29.917 μs | 244.62 μs | 187.99 μs | 316.79 μs | 1.02 | 0.22 | - | - | - | 5.26 KB |
| Get | Job-DJGXQP | /master/corerun/corerun | True | False | 100000 | 674.75 μs | 254.020 μs | 249.481 μs | 631.25 μs | 387.43 μs | 1,163.31 μs | 1.00 | 0.00 | - | - | - | 71.29 KB |
| Get | Job-LKRKYA | /sync_pr/corerun/corerun | True | False | 100000 | 513.09 μs | 146.182 μs | 156.413 μs | 440.12 μs | 346.33 μs | 737.11 μs | 0.91 | 0.44 | - | - | - | 5.27 KB |
| Get | Job-DJGXQP | /master/corerun/corerun | True | True | 1 | 220.30 μs | 37.672 μs | 41.872 μs | 228.83 μs | 149.07 μs | 282.84 μs | 1.00 | 0.00 | - | - | - | 5.22 KB |
| Get | Job-LKRKYA | /sync_pr/corerun/corerun | True | True | 1 | 225.02 μs | 38.637 μs | 44.495 μs | 204.35 μs | 175.35 μs | 307.61 μs | 1.05 | 0.16 | - | - | - | 5.25 KB |
| Get | Job-DJGXQP | /master/corerun/corerun | True | True | 100000 | 548.13 μs | 149.034 μs | 146.372 μs | 507.55 μs | 328.93 μs | 958.75 μs | 1.00 | 0.00 | - | - | - | 5.22 KB |
| Get | Job-LKRKYA | /sync_pr/corerun/corerun | True | True | 100000 | 507.01 μs | 102.817 μs | 110.013 μs | 479.67 μs | 353.45 μs | 681.92 μs | 0.99 | 0.29 | - | - | - | 5.25 KB |
I ran it several times. Sometimes it would show a regression in one test, other times in another. When I isolated them, I got somewhat unstable results (sometimes was my PR faster, sometimes master). Due to this, I don't think there's a perf regression per se.