Skip to content

HTTP Local Performance

Measurements come from independent local Release loopback processes, not CI.

Metric definitions and methodology.

Average process CPU uses 100% for one logical CPU and can exceed 100% with multiple threads; estimated values are labeled, and incomplete sets of three CPU samples show N/A. On this 14-logical-CPU host, machine share = process CPU% / 14; 100% of one CPU is 7.14% of the machine.

No added computation

Throughput

p99 RTT

All Measured Profiles

64 B / 1 clients / window=1

Dependency / model Round trips/s MiB/s p50 / us p95 / us p99 / us CPU / % RSS / MiB
Boost.Asio / 2 contexts / 2 threads 20,995 [20,801, 21,074] 2.56 46.88 52.79 59.25 51.2 (estimated) 7.77
Boost.Asio / 4 contexts / 4 threads 20,962 [20,945, 20,976] 2.56 47.12 53.00 60.25 51.1 (estimated) 7.83
Boost.Asio / 1 context / 1 thread 20,693 [20,583, 20,796] 2.53 47.75 54.79 62.75 42.8 (estimated) 7.78
Boost.Asio / 1 context / 2 threads 31,658 [31,371, 31,725] 3.86 31.12 36.17 43.25 81.1 (estimated) 7.80
Boost.Asio / 1 context / 4 threads 28,762 [28,526, 29,111] 3.51 33.08 41.62 58.00 142.9 (estimated) 7.86
Standalone Asio / 2 contexts / 2 threads 20,882 [20,775, 21,023] 2.55 47.71 53.54 61.58 51.5 (estimated) 8.25
Standalone Asio / 4 contexts / 4 threads 20,776 [20,759, 20,865] 2.54 47.62 53.08 60.54 51.3 (estimated) 8.30
Standalone Asio / 1 context / 1 thread 20,587 [20,212, 20,639] 2.51 47.83 55.42 64.42 43.0 (estimated) 8.22
Standalone Asio / 1 context / 2 threads 31,463 [31,352, 31,787] 3.84 31.08 37.42 50.12 81.7 (estimated) 8.25
Standalone Asio / 1 context / 4 threads 28,458 [28,360, 28,715] 3.47 33.38 43.46 65.96 142.5 (estimated) 8.30

1024 B / 1 clients / window=1

Dependency / model Round trips/s MiB/s p50 / us p95 / us p99 / us CPU / % RSS / MiB
Boost.Asio / 2 contexts / 2 threads 13,381 [13,318, 13,393] 26.13 73.08 81.75 89.92 38.6 (estimated) 7.75
Boost.Asio / 4 contexts / 4 threads 13,412 [13,353, 13,481] 26.20 73.50 81.50 91.04 38.8 (estimated) 7.81
Boost.Asio / 1 context / 1 thread 13,348 [13,299, 13,361] 26.07 73.50 82.42 91.17 34.2 (estimated) 7.72
Boost.Asio / 1 context / 2 threads 34,988 [34,841, 35,004] 68.34 27.54 31.83 37.79 90.9 (estimated) 7.78
Boost.Asio / 1 context / 4 threads 29,079 [28,901, 29,447] 56.80 32.29 40.21 56.62 146.8 (estimated) 7.84
Standalone Asio / 2 contexts / 2 threads 13,407 [13,271, 13,415] 26.19 73.50 81.79 91.62 38.8 (estimated) 8.23
Standalone Asio / 4 contexts / 4 threads 13,354 [13,319, 13,398] 26.08 72.92 81.96 92.29 38.7 (estimated) 8.27
Standalone Asio / 1 context / 1 thread 13,246 [12,259, 13,328] 25.87 73.58 84.33 98.67 34.5 (estimated) 8.22
Standalone Asio / 1 context / 2 threads 35,683 [32,269, 35,751] 69.69 27.00 31.88 40.29 91.5 (estimated) 8.25
Standalone Asio / 1 context / 4 threads 29,167 [28,813, 29,650] 56.97 32.00 42.42 62.21 148.8 (estimated) 8.33

16384 B / 1 clients / window=1

Dependency / model Round trips/s MiB/s p50 / us p95 / us p99 / us CPU / % RSS / MiB
Boost.Asio / 2 contexts / 2 threads 2,038 [2,036, 2,081] 63.68 478.21 519.04 541.12 22.6 (estimated) 7.78
Boost.Asio / 4 contexts / 4 threads 2,012 [2,001, 2,022] 62.87 483.42 553.50 572.62 22.9 (estimated) 7.84
Boost.Asio / 1 context / 1 thread 2,033 [2,016, 2,065] 63.52 478.54 530.75 560.38 21.6 (estimated) 7.70
Boost.Asio / 1 context / 2 threads 16,158 [16,042, 16,388] 504.92 50.04 56.83 66.21 105.1 (estimated) 7.81
Boost.Asio / 1 context / 4 threads 14,924 [14,919, 14,928] 466.36 55.75 66.42 87.83 127.1 (estimated) 7.89
Standalone Asio / 2 contexts / 2 threads 2,041 [2,015, 2,052] 63.77 472.83 542.00 564.58 22.7 (estimated) 8.22
Standalone Asio / 4 contexts / 4 threads 2,027 [2,026, 2,062] 63.35 476.50 550.17 568.75 22.8 (estimated) 8.31
Standalone Asio / 1 context / 1 thread 2,026 [1,997, 2,037] 63.32 478.21 537.83 568.38 22.2 (estimated) 8.19
Standalone Asio / 1 context / 2 threads 16,855 [16,187, 17,038] 526.70 47.04 56.79 74.46 105.8 (estimated) 8.27
Standalone Asio / 1 context / 4 threads 15,105 [14,617, 15,436] 472.03 54.29 67.46 89.88 126.5 (estimated) 8.38

64 B / 16 clients / window=16

Dependency / model Round trips/s MiB/s p50 / us p95 / us p99 / us CPU / % RSS / MiB
Boost.Asio / 2 contexts / 2 threads 217,845 [208,717, 225,168] 26.59 1,124.12 1,446.08 2,244.92 198.8 (estimated) 8.69
Boost.Asio / 4 contexts / 4 threads 285,384 [270,209, 294,985] 34.84 845.04 1,288.50 1,343.92 389.4 (estimated) 9.08
Boost.Asio / 1 context / 1 thread 150,384 [150,297, 151,671] 18.36 1,685.38 1,794.71 1,864.75 99.8 (estimated) 8.62
Boost.Asio / 1 context / 2 threads 182,970 [172,894, 187,171] 22.34 1,356.04 1,620.38 1,861.00 193.8 (estimated) 8.61
Boost.Asio / 1 context / 4 threads 209,663 [207,223, 212,841] 25.59 1,184.83 1,504.58 1,766.62 367.7 (estimated) 8.81
Standalone Asio / 2 contexts / 2 threads 229,220 [217,599, 236,317] 27.98 1,074.96 1,366.29 2,174.08 199.0 (estimated) 9.08
Standalone Asio / 4 contexts / 4 threads 284,782 [283,663, 287,910] 34.76 850.96 1,281.04 1,323.83 388.4 (estimated) 9.48
Standalone Asio / 1 context / 1 thread 157,694 [156,776, 157,773] 19.25 1,600.75 1,758.67 1,943.29 99.6 (estimated) 9.05
Standalone Asio / 1 context / 2 threads 191,849 [181,795, 195,437] 23.42 1,307.17 1,533.33 1,845.04 195.2 (estimated) 9.09
Standalone Asio / 1 context / 4 threads 201,333 [169,613, 216,029] 24.58 1,180.79 1,738.42 2,214.46 365.6 (estimated) 9.23

1024 B / 16 clients / window=16

Dependency / model Round trips/s MiB/s p50 / us p95 / us p99 / us CPU / % RSS / MiB
Boost.Asio / 2 contexts / 2 threads 180,211 [178,797, 180,583] 351.97 1,388.33 1,565.25 2,110.00 198.7 (estimated) 9.31
Boost.Asio / 4 contexts / 4 threads 230,485 [224,878, 238,748] 450.17 1,035.21 1,526.50 1,616.92 385.7 (estimated) 9.72
Boost.Asio / 1 context / 1 thread 101,056 [95,303, 101,646] 197.37 2,559.38 2,709.83 2,799.08 95.1 (estimated) 9.30
Boost.Asio / 1 context / 2 threads 147,334 [146,812, 148,043] 287.76 1,708.54 1,929.29 2,127.54 195.3 (estimated) 9.30
Boost.Asio / 1 context / 4 threads 161,784 [161,158, 165,824] 315.98 1,510.04 1,991.67 2,180.88 368.3 (estimated) 9.44
Standalone Asio / 2 contexts / 2 threads 180,744 [162,565, 184,333] 353.01 1,361.21 1,642.54 2,786.17 197.6 (estimated) 9.77
Standalone Asio / 4 contexts / 4 threads 234,535 [211,937, 238,696] 458.08 1,010.50 1,500.88 1,646.79 384.7 (estimated) 11.45
Standalone Asio / 1 context / 1 thread 98,353 [95,808, 99,450] 192.10 2,550.92 2,760.71 2,951.96 88.1 (estimated) 9.72
Standalone Asio / 1 context / 2 threads 149,186 [148,368, 149,892] 291.38 1,674.00 1,942.00 2,162.96 195.1 (estimated) 9.78
Standalone Asio / 1 context / 4 threads 158,108 [157,716, 166,065] 308.81 1,510.58 2,068.29 2,376.75 367.0 (estimated) 9.81

16384 B / 16 clients / window=16

Dependency / model Round trips/s MiB/s p50 / us p95 / us p99 / us CPU / % RSS / MiB
Boost.Asio / 2 contexts / 2 threads 32,430 [32,127, 32,622] 1,013.44 7,787.88 8,518.17 8,843.83 146.0 (estimated) 20.45
Boost.Asio / 4 contexts / 4 threads 39,021 [38,133, 39,200] 1,219.40 6,484.25 6,962.62 7,363.12 241.6 (estimated) 20.61
Boost.Asio / 1 context / 1 thread 18,309 [18,257, 18,410] 572.17 13,937.88 14,193.54 14,276.83 83.7 (estimated) 20.36
Boost.Asio / 1 context / 2 threads 37,626 [36,596, 37,719] 1,175.82 6,760.71 7,335.50 7,788.46 190.2 (estimated) 20.41
Boost.Asio / 1 context / 4 threads 42,983 [42,951, 43,408] 1,343.21 5,728.08 6,695.92 6,923.58 369.2 (estimated) 20.55
Standalone Asio / 2 contexts / 2 threads 32,485 [31,627, 34,492] 1,015.16 7,881.21 8,252.71 8,569.42 143.7 (estimated) 20.86
Standalone Asio / 4 contexts / 4 threads 39,319 [39,229, 39,451] 1,228.72 6,447.25 6,929.83 7,159.83 240.0 (estimated) 20.88
Standalone Asio / 1 context / 1 thread 18,870 [17,573, 18,997] 589.68 13,520.17 14,473.50 15,043.92 83.0 (estimated) 20.80
Standalone Asio / 1 context / 2 threads 38,472 [36,113, 38,709] 1,202.25 6,612.17 7,079.25 7,517.17 189.9 (estimated) 20.77
Standalone Asio / 1 context / 4 threads 43,545 [42,816, 44,201] 1,360.80 5,738.17 6,652.54 7,347.25 369.8 (estimated) 20.86

Tables show repeated-run medians; throughput brackets show min/max, not confidence intervals. Latencies are medians of per-run sampled percentiles, without pooling samples.

Navigation

Type to search…

↑↓ navigate↵ selectEsc close