HTTP (L7) proxy benchmark · open-loop ramp · c1k

request throughput 20261007-024759

Every proxy driven through the identical linear ramp (zrk, open-loop, coordinated-omission corrected) at 1000 connections. Throughput and CPU span the full ramp. The table's p50 / p99 are each proxy's latency at a common 2k req/s reference load — a shared, light, sub-knee rate where the number reflects per-request cost rather than queueing.

proxy	max sustained req/s	p50 @ 2k	p99 @ 2k	peak mem	status
zoxy0.9.0 @935a2f349	62.6829k	0.35ms	1.00ms	201.2MiB	ok
nginx1.31	28.0336k	0.38ms	0.87ms	61.2MiB	ok
haproxy3.4.6	22.0091k	0.37ms	1.17ms	87.1MiB	ok
pingora0.8	14.8509k	0.39ms	1.85ms	58.8MiB	ok
envoy—	9.1991k	0.53ms	2.32ms	79.0MiB	ok

zoxy compiled for this host's CPU with SIMD (upstream release v0.9.0 x86_64-linux ReleaseSafe cpu=x86_64_v3 (SIMD, sha256-verified, mode read from the artifact)); stock images in this comparison are generic x86-64 builds
zoxy built ReleaseSafe: Zig's bounds/overflow/unreachable checks stay on in the hot path, which measured -12.7% sustained against the same code at ReleaseFast (c1k, 2x2 over v0.1.0/v0.2.1); the other proxies here are built without equivalent checks
zoxy dropped 2052891 access-log lines (16.93% of completed requests) instead of blocking on stdout; the other proxies block or buffer, so this much logging work was skipped here and not by them
nginx latency histogram saturated at 60s; tail percentiles are a floor, not a value
haproxy latency histogram saturated at 60s; tail percentiles are a floor, not a value
pingora compiled for this host's CPU with SIMD (release target-cpu=native); stock images in this comparison are generic x86-64 builds
pingora latency histogram saturated at 60s; tail percentiles are a floor, not a value
envoy latency histogram saturated at 60s; tail percentiles are a floor, not a value

Successful req/s vs offered

open-loop ramp; dashed gray = perfect keep-up

Proxy CPU vs offered

container cores sampled from cAdvisor at 1Hz, on the ramp's own clock

p99 latency vs offered (while keeping up)

per-window tail (log scale); each line stops where that proxy stops keeping up

Load shed vs offered

offered load the proxy couldn't serve (1 − achieved/offered); the first few seconds include the generator's own ramp-up shortfall, which is not the proxy's

Latency distribution · HdrHistogram (whole run)

zoxy

latency by percentile · raw zoxy.hgrm

haproxy

latency by percentile · saturated at 60s — tail percentiles are a floor, not a value · raw haproxy.hgrm

nginx

latency by percentile · saturated at 60s — tail percentiles are a floor, not a value · raw nginx.hgrm

pingora

latency by percentile · saturated at 60s — tail percentiles are a floor, not a value · raw pingora.hgrm

envoy

latency by percentile · saturated at 60s — tail percentiles are a floor, not a value · raw envoy.hgrm