mirror of
https://github.com/zap-proto/fiber.git
synced 2026-10-30 15:06:23 +00:00
d94425b64f
Prove the specificity-routing fork's performance on a representative ~70-route Hanzo cloud table (static, param, multi-param, wildcard). - router_bench_test.go: match hot-path (app.next, 0-alloc), full dispatch, registration scaling (N=10/70/500/1000), and a Go 1.22 net/http.ServeMux routing + registration baseline, with precedence-correctness gates (static/param beat wildcard; fork agrees with ServeMux on every request). - bench/: separate module (own go.mod, replace ../) that isolates the upstream gofiber/fiber v3.2.0 dependency — fork vs upstream match (per-route + aggregate sweep) and registration. Never pollutes the fork's own module. - middleware/static: embed.FS serve benchmark (small 1 KiB + medium 64 KiB). - BENCHMARKS.md: methodology (M1 Max, go1.26.4, -benchtime=1s -count=6, benchstat medians) and the real numbers. Findings: the fork adds ZERO aggregate match cost (sweep 10.77us vs upstream 10.96us; per-route latency is a scan-order permutation, not a regression); fiber routing is 0-alloc vs ServeMux's 1-5 allocs per dynamic route; registration is O(n^2) (9.5ms at N=1000/method, <110us at N<=70) — the one honest regression, and fixable with an index-based conflict check. Co-authored-by: Hanzo Dev <[email protected]>
73 lines
2.0 KiB
Go
73 lines
2.0 KiB
Go
// ⚡️ Fiber is an Express inspired web framework written in Go with ☕️
|
|
// 🤖 GitHub Repository: https://github.com/gofiber/fiber
|
|
// 📌 API Documentation: https://docs.gofiber.io
|
|
|
|
package static
|
|
|
|
// Static-serving benchmarks for the embed.FS path: the ZAP web stack ships
|
|
// assets straight out of a compiled-in embed.FS (no disk touch), so this
|
|
// measures the middleware's per-request cost (path sanitize → open → detect
|
|
// content-type → stream) plus the fixed + per-byte serve cost for a small
|
|
// (1 KiB) and a medium (64 KiB) file.
|
|
//
|
|
// Served through app.Handler() on a reused *fasthttp.RequestCtx — the same
|
|
// low-overhead idiom as the router benchmarks — so the numbers are the serve
|
|
// path, not net/http↔fasthttp conversion.
|
|
//
|
|
// Run:
|
|
//
|
|
// go test -run='^$' -bench='Benchmark_Static_Serve' -benchmem -count=6 ./middleware/static
|
|
//
|
|
// Numbers and methodology: BENCHMARKS.md (fiber).
|
|
|
|
import (
|
|
"embed"
|
|
"testing"
|
|
|
|
"github.com/valyala/fasthttp"
|
|
|
|
"github.com/zap-proto/fiber/v3"
|
|
)
|
|
|
|
//go:embed testdata/small.txt testdata/medium.txt
|
|
var benchFS embed.FS
|
|
|
|
var staticServeFiles = []struct {
|
|
name string
|
|
path string
|
|
size int
|
|
}{
|
|
{"small_1KiB", "/testdata/small.txt", 1024},
|
|
{"medium_64KiB", "/testdata/medium.txt", 65536},
|
|
}
|
|
|
|
// Benchmark_Static_Serve serves a small and a medium file from embed.FS.
|
|
func Benchmark_Static_Serve(b *testing.B) {
|
|
app := fiber.New()
|
|
app.Get("/*", New("", Config{FS: benchFS}))
|
|
h := app.Handler()
|
|
|
|
for _, f := range staticServeFiles {
|
|
b.Run(f.name, func(b *testing.B) {
|
|
fctx := &fasthttp.RequestCtx{}
|
|
fctx.Request.Header.SetMethod(fiber.MethodGet)
|
|
fctx.URI().SetPath(f.path)
|
|
|
|
// Correctness gate: 200 with the full file body.
|
|
h(fctx)
|
|
if sc := fctx.Response.StatusCode(); sc != fiber.StatusOK {
|
|
b.Fatalf("%s: status %d, want 200", f.path, sc)
|
|
}
|
|
if got := len(fctx.Response.Body()); got != f.size {
|
|
b.Fatalf("%s: served %d bytes, want %d", f.path, got, f.size)
|
|
}
|
|
|
|
b.SetBytes(int64(f.size))
|
|
b.ReportAllocs()
|
|
for b.Loop() {
|
|
h(fctx)
|
|
}
|
|
})
|
|
}
|
|
}
|