go http 架構性能大幅下降原因分析

最近在開發一個web 架構，然後業務方使用過程中，跟我們說，壓測qps 上不去，我就很納悶，httprouter + net/http.httpserver , 性能不可能這麼差啊，網上的壓測結果都是10w qps 以上，幾個middleware 至于将性能拖垮？後來一番排查，發現些有意思的東西。

首先，我就簡單壓測hello world, 每個請求進來，我日志都不打，然後，打開pprof ,顯示的情況如下：

這裡futex 怎麼這麼高？看着上面的一些操作，addtimer, deltimer 我想到以前的自己實作的定時器，這估計是逾時引起的。然後檢查版本，go1.9,  然後架構預設為每個conn 設定了4個timeout，readtimeout, writetimeout, idletimeout, headertimeout ，這直接導緻了定時器在添加和删除回調的時候，鎖的壓力特别大。

下面我們分析下，正常的加逾時操作，到底發生了些什麼，下面是個最簡單的例子，為了安全，每個連接配接設定逾時。

package main

import (
	"fmt"
	"github.com/julienschmidt/httprouter"
	"log"
	"net/http"
	"time"
)

func Index(w http.ResponseWriter, r *http.Request, _ httprouter.Params) {
	fmt.Fprint(w, "Welcome!\n")
}

func Hello(w http.ResponseWriter, r *http.Request, ps httprouter.Params) {
	fmt.Fprintf(w, "hello, %s!\n", ps.ByName("name"))
}

func main() {

	router := httprouter.New()
	router.GET("/", Index)
	router.GET("/hello/:name", Hello)

	srv := &http.Server{
		ReadTimeout:       5 * time.Second,
		WriteTimeout:      10 * time.Second,
		ReadHeaderTimeout: 10 * time.Second,
		IdleTimeout:       10 * time.Second,
		Addr:              "0.0.0.0:8998",
		Handler:           router,
	}

	log.Fatal(srv.ListenAndServe())
}           

其中，ListenAndServe() 在調用accept 每個連接配接後，會調用 server.serve(), 根據是否添加逾時，調用conn.SetReadDeadline等函數，對應的是 net/http/server.go,如下：

// Serve a new connection.
func (c *conn) serve(ctx context.Context) {
	...

	if tlsConn, ok := c.rwc.(*tls.Conn); ok {
		if d := c.server.ReadTimeout; d != 0 {
			c.rwc.SetReadDeadline(time.Now().Add(d)) // 設定讀逾時
		}
		if d := c.server.WriteTimeout; d != 0 {
			c.rwc.SetWriteDeadline(time.Now().Add(d))// 設定寫逾時
		}
		if err := tlsConn.Handshake(); err != nil {
			c.server.logf("http: TLS handshake error from %s: %v", c.rwc.RemoteAddr(), err)
			return
		}
		c.tlsState = new(tls.ConnectionState)
		*c.tlsState = tlsConn.ConnectionState()
		if proto := c.tlsState.NegotiatedProtocol; validNPN(proto) {
			if fn := c.server.TLSNextProto[proto]; fn != nil {
				h := initNPNRequest{tlsConn, serverHandler{c.server}}
				fn(c.server, tlsConn, h)
			}
			return
		}
	}
   ...           

之後，con.SetReadDeadline 會調用 internal/poll/fd_poll_runtime.go的 fd.setReadDeadline，最後調用runtime/netpoll.go 的poll_runtime_pollSetDeadline， 這個函數會連結成internal/poll.runtime_pollSetDeadline。這個函數比較關鍵：

//go:linkname poll_runtime_pollSetDeadline internal/poll.runtime_pollSetDeadline
func poll_runtime_pollSetDeadline(pd *pollDesc, d int64, mode int) {
	lock(&pd.lock)
	if pd.closing {
		unlock(&pd.lock)
		return
	}
	pd.seq++ // invalidate current timers
	// Reset current timers.
	if pd.rt.f != nil {
		deltimer(&pd.rt)
		pd.rt.f = nil
	}
	if pd.wt.f != nil {
		deltimer(&pd.wt)
		pd.wt.f = nil
	}
	// Setup new timers.
	if d != 0 && d <= nanotime() {
		d = -1
	}
	if mode == 'r' || mode == 'r'+'w' {
		pd.rd = d
	}
	if mode == 'w' || mode == 'r'+'w' {
		pd.wd = d
	}
	if pd.rd > 0 && pd.rd == pd.wd {
		pd.rt.f = netpollDeadline
		pd.rt.when = pd.rd
		// Copy current seq into the timer arg.
		// Timer func will check the seq against current descriptor seq,
		// if they differ the descriptor was reused or timers were reset.
		pd.rt.arg = pd
		pd.rt.seq = pd.seq
		addtimer(&pd.rt)
	} else {
		if pd.rd > 0 {
			pd.rt.f = netpollReadDeadline // 設定讀的定時回調
			pd.rt.when = pd.rd
			pd.rt.arg = pd
			pd.rt.seq = pd.seq
			addtimer(&pd.rt)             // 添加到系統定時器中
		}
		if pd.wd > 0 {
			pd.wt.f = netpollWriteDeadline // 設定寫的定時回調
			pd.wt.when = pd.wd
			pd.wt.arg = pd
			pd.wt.seq = pd.seq
			addtimer(&pd.wt)             // 添加到系統定時器中
		}
	}
	// If we set the new deadline in the past, unblock currently pending IO if any.
	var rg, wg *g
	atomicstorep(unsafe.Pointer(&wg), nil) // full memory barrier between stores to rd/wd and load of rg/wg in netpollunblock
	if pd.rd < 0 {
		rg = netpollunblock(pd, 'r', false)
	}
	if pd.wd < 0 {
		wg = netpollunblock(pd, 'w', false)
	}
	unlock(&pd.lock)
	if rg != nil {
		netpollgoready(rg, 3)
	}
	if wg != nil {
		netpollgoready(wg, 3)
	}
}           

這裡主要工作就是檢查過期定時器，然後添加定時器，設定回調函數為netpollReadDeadline 或者netpollWriteDeadline。 從中可以看出添加和删除定時器操作為addtimer(&pd.rt), deltimer(&pd.rt)。

後面就是核心了，為啥加逾時後這麼慢，看下addtimer 的實作，timer 是個四叉小頂堆，每次添加一個逾時，最後都需要對一個全局的timers 進行加鎖，當qps 很高，一個請求，多次加鎖，這性能能很高嗎？

type timer struct {
	i int // heap index

	// Timer wakes up at when, and then at when+period, ... (period > 0 only)
	// each time calling f(arg, now) in the timer goroutine, so f must be
	// a well-behaved function and not block.
	when   int64
	period int64
	f      func(interface{}, uintptr)
	arg    interface{}
	seq    uintptr
}

var timers struct {
	lock         mutex
	gp           *g
	created      bool
	sleeping     bool
	rescheduling bool
	sleepUntil   int64
	waitnote     note
	t            []*timer
}

//添加一個定時器

func addtimer(t *timer) {
	lock(&timers.lock)
	addtimerLocked(t)
	unlock(&timers.lock)
}           

解決鎖沖突改怎麼辦？分段鎖是很常見一個思路，在go1.10 後，timers 由一個，變成64個，定時器被打散到64個鎖上去，自然鎖沖突就降低了。看1.10的runtime/time.go 可以發現定義如下，每個p有單獨的timer， 每個timer能被多個p使用：

// Package time knows the layout of this structure.
// If this struct changes, adjust ../time/sleep.go:/runtimeTimer.
// For GOOS=nacl, package syscall knows the layout of this structure.
// If this struct changes, adjust ../syscall/net_nacl.go:/runtimeTimer.
type timer struct {
	tb *timersBucket // the bucket the timer lives in
	i  int           // heap index

	// Timer wakes up at when, and then at when+period, ... (period > 0 only)
	// each time calling f(arg, now) in the timer goroutine, so f must be
	// a well-behaved function and not block.
	when   int64
	period int64
	f      func(interface{}, uintptr)
	arg    interface{}
	seq    uintptr
}

// timersLen is the length of timers array.
//
// Ideally, this would be set to GOMAXPROCS, but that would require
// dynamic reallocation
//
// The current value is a compromise between memory usage and performance
// that should cover the majority of GOMAXPROCS values used in the wild.
const timersLen = 64 //64個bucket

// timers contains "per-P" timer heaps.
//
// Timers are queued into timersBucket associated with the current P,
// so each P may work with its own timers independently of other P instances.
//
// Each timersBucket may be associated with multiple P
// if GOMAXPROCS > timersLen.
var timers [timersLen]struct {
	timersBucket

	// The padding should eliminate false sharing
	// between timersBucket values.
	pad [sys.CacheLineSize - unsafe.Sizeof(timersBucket{})%sys.CacheLineSize]byte
}           

下面是go1.10 後的timer 資料結構（此圖來源于網絡）：

總結，網上很多httpserver 架構壓測 qps 很高，但是它們的demo并沒有設定逾時，資料真實值會差很多。線上如果需要設定逾時，需要注意go 的版本，qps 很高的情況下，最好使用1.10以上。最終我們不做任何其他操作情況下，僅将go 版本提高到1.10，qps 提高接近2倍。

本文來自雲栖社群合作夥伴“開源中國”

本文作者：鼎銘 

原文連結