Go tool使用记录

Go tool使用记录

1、trace

https://mp.weixin.qq.com/s/ugXT_l-fNHEGANoLvirRLA

示例程序

package main

import (
	"os"
	"runtime/trace"
)

func main() {
	trace.Start(os.Stderr)
	defer trace.Stop()
	var ch1 = make(chan string)

	go func() {
		ch1 <- "hello world"
	}()

	<-ch1

}

生成trace文件

go run main.go. 2>trace.out

使用go tool trace打开

go tool trace trace.out

out

2021/09/01 10:15:26 Parsing trace...
2021/09/01 10:15:26 Splitting trace...
2021/09/01 10:15:26 Opening browser. Trace viewer is listening on http://127.0.0.1:57123

浏览器显示：

解释：

• View trace：查看跟踪
• Goroutine analysis：Goroutine 分析
• Network blocking profile：网络阻塞概况
• Synchronization blocking profile：同步阻塞概况
• Syscall blocking profile：系统调用阻塞概况
• Scheduler latency profile：调度延迟概况
• User defined tasks：用户自定义任务
• User defined regions：用户自定义区域
• Minimum mutator utilization：最低 Mutator 利用率

打开goroutine分析

按理来说，N应该是3，目前还不清楚为什么是4.链接中的演示显示的是3

详细使用看上面的链接

方法二

package main

import (
	"os"
	"runtime/trace"
)

func main() {
	f, err := os.Create("trace1.out")
	if err != nil {
		panic(err)
	}
	defer f.Close()

	err = trace.Start(f)
	if err != nil {
		panic(err)
	}
	defer trace.Stop()
	// trace.Start(os.Stderr)
	// defer trace.Stop()
	ch1 := make(chan string)

	go func() {
		ch1 <- "hello world"
	}()

	<-ch1

}

直接启动就可以得到trace1.out文件

2、pprof

参考

1）查看内存使用情况

示例代码

package main

import (
	"log"
	"net/http"
	_ "net/http/pprof"
	"os"
	"runtime"
	"runtime/trace"
	"time"
)

func readMemStats() {

	var ms runtime.MemStats

	runtime.ReadMemStats(&ms)

	log.Printf(" ===> Alloc:%d(bytes) HeapIdle:%d(bytes) HeapReleased:%d(bytes)", ms.Alloc, ms.HeapIdle, ms.HeapReleased)
}

func test() {
	//slice 会动态扩容，用slice来做堆内存申请
	container := make([]int, 8)

	log.Println(" ===> loop begin.")
	for i := 0; i < 32*1000*1000; i++ {
		container = append(container, i)
		if i == 16*1000*1000 {
			readMemStats()
		}
	}

	log.Println(" ===> loop end.")
}

func main() {
	f, err := os.Create("trace1.out")
	if err != nil {
		panic(err)
	}
	defer f.Close()

	err = trace.Start(f)
	if err != nil {
		panic(err)
	}
	defer trace.Stop()
	//启动pprof
	go func() {
		log.Println(http.ListenAndServe("0.0.0.0:10000", nil))
	}()

	log.Println(" ===> [Start].")

	readMemStats()
	test()
	readMemStats()

	log.Println(" ===> [force gc].")
	runtime.GC() //强制调用gc回收

	log.Println(" ===> [Done].")
	readMemStats()

	go func() {
		for {
			readMemStats()
			time.Sleep(5 * time.Second)
		}
	}()

	time.Sleep(20 * time.Second) //睡眠，保持程序不退出
}

启动

go run main.go

pprof

"打开浏览器"
http://127.0.0.1:10000/debug/pprof/heap?debug=1

就能看到目前的内存状态

代码中38–48的代码是为了程序结束后生成trace.out文件，同样可以使用trace工具查看

2）查看cpu使用

示例代码

package main

import (
    "bytes"
    "math/rand"
    "time"
    "log"
    "net/http"
    _ "net/http/pprof"
)


func test() {

    log.Println(" ===> loop begin.")
    for i := 0; i < 1000; i++ {
        log.Println(genSomeBytes())
    }

    log.Println(" ===> loop end.")
}

//生成一个随机字符串
func genSomeBytes() *bytes.Buffer {

    var buff bytes.Buffer

    for i := 1; i < 20000; i++ {
        buff.Write([]byte{'0' + byte(rand.Intn(10))})
    }

    return &buff
}

func main() {

    go func() {
        for {
            test()
            time.Sleep(time.Second * 1)
        }
    }()

    //启动pprof
    http.ListenAndServe("0.0.0.0:10000", nil)

}

编译后启动

go build .

浏览器打开

http://127.0.0.1:10000/debug/pprof/

点击profile，程序需要运行30s就能得到一个可以下载到本地的二进制文件profile。

使用pprof工具

go tool pprof [binary] [profile]

binary: 必须指向生成这个性能分析数据的那个二进制可执行文件；

profile: 必须是该二进制可执行文件所生成的性能分析数据文件

二者要严格匹配

执行命令后出现：

ype: cpu
Time: Sep 1, 2021 at 2:30pm (CST)
Duration: 10s, Total samples = 1.38s (13.79%)
Entering interactive mode (type "help" for commands, "o" for options)
(pprof) top
Showing nodes accounting for 1230ms, 89.13% of 1380ms total
Showing top 10 nodes out of 86
      flat  flat%   sum%        cum   cum%
     610ms 44.20% 44.20%      610ms 44.20%  syscall.syscall
     140ms 10.14% 54.35%      140ms 10.14%  runtime.nanotime1
     100ms  7.25% 61.59%      100ms  7.25%  runtime.madvise
     100ms  7.25% 68.84%      100ms  7.25%  runtime.usleep
      90ms  6.52% 75.36%       90ms  6.52%  runtime.pthread_cond_wait
      50ms  3.62% 78.99%       50ms  3.62%  sync.(*Mutex).Unlock (inline)
      40ms  2.90% 81.88%       90ms  6.52%  math/rand.(*lockedSource).Int63
      40ms  2.90% 84.78%       40ms  2.90%  runtime.pthread_cond_signal
      30ms  2.17% 86.96%       50ms  3.62%  bytes.(*Buffer).Write
      30ms  2.17% 89.13%       30ms  2.17%  runtime.kevent

可以查看程序中各个函数的耗时和cpu使用情况