golang 里出现多 goroutine 的场景很常见, 最常用的两种方式就是 WaitGroupContext, 今天我们了解一下 Context 的应用场景

使用场景

场景一: 多goroutine执行超时通知

并发执行的业务中最常见的就是有协程执行超时, 如果不做超时处理就会出现一个僵尸进程, 这累计的多了就会有一阵手忙脚乱了, 所以我们要在源头上就避免它们

看下面这个示例:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
package main

import (
	"context"
	"fmt"
	"time"
)

/**
同一个content可以控制多个goroutine, 确保线程可控, 而不是每新建一个goroutine就要有一个chan去通知他关闭
有了他代码更加简洁
*/

func main() {
	fmt.Println("run demo \n\n\n")
	demo()
}

func demo() {
	ctx, cancel := context.WithTimeout(context.Background(), 9*time.Second)
	go watch(ctx, "[线程1]")
	go watch(ctx, "[线程2]")
	go watch(ctx, "[线程3]")

	index := 0
	for {
		index++
		fmt.Printf("%d 秒过去了 \n", index)
		time.Sleep(1 * time.Second)
		if index > 10 {
			break
		}
	}

    fmt.Println("通知停止监控")
    // 其实此时已经超时, 协程已经提前退出
	cancel()

	// 防止主进程提前退出
	time.Sleep(3 * time.Second)
	fmt.Println("done")
}

func watch(ctx context.Context, name string) {
	for {
		select {
		case <-ctx.Done():
			fmt.Printf("%s  监控退出, 停止了...\n", name)
			return
		default:
			fmt.Printf("%s goroutine监控中... \n", name)
			time.Sleep(2 * time.Second)
		}
	}
}

使用 context.WithTimeout() 给文本流设置一个时间上限, 结合 for+select 去接收消息. 当执行超时,或手动关闭都会给 <-ctx.Done() 发送消息, 而且所有使用同一个 context 都会收到这个通知, 免去了一个一个通知的繁琐代码

场景二: 类似web服务器中的session

比如在php中(没用swoole扩展), 一个请求进来, 从 $_REQUEST $_SERVER 能获取到的是有关这一条请求的所有信息, 哪怕是使用全局变量也是给这一个请求来服务的, 是线程安全的

但是 golang 就不一样了, 因为程序本身就能起一个 web sever, 因此就不能随便使用全局变量了, 不然就是内存泄露警告. 但是实际业务当中需要有一个类似session 的东西来承载单次请求的信息, 举一个具体的例子就是: 给每次请求加一个 uniqueID 该如何处理? 有了这个 uniqueID, 请求的所有日志都能带上它, 这样排查问题的时候方便追踪一次请求发生了什么

如下:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

func demo2() {
	pCtx, pCancel := context.WithCancel(context.Background())
	pCtx = context.WithValue(pCtx, "parentKey", "parentVale")
	go watch(pCtx, "[父进程1]")
	go watch(pCtx, "[父进程2]")

	cCtx, cCancel := context.WithCancel(pCtx)
	go watch(cCtx, "[子进程1]")
	go watch(cCtx, "[子进程2]")
	fmt.Println(pCtx.Value("parentKey"))
	fmt.Println(cCtx.Value("parentKey"))

	time.Sleep(10 * time.Second)
	fmt.Println("子进程关闭")
	cCancel()
	time.Sleep(5 * time.Second)
	fmt.Println("父进程关闭")
	pCancel()

	time.Sleep(3 * time.Second)
	fmt.Println("done")
}

最开始的 context.WithCancel(context.Background())context.Background() 就是一个新建的 context, 利用 context 能继承的特性, 可以将自己的程序构建出一个 context 树, context 执行 cancel() 将影响到当前 context 和子 context, 不会影响到父级.

同时 context.WithValue 也会给 context 带上自定义的值, 这样 uniqueID 就能轻松的传递了下去, 而不是一层层的传递参数, 改func什么的

对于 context 很值得参考的应用有:

Context 相关 func 和接口

继承 context 需要实现如下四个接口

 1
 2
 3
 4
 5
 6
 7
 8
 9
10

type Context interface {
	Deadline() (deadline time.Time, ok bool)

	Done() <-chan struct{}

	Err() error

	Value(key interface{}) interface{}
}

当使用的时候不需要实现接口, 因为官方包里已经基于 emptyCtx 实现了一个, 调用方法有

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
var (
	background = new(emptyCtx)
	todo       = new(emptyCtx)
)

// 这个是最初始的ctx, 之后的子ctx都是继承自它
func Background() Context {
	return background
}

// 不清楚context要干嘛, 但是就得有一个ctx的用这个
func TODO() Context {
	return todo
}

继承用的函数

1
2
3
4
func WithCancel(parent Context) (ctx Context, cancel CancelFunc)
func WithDeadline(parent Context, deadline time.Time) (Context, CancelFunc)
func WithTimeout(parent Context, timeout time.Duration) (Context, CancelFunc)
func WithValue(parent Context, key, val interface{}) Context
  • WithCancel 返回一个带 cancel 函数的ctx,
  • WithDeadline 在到达指定时间时自动执行 cancel()
  • WithTimeoutWithDeadline的壳子, 区别就是这个函数是多少时间过后执行 cancel
1
2
3
func WithTimeout(parent Context, timeout time.Duration) (Context, CancelFunc) {
	return WithDeadline(parent, time.Now().Add(timeout))
}
  • WithValue 继承父类ctx时顺便带上一个值