etcd实现分布式锁

2022-12-11,

转载自:etcd实现分布式

当并发的访问共享资源的时候,如果没有加锁的话,无法保证共享资源安全性和正确性。这个时候就需要用到锁

1、需要具备的特性

    需要保证互斥访问(分布式环境需要保证不同节点、不同线程的互斥访问)
    需要有超时机制,防止锁意外未释放,其他节点无法获取到锁;也要保证任务能够正常执行完成,不能超时了任务还没结束,导致任务执行一般被释放锁
    需要有阻塞和非阻塞两种请求锁的接口

2、本地锁

当业务执行在同一个线程内,也就是我初始化一个本地锁,其他请求也认这把锁。一般是服务部署在单机环境下。

我们可以看下下面的例子,开1000个goroutine并发的给Counter做自增操作,结果会是什么样的呢?

package main

import (
"fmt"
"sync"
) var sg sync.WaitGroup type Counter struct {
count int
} // 自增操作
func (m *Counter) Incr() {
m.count++
} // 获取总数
func (m *Counter) Count() int {
return m.count
} func main() {
c := &Counter{}
for i := 0; i < 1000; i++ {
sg.Add(1)
// 模拟并发请求
go func() {
c.Incr()
sg.Done()
}()
}
sg.Wait() fmt.Println(c.Count())
}

结果是count的数量并不是预想中的1000,而是下面这样,每次打印出的结果都不一样,但是接近1000

user@userdeMacBook-Pro  ~/go/src/go-demo/mutex  go run main.go
953
user@userdeMacBook-Pro  ~/go/src/go-demo/mutex  go run main.go
982
user@userdeMacBook-Pro  ~/go/src/go-demo/mutex  go run main.go
984

出现这个问题的原因就是没有给自增操作加锁

下面我们修改代码如下,在Incr中加上go的mutex互斥锁

package main

import (
"fmt"
"sync"
) var sg sync.WaitGroup type Counter struct {
count int
mu sync.Mutex
} func (m *Counter) Incr() {
// 每次写之前先加锁,写完之后释放锁
m.mu.Lock()
defer m.mu.Unlock()
m.count++
} func (m *Counter) Count() int {
return m.count
} func main() {
c := &Counter{}
for i := 0; i < 1000; i++ {
sg.Add(1)
go func() {
c.Incr()
sg.Done()
}()
}
sg.Wait() fmt.Println(c.Count())
}

可以看到现在count正常输出1000了

user@userdeMacBook-Pro  ~/go/src/go-demo/mutex  go run main.go
1000
user@userdeMacBook-Pro  ~/go/src/go-demo/mutex  go run main.go
1000
user@userdeMacBook-Pro  ~/go/src/go-demo/mutex  go run main.go
1000

3、etcd分布式锁

简单部署一个etcd集群

├── docker-compose.yml
├── etcd
│ └── Dockerfile

Dockerfile文件内容

FROM bitnami/etcd:latest

LABEL maintainer="liuyuede123 <liufutianoppo@163.com>"

Docker-compose.yml内容

version: '3.5'
# 网络配置
networks:
backend:
driver: bridge # 服务容器配置
services:
etcd1: # 自定义容器名称
build:
context: etcd # 指定构建使用的 Dockerfile 文件
environment:
- TZ=Asia/Shanghai
- ALLOW_NONE_AUTHENTICATION=yes
- ETCD_NAME=etcd1
- ETCD_INITIAL_ADVERTISE_PEER_URLS=http://etcd1:2380
- ETCD_LISTEN_PEER_URLS=http://0.0.0.0:2380
- ETCD_LISTEN_CLIENT_URLS=http://0.0.0.0:2379
- ETCD_ADVERTISE_CLIENT_URLS=http://etcd1:2379
- ETCD_INITIAL_CLUSTER_TOKEN=etcd-cluster
- ETCD_INITIAL_CLUSTER=etcd1=http://etcd1:2380,etcd2=http://etcd2:2380,etcd3=http://etcd3:2380
- ETCD_INITIAL_CLUSTER_STATE=new
ports: # 设置端口映射
- "12379:2379"
- "12380:2380"
networks:
- backend
restart: always etcd2: # 自定义容器名称
build:
context: etcd # 指定构建使用的 Dockerfile 文件
environment:
- TZ=Asia/Shanghai
- ALLOW_NONE_AUTHENTICATION=yes
- ETCD_NAME=etcd2
- ETCD_INITIAL_ADVERTISE_PEER_URLS=http://etcd2:2380
- ETCD_LISTEN_PEER_URLS=http://0.0.0.0:2380
- ETCD_LISTEN_CLIENT_URLS=http://0.0.0.0:2379
- ETCD_ADVERTISE_CLIENT_URLS=http://etcd2:2379
- ETCD_INITIAL_CLUSTER_TOKEN=etcd-cluster
- ETCD_INITIAL_CLUSTER=etcd1=http://etcd1:2380,etcd2=http://etcd2:2380,etcd3=http://etcd3:2380
- ETCD_INITIAL_CLUSTER_STATE=new
ports: # 设置端口映射
- "22379:2379"
- "22380:2380"
networks:
- backend
restart: always etcd3: # 自定义容器名称
build:
context: etcd # 指定构建使用的 Dockerfile 文件
environment:
- TZ=Asia/Shanghai
- ALLOW_NONE_AUTHENTICATION=yes
- ETCD_NAME=etcd3
- ETCD_INITIAL_ADVERTISE_PEER_URLS=http://etcd3:2380
- ETCD_LISTEN_PEER_URLS=http://0.0.0.0:2380
- ETCD_LISTEN_CLIENT_URLS=http://0.0.0.0:2379
- ETCD_ADVERTISE_CLIENT_URLS=http://etcd3:2379
- ETCD_INITIAL_CLUSTER_TOKEN=etcd-cluster
- ETCD_INITIAL_CLUSTER=etcd1=http://etcd1:2380,etcd2=http://etcd2:2380,etcd3=http://etcd3:2380
- ETCD_INITIAL_CLUSTER_STATE=new
ports: # 设置端口映射
- "32379:2379"
- "32380:2380"
networks:
- backend
restart: always

执行docker-compose up -d启动etcd服务,可以看到docker中已经启动了3个服务

实现互斥访问

package main

import (
"fmt"
clientv3 "go.etcd.io/etcd/client/v3"
"go.etcd.io/etcd/client/v3/concurrency"
"sync"
) var sg sync.WaitGroup type Counter struct {
count int
} func (m *Counter) Incr() {
m.count++
} func (m *Counter) Count() int {
return m.count
} func main() {
endpoints := []string{"http://127.0.0.1:12379", "http://127.0.0.1:22379", "http://127.0.0.1:32379"}
// 初始化etcd客户端
client, err := clientv3.New(clientv3.Config{Endpoints: endpoints})
if err != nil {
fmt.Println(err)
return
}
defer client.Close() counter := &Counter{} sg.Add(100)
for i := 0; i < 100; i++ {
go func() {
// 这里会生成租约,默认是60秒
session, err := concurrency.NewSession(client)
if err != nil {
panic(err)
}
defer session.Close() locker := concurrency.NewLocker(session, "/my-test-lock")
locker.Lock()
counter.Incr()
locker.Unlock()
sg.Done()
}()
}
sg.Wait() fmt.Println("count:", counter.Count())
}

执行结果:

user@userdeMacBook-Pro  ~/go/src/go-demo/mutex  go run main.go
count: 100
user@userdeMacBook-Pro  ~/go/src/go-demo/mutex  go run main.go
count: 100
user@userdeMacBook-Pro  ~/go/src/go-demo/mutex  go run main.go
count: 100

实现超时机制

当某个客户端持有锁时,由于某些原因导致锁未释放,就会导致这个客户端一直持有这把锁,其他客户端一直获取不到锁。所以需要分布式锁实现超时机制,当锁未释放时,会因为etcd的租约会到期而释放锁。当业务正常处理时,租约到期之前会继续续约,知道业务处理完毕释放锁。

package main

import (
"fmt"
clientv3 "go.etcd.io/etcd/client/v3"
"go.etcd.io/etcd/client/v3/concurrency"
"sync"
"time"
) var sg sync.WaitGroup type Counter struct {
count int
} func (m *Counter) Incr() {
m.count++
} func (m *Counter) Count() int {
return m.count
} func main() {
endpoints := []string{"http://127.0.0.1:12379", "http://127.0.0.1:22379", "http://127.0.0.1:32379"}
client, err := clientv3.New(clientv3.Config{Endpoints: endpoints})
if err != nil {
fmt.Println(err)
return
}
defer client.Close() counter := &Counter{} session, err := concurrency.NewSession(client)
if err != nil {
panic(err)
}
defer session.Close() locker := concurrency.NewLocker(session, "/my-test-lock")
fmt.Println("locking...", time.Now().Format("2006-01-02 15:04:05"))
locker.Lock()
fmt.Println("locked...", time.Now().Format("2006-01-02 15:04:05"))
// 模拟业务
time.Sleep(100 * time.Second)
counter.Incr()
locker.Unlock()
fmt.Println("released...", time.Now().Format("2006-01-02 15:04:05")) fmt.Println("count:", counter.Count())
}

命令行开2个窗口,第一个窗口执行程序并获取锁,之后模拟意外退出并没有调用unlock方法

go run main.go
locking... 2022-09-03 23:41:48 # 租约生成时间
locked... 2022-09-03 23:41:48
^Csignal: interrupt

第二个窗口,在第一个窗口退出之前尝试获取锁,此时是阻塞状态。第一个窗口退出之后由于租约还没到期,第二个窗口还是获取锁的状态。等到第一个窗口租约到期(默认60秒),第二个获取锁成功

locking... 2022-09-03 23:41:52
locked... 2022-09-03 23:42:48 # 第一个租约60秒到期,获取锁成功
released... 2022-09-03 23:44:28
count: 1

实现阻塞和非阻塞接口

上面的例子中已经实现了阻塞接口,即当前有获取到锁的请求,则其他请求阻塞等待锁释放

非阻塞的方式就是尝试获取锁,如果失败立即返回。etcd中是实现了tryLock方法

// TryLock locks the mutex if not already locked by another session.
// If lock is held by another session, return immediately after attempting necessary cleanup
// The ctx argument is used for the sending/receiving Txn RPC.
func (m *Mutex) TryLock(ctx context.Context) error {

具体看下面的例子

package main

import (
"context"
"fmt"
clientv3 "go.etcd.io/etcd/client/v3"
"go.etcd.io/etcd/client/v3/concurrency"
"sync"
"time"
) var sg sync.WaitGroup type Counter struct {
count int
} func (m *Counter) Incr() {
m.count++
} func (m *Counter) Count() int {
return m.count
} func main() {
endpoints := []string{"http://127.0.0.1:12379", "http://127.0.0.1:22379", "http://127.0.0.1:32379"}
client, err := clientv3.New(clientv3.Config{Endpoints: endpoints})
if err != nil {
fmt.Println(err)
return
}
defer client.Close() counter := &Counter{} session, err := concurrency.NewSession(client)
if err != nil {
panic(err)
}
defer session.Close() // 此处使用newMutex初始化
locker := concurrency.NewMutex(session, "/my-test-lock")
fmt.Println("locking...", time.Now().Format("2006-01-02 15:04:05"))
err = locker.TryLock(context.Background())
// 获取锁失败就抛错
if err != nil {
fmt.Println("lock failed", err)
return
}
fmt.Println("locked...", time.Now().Format("2006-01-02 15:04:05"))
time.Sleep(100 * time.Second)
counter.Incr()
err = locker.Unlock(context.Background())
if err != nil {
fmt.Println("unlock failed", err)
return
}
fmt.Println("released...", time.Now().Format("2006-01-02 15:04:05")) fmt.Println("count:", counter.Count())
}

窗口1、窗口2执行结果

go run main.go
locking... 2022-09-04 00:00:21
locked... 2022-09-04 00:00:21
released... 2022-09-04 00:02:01
count: 1
go run main.go
locking... 2022-09-04 00:00:27
lock failed mutex: Locked by another session

etcd实现分布式锁的相关教程结束。

《etcd实现分布式锁.doc》

下载本文的Word格式文档,以方便收藏与打印。