引言
ZooKeeper是一个分布式的，开放源码的分布式应用程序协调服务，是Google的Chubby一个开源的实现，是Hadoop和Hbase的重要组件。它是一个为分布式应用提供一致性服务的软件，提供的功能包括：配置维护、域名服务、分布式同步、组服务等。
ZooKeeper的架构通过冗余服务实现高可用性。因此，如果第一次无应答，客户端就可以询问另一台ZooKeeper主机。ZooKeeper节点将它们的数据存储于一个分层的命名空间，非常类似于一个文件系统或一个前缀树结构。客户端可以在节点读写，从而以这种方式拥有一个共享的配置服务。更新是全序的。
基于ZooKeeper分布式锁的原理
- 在zookeeper指定节点（locks）下创建临时顺序节点node_n
- 获取locks下所有子节点children
- 对子节点按节点自增序号从小到大排序
- 判断本节点是不是第一个子节点，若是，则获取锁；若不是，则监听比该节点小的那个节点的删除事件
- 若监听事件生效，则回到第二步重新进行判断，直到获取到锁
具体实现
下面就具体使用java和zookeeper实现分布式锁，操作zookeeper使用的是apache提供的zookeeper的包。
- 通过实现Watch接口，实现process(WatchedEvent event)方法来实施监控，使CountDownLatch来完成监控，在等待锁的时候使用CountDownLatch来计数，等到后进行countDown，停止等待，继续运行。
- 以下整体流程基本与上述描述流程一致，只是在监听的时候使用的是CountDownLatch来监听前一个节点。
package org.java.base.zookeeper;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.List;
import java.util.concurrent.CountDownLatch;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.locks.Condition;
import java.util.concurrent.locks.Lock;
import org.apache.zookeeper.CreateMode;
import org.apache.zookeeper.KeeperException;
import org.apache.zookeeper.WatchedEvent;
import org.apache.zookeeper.Watcher;
import org.apache.zookeeper.ZooDefs;
import org.apache.zookeeper.ZooKeeper;
import org.apache.zookeeper.data.Stat;
import org.codehaus.jettison.json.JSONString;
import com.fasterxml.jackson.databind.util.JSONPObject;
import com.google.gson.JsonObject;
public class DistributedLock implements Lock, Watcher {
private ZooKeeper zk = null;
// 根节点
private String ROOT_LOCK = "/locks";
// 竞争的资源
private String lockName;
// 等待的前一个锁
private String WAIT_LOCK;
// 当前锁
private String CURRENT_LOCK;
// 计数器
private CountDownLatch countDownLatch;
private int sessionTimeout = 30000;
private List<Exception> exceptionList = new ArrayList<Exception>();
/**
* 配置分布式锁
* @param config 连接的url
* @param lockName 竞争资源
*/
public DistributedLock(String config, String lockName) {
this.lockName = lockName;
try {
// 连接zookeeper
zk = new ZooKeeper(config, sessionTimeout, this);
Stat stat = zk.exists(ROOT_LOCK, false);
if (stat == null) {
// 如果根节点不存在，则创建根节点
zk.create(ROOT_LOCK, new byte[0], ZooDefs.Ids.OPEN_ACL_UNSAFE, CreateMode.PERSISTENT);
}
} catch (IOException e) {
e.printStackTrace();
} catch (InterruptedException e) {
e.printStackTrace();
} catch (KeeperException e) {
e.printStackTrace();
}
}
// 节点监视器
public void process(WatchedEvent event) {
if (this.countDownLatch != null) {
this.countDownLatch.countDown();
}
}
public void lock() {
if (exceptionList.size() > 0) {
throw new LockException(exceptionList.get(0));
}
try {
if (this.tryLock()) {
System.out.println(Thread.currentThread().getName() + " " + lockName + "获得了锁");
return;
} else {
// 等待锁
waitForLock(WAIT_LOCK, sessionTimeout);
}
} catch (InterruptedException e) {
e.printStackTrace();
} catch (KeeperException e) {
e.printStackTrace();
}
}
public boolean tryLock() {
try {
String splitStr = "_lock_";
if (lockName.contains(splitStr)) {
throw new LockException("锁名有误");
}
// 创建临时有序节点
CURRENT_LOCK = zk.create(ROOT_LOCK + "/" + lockName + splitStr, new byte[0],
ZooDefs.Ids.OPEN_ACL_UNSAFE, CreateMode.EPHEMERAL_SEQUENTIAL);
System.out.println(CURRENT_LOCK + " 已经创建");
// 取所有子节点
List<String> subNodes = zk.getChildren(ROOT_LOCK, false);
// 取出所有lockName的锁
List<String> lockObjects = new ArrayList<String>();
for (String node : subNodes) {
String _node = node.split(splitStr)[0];
if (_node.equals(lockName)) {
lockObjects.add(node);
}
}
Collections.sort(lockObjects);
System.out.println(Thread.currentThread().getName() + " 的锁是 " + CURRENT_LOCK);
// 若当前节点为最小节点，则获取锁成功
if (CURRENT_LOCK.equals(ROOT_LOCK + "/" + lockObjects.get(0))) {
return true;
}
// 若不是最小节点，则找到自己的前一个节点
String prevNode = CURRENT_LOCK.substring(CURRENT_LOCK.lastIndexOf("/") + 1);
WAIT_LOCK = lockObjects.get(Collections.binarySearch(lockObjects, prevNode) - 1);
} catch (InterruptedException e) {
e.printStackTrace();
} catch (KeeperException e) {
e.printStackTrace();
}
return false;
}
public boolean tryLock(long timeout, TimeUnit unit) {
try {
if (this.tryLock()) {
return true;
}
return waitForLock(WAIT_LOCK, timeout);
} catch (Exception e) {
e.printStackTrace();
}
return false;
}
// 等待锁
private boolean waitForLock(String prev, long waitTime) throws KeeperException, InterruptedException {
Stat stat = zk.exists(ROOT_LOCK + "/" + prev, true);
if (stat != null) {
System.out.println(Thread.currentThread().getName() + "等待锁 " + ROOT_LOCK + "/" + prev);
this.countDownLatch = new CountDownLatch(1);
// 计数等待，若等到前一个节点消失，则precess中进行countDown，停止等待，获取锁
this.countDownLatch.await(waitTime, TimeUnit.MILLISECONDS);
this.countDownLatch = null;
System.out.println(Thread.currentThread().getName() + " 等到了锁");
}
return true;
}
public void unlock() {
try {
System.out.println("释放锁 " + CURRENT_LOCK);
zk.delete(CURRENT_LOCK, -1);
CURRENT_LOCK = null;
zk.close();
} catch (InterruptedException e) {
e.printStackTrace();
} catch (KeeperException e) {
e.printStackTrace();
}
}
public Condition newCondition() {
return null;
}
public void lockInterruptibly() throws InterruptedException {
this.lock();
}
public class LockException extends RuntimeException {
private static final long serialVersionUID = 1L;
public LockException(String e){
super(e);
}
public LockException(Exception e){
super(e);
}
}
}
测试类
package org.java.base.zookeeper;
/**
* 这段代码有一个问题，看大家能否发现？
* @author Liuhaihua
*
*/
public class Test {
static int n = 500;
public static void secskill() {
System.out.println(--n);
}
public static void main(String[] args) {
Runnable runnable = new Runnable() {
public void run() {
DistributedLock lock = null;
try {
lock = new DistributedLock("10.152.17.73:2181", "test1");
lock.lock();
secskill();
System.out.println(Thread.currentThread().getName() + "正在运行");
} finally {
if (lock != null) {
lock.unlock();
}
}
}
};
for (int i = 0; i < 10; i++) {
Thread t = new Thread(runnable);
t.start();
}
}
}
zk 实现的分布式锁一般性能是较差的，大家有什么不同的想法建议可以再下方评论出来，欢迎大家交流！