对于Zookeeper的认知

说一下zookeeper
zookeeper是分布式协调服务
其中配置有
zoo.cfg
datadir=
cd /etc/profile ----->

server.1=bin02:2888:3888
server.2=bin03:2888:3888
server.3=bin04:2888:3888

其中2888端口提供zookeeper对外通信，
3888端口：当leader挂掉之后，重新选择leader的时候提供对外通信；

切记：关闭防火墙！！！！！

角色：
leader 领导者，发起一个请求，并且开始投票
follower 是跟从者，响应leader的请求并发起投票
observer 监视leader 并将leader的状态告诉follower

有：–》选举机制、原子广播和zab协议、监听机制
如下：
1：一个外部客户端写入请求，将请求传给follower
2：follower将请求转发给leader
3：leader将请求发到各个follower，进行投票
4：follower将自己的想法返回给leader
原则（少数服从多数）过半原则：节点数最好是单数；
3+4是原子广播

 zab协议：
 		广播模式-------已经选举出来的leader，开始对外提供服务------------原子广播
 		恢复模式----还没有leader（这个情况可能是集群刚启动或者是leader挂掉了新的还没有起来）
 					-------looking  观望
 					-------following----跟从但是有想法
 					-----leading   继承人（准备被继承）leader
 					observing--------observer

监听机制
follower 查看leader的状态。然后将状态信息进行更新和发送

zookeeper的文件节点----------znode
启动客户端： zkCli.sh -server 192.168.35.125:2181

文件节点操作文件的命令
在这里插入图片描述

上手操作：

[zk: 192.168.35.125:2181(CONNECTED) 21] ls /
[name, zookeeper, name20000000002, name10000000001]
[zk: 192.168.35.125:2181(CONNECTED) 22] rmr /name 
[zk: 192.168.35.125:2181(CONNECTED) 23] ls /
[zookeeper, name20000000002, name10000000001]

[zk: 192.168.35.125:2181(CONNECTED) 19] set /name buzhiushi123
cZxid = 0x200000004
ctime = Tue Jun 18 01:45:41 CST 2019
mZxid = 0x200000008
mtime = Tue Jun 18 01:51:03 CST 2019
pZxid = 0x200000004
cversion = 0
dataVersion = 1
aclVersion = 0
ephemeralOwner = 0x16b66d95ad00001
dataLength = 12
numChildren = 0
[zk: 192.168.35.125:2181(CONNECTED) 20] get /name
buzhiushi123
cZxid = 0x200000004
ctime = Tue Jun 18 01:45:41 CST 2019
mZxid = 0x200000008
mtime = Tue Jun 18 01:51:03 CST 2019
pZxid = 0x200000004
cversion = 0
dataVersion = 1
aclVersion = 0
ephemeralOwner = 0x16b66d95ad00001
dataLength = 12
numChildren = 0

[zk: 192.168.35.125:2181(CONNECTED) 2] create -e /name "123"
Created /name
[zk: 192.168.35.125:2181(CONNECTED) 3] ls /
[name, zookeeper]
[zk: 192.168.35.125:2181(CONNECTED) 4] create -s /name1 xiaopengyou
Created /name10000000001
[zk: 192.168.35.125:2181(CONNECTED) 5] create -s /name2 xiaojiejie 
Created /name20000000002
[zk: 192.168.35.125:2181(CONNECTED) 6] ls /
[name, zookeeper, name20000000002, name10000000001]

在eclipse中使用zookeeper

public class TestZ {
	public static void main(String[] args) throws Exception {
		ZooKeeper zooKeeper = new ZooKeeper("192.168.35.125", 1000, null);
		//创建主节点
		//String create = zooKeeper.create("/name", "dididididi".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.PERSISTENT);
		
		//创建子节点
		/*String create = zooKeeper.create("/name/sex", "didididadada".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.PERSISTENT);
		System.out.println(create);*/
		//获取节点内容
		/*byte[] data = zooKeeper.getData("/name", true, null);	
		String string = new String(data);
		System.out.println(string);//dididididi
*/		//删除节点
		//zooKeeper.delete("/name/sex", -1);
		
		//设置节点内容
		//zooKeeper.setData("/name", "2222222222".getBytes(), zooKeeper.exists("/name", true).getAversion());
		byte[] data = zooKeeper.getData("/name", true, null);
		String string = new String(data);
		System.out.println(string);
		
	}

}

public class TestWatcher implements Watcher{
	
	private static final int SESSION_TIMEOUT=1000;
	private ZooKeeper zk=null;
	/**
	 * @throws Exception 
	 * @throws Exception 
	 * 
	 */
	public static void main(String[] args) throws Exception {
		ZooKeeper zooKeeper = new ZooKeeper("192.168.35.125", 1000, null);
		byte[] data = zooKeeper.getData("/learning", new TestWatcher(), null);
		System.out.println(new String(data));
		Thread.sleep(Long.MAX_VALUE);
	}
	

	@Override
	public void process(WatchedEvent event) {

		System.out.println(event.getPath());
	}

}

对于Zookeeper的认知

猜你喜欢