ZooKeeper单机&伪分布式集群搭建实战

1.简介

ZooKeeper是一个为分布式应用所设计的分布的、开源的协调服务，它主要是用来解决分布式应用中经常遇到的一些数据管理问题，简化分布式应用协调及其管理的难度，提供高性能的分布式服务。ZooKeeper本身可以以Standalone模式安装运行，不过它的长处在于通过分布式ZooKeeper集群（一个Leader，多个Follower），基于一定的策略来保证ZooKeeper集群的稳定性和可用性，从而实现分布式应用的可靠性。
ZooKeeper是作为分布式协调服务，是不需要依赖于Hadoop的环境，也可以为其他的分布式环境提供服务。

2. 单节点安装

下载,解压,配置

cd /data0/java/deploy/
curl -O http://apache.mirrors.tds.net/zookeeper/stable/zookeeper-3.4.6.tar.gz
# 解压
tar xzvf zookeeper-3.4.6.tar.gz
mv zookeeper-3.4.6 zookeeper346
cd zookeeper346/conf
cp zoo_sample.cfg zoo.cfg

vim zoo.cfg
修改内容如下, 需手动创建定义的数据目录

dataDir=/data0/java/deploy/zookeeper/nodes/zk0

启动

# zookeeper 可执行文件目录
cd ../bin
./zkServer.sh start

至此单节点配置,启动完成.
启动 ：./zkServer.sh start
停止 ：./zkServer.sh stop
查看状态 ：./zkServer.sh status

3 .集群安装

集群中节点树结构

└── nodes
    ├── zk0
    │   └── version-2
    │       └── log.1
    ├── zk1
    │   ├── data
    │   │   ├── myid
    │   │   └── version-2
    │   │       ├── acceptedEpoch
    │   │       ├── currentEpoch
    │   │       └── snapshot.0
    │   └── logs
    │       └── version-2
    │           └── log.100000001
    ├── zk2
    │   ├── data
    │   │   ├── myid
    │   │   └── version-2
    │   │       ├── acceptedEpoch
    │   │       └── currentEpoch
    │   └── logs
    │       └── version-2
    │           └── log.100000001
    └── zk3
        ├── data
        │   ├── myid
        │   └── version-2
        │       ├── acceptedEpoch
        │       ├── currentEpoch
        │       └── snapshot.100000000
        └── logs
            └── version-2
                └── log.100000001

配置节点一
配置文件(/data0/java/deploy/zookeeper/conf/zk1.cfg)

dataDir=/data0/java/deploy/zookeeper/nodes/zk1/data
dataLogDir=/data0/java/deploy/zookeeper/nodes/zk1/logs
clientPort=2181
# 集群实例
server.1=127.0.0.1:2888:3888
server.2=127.0.0.1:2889:3889
server.3=127.0.0.1:2890:3890

根据dataDir,dataLogDir创建相应的目录并在data目录执行如下命令

echo 1 > myid

配置节点二

不解释, 直接上配置文件(/data0/java/deploy/zookeeper/conf/zk2.cfg)

dataDir=/data0/java/deploy/zookeeper/nodes/zk2/data
dataLogDir=/data0/java/deploy/zookeeper/nodes/zk2/logs
clientPort=2182
# 集群实例
server.1=127.0.0.1:2888:3888
server.2=127.0.0.1:2889:3889
server.3=127.0.0.1:2890:3890

根据dataDir,dataLogDir创建相应的目录并在data目录执行如下命令

echo 2 > myid

配置节点三

不解释, 直接上配置文件(/data0/java/deploy/zookeeper/conf/zk3.cfg)

dataDir=/data0/java/deploy/zookeeper/nodes/zk3/data
dataLogDir=/data0/java/deploy/zookeeper/nodes/zk3/logs
clientPort=2183
# 集群实例
server.1=127.0.0.1:2888:3888
server.2=127.0.0.1:2889:3889
server.3=127.0.0.1:2890:3890

根据dataDir,dataLogDir创建相应的目录并在data目录执行如下命令

echo 3 > myid

启动Zookeeper集群

# zookeeper可执行文件目录
cd bin
./zkServer.sh start zk1.cfg
./zkServer.sh start zk2.cfg
./zkServer.sh start zk3.cfg

查看Zookeeper集群中各个节点的状态

./zkServer.sh status zk1.cfg
JMX enabled by default
Using config: /data0/java/deploy/zookeeper/bin/../conf/zk1.cfg
Mode: follower
./zkServer.sh status zk2.cfg
JMX enabled by default
Using config: /data0/java/deploy/zookeeper/bin/../conf/zk2.cfg
Mode: leader
./zkServer.sh status zk3.cfg
JMX enabled by default
Using config: /data0/java/deploy/zookeeper/bin/../conf/zk3.cfg
Mode: follower

注: 可以考虑多种启动或停止zookeeper的情况进行状态查看, 可发现zookeeper的高可用(HA);

4.命令行操作

4.1 使用命令行client连接zookeeper并操作API

连接到任意节点
./zkCli.sh -server 127.0.0.1:2182

4.2 help 查看指令

ZooKeeper -server host:port cmd args
    connect host:port
    get path [watch]
    ls path [watch]
    set path data [version]
    rmr path
    delquota [-n|-b] path
    quit
    printwatches on|off
    create [-s] [-e] path data acl
    stat path [watch]
    close
    ls2 path [watch]
    history
    listquota path
    setAcl path acl
    getAcl path
    sync path
    redo cmdno
    addauth scheme auth
    delete path [version]
    setquota -n|-b val path

4.3其他常用命令

• 查看/目录节点
  ls /
• 创建znode节点
  create /config db.properties
• 查看/下节点, ls /
  [config, zookeeper]
• 查看节点内容, get /config

db.properties
cZxid = 0x300000011
ctime = Fri Nov 14 23:05:05 CST 2014
mZxid = 0x300000011
mtime = Fri Nov 14 23:05:05 CST 2014
pZxid = 0x300000011
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 13
numChildren = 0

• 修改节点内容, set /config db2.properties
• 再次查看 get /config

db2.properties
cZxid = 0x300000011
ctime = Fri Nov 14 23:05:05 CST 2014
mZxid = 0x300000012
mtime = Fri Nov 14 23:08:10 CST 2014
pZxid = 0x300000011
cversion = 0
dataVersion = 1
aclVersion = 0
ephemeralOwner = 0x0
dataLength = 14
numChildren = 0

• 删除节点, delete /config
• 查看, ls /
• 退出, quit

5.Java操作

源码

package com.yangyang.zookeeper;
import java.io.IOException;
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.Ids;
import org.apache.zookeeper.ZooKeeper;
/**
 * Java 操作zookeeper API
 *
 */
public class Main
{
    public static void main( String[] args ) throws IOException, InterruptedException, KeeperException
    {
        // 创建连接
        ZooKeeper zk = new ZooKeeper("127.0.0.1:2182", 30000, new Watcher(){
            // 监听的事件发生变化时, 调用该函数
            public void process(WatchedEvent event) {
                System.out.println("Event emit -> " + event.getType());
            }
        });
       
        // 查看根目录, ls /
        System.out.println("[line: 29]command: ls / -> " + zk.getChildren("/", true));
       
        // 创建数据节点config
        if(zk.exists("/config", true) == null) {
            zk.create("/config", "db.properties".getBytes(), Ids.OPEN_ACL_UNSAFE, CreateMode.PERSISTENT);
           
            System.out.println("创建数据节点config成功.");
           
            // 查看数据节点数据
            System.out.println("[line: 38]command get /config: " + new String(zk.getData("/config", true, null)));
           
            // 查看根目录, ls /
            System.out.println("[line: 41]command: ls / -> " + zk.getChildren("/", true));
        }
       
        // 修改数据节点
        if(zk.exists("/config", true) != null) {
            zk.setData("/config", "db2.properteis".getBytes(), -1);
            // 查看数据节点数据
            System.out.println("[line: 48]command get /config: " + new String(zk.getData("/config", true, null)));
            // 查看根目录, ls /
            System.out.println("[line: 50]command: ls / -> " + zk.getChildren("/", true));
        }
       
        // 删除数据节点
        if(zk.exists("/config", true) != null) {
            zk.delete("/config", -1);
            // 查看根目录, ls /
            System.out.println("[line: 57]command: ls / -> " + zk.getChildren("/", true));
        }
       
        // 关闭连接
        zk.close();
    }
}

结果

Event emit -> None
[line: 29]command: ls / -> [zookeeper]
Event emit -> NodeCreated
Event emit -> NodeChildrenChanged
创建数据节点config成功.
[line: 38]command get /config: db.properties
[line: 41]command: ls / -> [config, zookeeper]
Event emit -> NodeDataChanged
[line: 48]command get /config: db2.properteis
[line: 50]command: ls / -> [config, zookeeper]
Event emit -> NodeDeleted
Event emit -> NodeChildrenChanged
[line: 57]command: ls / -> [zookeeper]

参考资料
ZooKeeper伪分布式集群安装及使用

欢迎大家扫码关注我的微信公众号，与大家一起分享技术与成长中的故事。

我的微信公众号.jpg