Zookeeper集群安装与配置

在本文中，Zookeeper节点个数（奇数）为3个。Zookeeper默认对外提供服务的端口号2181。Zookeeper集群内部3个节点之间通信默认使用2888:3888。

1. Hadoop集群环境

2. 下载Zookeeper安装包

下载地址：http://mirrors.shu.edu.cn/apache/zookeeper/
根据需要选择合适的版本，本文为zookeeper-3.4.14.tar.gz.

3. 上传、解压Zookeeper包

进入/opt目录，使用rz命令上传到hadoop1服务器并解压，命令如下：

#cd /opt
#rz
#tar -zxvf zookeeper-3.4.14.tar.gz
#生成软链接
#ln -s /opt/zookeeper-3.4.14 /opt/zookeeper

4. 配置环境变量（每个节点都需要配置）

# vim /etc/profile
# zookeeper
export ZK_HOME=/opt/zookeeper # zookeeper解压安装目录
export PATH=$PATH:$ZK_HOME/bin

#配置后，使之生效
# source /etc/profile

依次在hadoop1、hadoop2、hadoop3、hadoop4节点上执行上述操作。

5. 配置Zookeeper参数文件（每个节点都需要配置）

（1）配置/opt/zookeeper/conf/zoo.cfg文件

首先，复制zoo_sample.cfg文件，命名为zoo.cfg。

[root@hadoop1 conf]# ls
configuration.xsl  log4j.properties  zoo_sample.cfg
[root@hadoop1 conf]# cp zoo_sample.cfg zoo.cfg
[root@hadoop1 conf]# ls
configuration.xsl  log4j.properties  zoo.cfg  zoo_sample.cfg

其次，编辑zoo.cfg配置文件（每个节点配置相同）

#保持默认值
tickTime=2000
initLimit=10
syncLimit=5
#修改存储快照文件snapshot的目录
dataDir=/opt/zookeeper/data
clientPort=2181
#这里server.x中的x是一个数字，与myid文件中的id一致
server.1=hadoop1:2888:3888
server.2=hadoop2:2888:3888
server.3=hadoop3:2888:3888

（2）配置/opt/zookeeper/data/myid文件（三个节点均需配置）

创建data文件夹，新建myid文件，路径为zoo.cfg文件中dataDir指定的路径，本文为/opt/zookeeper/data。

hadoop1节点：

# cd /opt/zookeeper
# mkdir data
# cd data
# echo 1  > myid

6. 分发安装包
将配置好的/opt/zookeeper文件夹拷贝到hadoop2、hadoop3节点。

# scp -r /opt/zookeeper hadoop2:/opt/
# scp -r /opt/zookeeper hadoop3:/opt/
# scp -r /opt/zookeeper hadoop4:/opt/

# scp -r /opt/zookeeper-3.4.14 hadoop2:/opt/
# scp -r /opt/zookeeper-3.4.14 hadoop3:/opt/
# scp -r /opt/zookeeper-3.4.14 hadoop4:/opt/

注意：修改hadoop2、hadoop3服务器对应/opt/zookeeper/data/myid文件内容。

hadoop2节点：

# echo 2  > /opt/zookeeper/data/myid

hadoop3节点：

# echo 3  > /opt/zookeeper/data/myid

7. 启动Zookeeper集群（分别在每台服务器启动Zookeeper）

[root@hadoop1 ~]# /opt/zookeeper/bin/zkServer.sh start

　　ZooKeeper JMX enabled by default
　　Using config: /opt/zookeeper/bin/../conf/zoo.cfg
　　Starting zookeeper ... STARTED

[root@hadoop2 ~]# /opt/zookeeper/bin/zkServer.sh start

[root@hadoop3 ~]# /opt/zookeeper/bin/zkServer.sh start

使用zkServer.sh status命令查看状态

[root@hadoop1 conf]# zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /opt/zookeeper/bin/../conf/zoo.cfg
Mode: follower

==============================以下是Zookeeper集群启动自动化脚本========================

zk集群脚本编写

将脚本放到/usr/local/bin目录下，方便调用。

# vim /usr/local/bin/xzk-cluster.sh
#修改脚本执行权限
# chmod 755 /usr/local/bin/xzk-cluster.sh

脚本内容如下：

#!/bin/bash
cmd=$1
servers="hadoop1 hadoop2 hadoop3"
for s in $servers ; do
   tput setaf 3
   echo ========== $s ==========
   tput setaf 7
   ssh $s "source /etc/profile ; zkServer.sh $cmd"
done

使用范例：

[root@hadoop1 conf]# xzk-cluster.sh status
========== hadoop1 ==========
ZooKeeper JMX enabled by default
Using config: /opt/zookeeper/bin/../conf/zoo.cfg
Mode: follower
========== hadoop2 ==========
ZooKeeper JMX enabled by default
Using config: /opt/zookeeper/bin/../conf/zoo.cfg
Mode: leader
========== hadoop3 ==========
ZooKeeper JMX enabled by default
Using config: /opt/zookeeper/bin/../conf/zoo.cfg
Mode: follower

[root@hadoop1 conf]# xzk-cluster.sh stop
========== hadoop1==========
ZooKeeper JMX enabled by default
Using config: /opt/zookeeper/bin/../conf/zoo.cfg
Stopping zookeeper ... STOPPED
========== hadoop2 ==========
ZooKeeper JMX enabled by default
Using config: /opt/zookeeper/bin/../conf/zoo.cfg
Stopping zookeeper ... STOPPED
========== hadoop3 ==========
ZooKeeper JMX enabled by default
Using config: /opt/zookeeper/bin/../conf/zoo.cfg
Stopping zookeeper ... STOPPED