Centos7 搭建Zookeeper集群

最新推荐文章于 2023-07-26 22:58:20 发布

蜗牛杨哥

最新推荐文章于 2023-07-26 22:58:20 发布

阅读量314

点赞数

分类专栏： Zookeeper集群搭建

本文链接：https://blog.youkuaiyun.com/u014635374/article/details/105104944

版权

Zookeeper集群搭建专栏收录该内容

2 篇文章

订阅专栏

本文详细介绍如何在三台服务器上搭建Zookeeper集群，包括下载安装、配置文件修改、节点角色规划、防火墙设置、开机启动脚本创建等步骤，确保集群稳定运行。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

下载安装包:(hadoop/hadoop)

wget https://mirrors.tuna.tsinghua.edu.cn/apache/zookeeper/ 该路径下寻对应版本

https://mirrors.tuna.tsinghua.edu.cn/apache/zookeeper/zookeeper-3.4.14/zookeeper-3.4.14.tar.gz

Zookeeper集群角色规划

主机	IP	zookeeper 服务之间通信的端口	zookeeper 与其他应用程序通信的端口
centoshadoop1	192.168.227.128	2888	2888+1000
centoshadoop2	192.168.227.129	2888	2888+1000
centoshadoop3	192.168.227.130	2888	2888+1000

mkdir -p /home/hadoop/zookeeper

mkdir -p /home/hadoop/zookeeper/data

mkdir -p /home/hadoop/zookeeper/data/logs

tar zxvf zookeeper-3.4.14.tar.gz -C /home/hadoop/zookeeper

cd /home/hadoop/zookeeper/zookeeper-3.4.14/conf

修改配置文件zoo.cfg

cp zoo_sample.cfg zoo.cfg

先把dataDir=/tmp/zookeeper注释，然后将下面四行代码添加到文件末尾

vi zoo.cfg

dataDirLog=/home/hadoop/zookeeper/data/logs/
dataDir=/home/hadoop/zookeeper/data
server.1=centoshadoop1:2888:3888
server.2=centoshadoop2:2888:3888
server.3=centoshadoop2:2888:3888

分发节点一centoshadoop1上配置完成的zookeeper文件到其他机器(各机器免密登录)

scp -r zookeeper/ hadoop@centoshadoop2:~

scp -r zookeeper/ hadoop@centoshadoop3:~

在centoshadoop1,centoshadoop2,centoshadoop3节点执行，下面写入配置

cd ../../data

touch myid

echo 1 >> myid

cat myid

=============================centoshadoop2==========================

cd ../../data

touch myid

echo 2 >> myid

cat myid

=============================centoshadoop3===========================

cd ../../data

touch myid

echo 3 >> myid

cat myid

各台主机上配置主机名与IP映射关系

vi /etc/hosts

192.168.227.128 centoshadoop1
192.168.227.129 centoshadoop2
192.168.227.130 centoshadoop3

三台机器上分别执行:

firewall-cmd --zone=public --add-port=2888/tcp --permanent

firewall-cmd --zone=public --add-port=3888/tcp --permanent

firewall-cmd --zone=public --add-port=2181/tcp --permanen

重启防火墙

firewall-cmd --reload

=========centoshadoop1,centoshadoop2,centoshadoop3制作开机启动脚本=========================

vi /usr/lib/systemd/system/zookeeper.service

#文件内容

[Unit]

Description=Zookeeper Server

After=network.target

[Service]

Type=forking

Environment=ZOO_LOG_DIR=/home/hadoop/zookeeper/data/logs/

Environment=JAVA_HOME=/home/hadoop/java/jdk1.8.0_192

ExecStart=/home/hadoop/zookeeper/zookeeper-3.4.14/bin/zkServer.sh start

ExecStop=/home/hadoop/zookeeper/zookeeper-3.4.14/bin/zkServer.sh stop

Restart=always

#User=zookeeper

#Group=zookeeper

[Install]

WantedBy=multi-user.target

先前安装时,没有配置这个使用systemctl start zookeeper 一直没有启动,最后发现该处有问题

Environment=JAVA_HOME=/home/hadoop/java/jdk1.8.0_192/

开机启动

systemctl enable zookeeper

重新加载信息

systemctl daemon-reload

启动zookeeper：systemctl start zookeeper.service

关掉zookeeper：systemctl stop zookeeper.service

查看进程状态及日志：systemctl status zookeeper.service

开机自启动：systemctl enable zookeeper.service

关闭自启动：systemctl disable zookeeper.service

firewall-cmd --zone=public --list-ports 查看已经打开的防火墙

chown -R hadoop:hadoop zookeeper-3.4.14

递归地修改zookeeper-3.4.14目录及子目录或者文件为hadoop用户及hadoop组

netstat -apn | grep 2181 查看每个端口是否处于使用状态

cd /home/hadoop/zookeeper/zookeeper-3.4.14/bin

在bin目录下查看客户端连接集群

./zkCli.sh -server 192.168.227.128:2181, 192.168.227.129:2181, 192.168.227.130:2181

查看集群中的某台集群是否为主或者从节点

[root@node-3 bin]# ./zkServer.sh status

ZooKeeper JMX enabled by default

Using config: /home/hadoop/zookeeper/zookeeper-3.4.14/bin/../conf/zoo.cfg

Mode: leader

注意：当不在该文件中/usr/lib/systemd/system/zookeeper.service添加下面hadoop用户及hadoop组的配置时,生成的日志为root账号及root组

/usr/lib/systemd/system/目录下的启动一般需要root权限,所以root开机启动后,生成的动态日志为

/home/hadoop/zookeeper/data/logs/zookeeper.out, 日志为root账号及root用户组,

所以zookeeper账号与zookeeper组不能执行直接这些systemctl start|stop| zookeeper命令,

此时,只能通过sudo获得root权限, 具体设置如下:

方案一:

给hadoop账户添加sudo权限：
       在root下：
           chmod u+w /etc/sudoers
           vi /etc/sudoers
         在 root       ALL=(ALL)    ALL下添加一行：
hadoop  ALL=(ALL)   ALL   //添加sudo授权