Big data cluster construction (zookeeper, high-availability hadoop, high-availability hbase)
1. Preparation
- Architecture
- Three mutual password-free login
//1.进入ssh目录
cd ~/.ssh/
//若出现 没有那个文件或目录错误 “ssh localhost”再进入
//2. 生成秘钥对
ssh-keygen -t rsa
//3. 拷贝秘钥到指定ip,先对自己免密然后另外两个，共9次
ssh-copy-id 192.168.80.128
ssh-copy-id 192.168.80.129
ssh-copy-id 192.168.80.130
...
//4.测试免密是否成功
ssh 192.168.80.129
ssh 192.168.80.130
ssh 192.168.80.128
3. Turn off the firewall
systemctl stop firewalld.service
systemctl disable firewalld.service
- Configure the domain name in the hosts file of the zhiyou001 host (all installations are in zhiyou001, and finally copy the other two.)
vi /etc/hosts
//修改为下面内容
#127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
#::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
# 上面的给注释掉或者删除
192.168.80.128 zhiyou001
192.168.80.129 zhiyou002
192.168.80.130 zhiyou003
//复制给另外两台
scp -r /etc/hosts [email protected]:/etc/hosts
scp -r /etc/hosts [email protected]:/etc/hosts
- Install jdk
- The tutorial is out, so I won’t repeat it here
- Server deployment-"jdk articles"
- Copy installation files and environment variables to zhiyou002 and zhiyou003
//复制jdk
scp -r /opt/java/ root@zhiyou002:/opt/java/
scp -r /opt/java/ root@zhiyou003:/opt/java/
//复制配置文件
scp -r /etc/profile root@zhiyou002:/etc/profile
scp -r /etc/profile root@zhiyou003:/etc/profile
//刷新配置文件；可右键“发送键输入到所有对话”同时刷新
source /etc/profile
//查看配置是否成功（三台）
java javac java -version
2. Title technology installation method. It’s too long to read, so it’s released separately.
1. Big data cluster construction-"zookeeper"
2. Building a big data cluster-"High-availability hadoop"
3. Building a Big Data Cluster-"Highly Available Habse"