Hadoop安装
需要先安装JDK 通过环境变量查看jdk位置:echo $JAVA_HOME
修改环境变量:vim /etc/profile
环境变量生效:source /etc/profile
下载Hadoop:
wget https://mirrors.tuna.tsinghua.edu.cn/apache/hadoop/common/hadoop-2.9.2/hadoop-2.9.2.tar.gz
修改配置文件:
vim /root/software/hadoop-2.7.7/etc/hadoop/hadoop-env.sh
vim /root/software/hadoop-2.7.7/etc/hadoop/core-site.xml
fs.defaultFS hdfs://localhost:9000 hadoop.tmp.dir /root/hadoopData/tempvim /root/software/hadoop-2.7.7/etc/hadoop/hdfs-site.xml
dfs.namenode.name.dir /root/hadoopData/name dfs.datanode.data.dir /root/hadoopData/data dfs.replication 1vim /root/software/hadoop-2.7.7/etc/hadoop/slaves
单台默认为localhost,无需修改
修改配置Hadoop 系统环境变量:
vim /etc/profile
配置Hadoop的安装目录
export HADOOP_HOME=/root/software/hadoop-2.7.7
在原PATH的基础上加入Hadoop的bin和sbin目录
export PATH= P A T H : PATH: PATH:HADOOP_HOME/bin:$HADOOP_HOME/sbin
让配置文件立即生效:source /etc/profile
检测 Hadoop 环境变量是否设置成功:hadoop version
HDFS 集群测试
hdfs namenode -format
启动 HDFS 集群: start-dfs.sh
关闭 HDFS 集群: stop-dfs.sh
默认端口: http://localhost:50070
出现这 4 个进程表示进程启动成功:
YARN 集群配置:
vim /usr/local/hadoop-2.9.2/etc/hadoop/yarn-env.sh
配置计算框架 mapred-site.xml:
mapreduce.framework.name yarn配置 YARN 系统 yarn-site.xml:
vim yarn-site.xml
yarn.nodemanager.aux-services mapreduce_shuffle脚本一键启动和关闭:
首先保证 HDFS 集群处于启动状态
启动 yarn集群: start-yarn.sh
关闭 yarn集群: stop-yarn.sh
默认端口: http://localhost:8088
全部启动/关闭:start-all.sh / stop-all.sh
集群搭建
修改主机名:
vim /etc/hostname
配置hosts:
vi /etc/hosts
192.168.102.135 master
192.168.102.136 slave1
使hosts生效:/bin/hostname hostsname
关闭防火墙:
查看防火墙状态
firewall-cmd --state
停止firewall
systemctl stop firewalld.service
禁止firewall开机启动
systemctl disable firewalld.service
时间同步:
yum更新源:yum update
下载ntp插件:yum install ntp
ntpdate -u ntp1.aliyun.com
date
Master与slave安装配置ssh:
安装SSH: yum -y install openssh-server
生成一个无密码密钥对: ssh-keygen -t rsa
id_rsa.pub(公钥)追加到授权的key里面:cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
修改.ssh目录下authorized_keys文件权限: chmod 600 authorized_keys
修改SSH配置文件"/etc/ssh/sshd_config"的下列内容:
RSAAuthentication yes # 启用 RSA 认证
PubkeyAuthentication yes # 启用公钥私钥配对认证方式
AuthorizedKeysFile %h/.ssh/authorized_keys # 公钥文件路径
重启SSH服务: service sshd restart
将公钥传送到远程主机上: ssh-copy-id root@xxx.xxx.xxx.xxx
测试是否无密码登录其它机器成功: ssh xxx.xxx.xxx.xxx
断开ssh连接: CTRL+D 或 logout
配置集群节点地址:
vi /usr/local/hadoop-2.9.2/etc/hadoop/slaves
将Master中配置好的hadoop传入Slave中:
scp -r /usr/local/hadoop-2.9.2/ root@47.95.216.255:/usr/local/
修改Master主机上的workers文件:
vi /usr/local/hadoop-2.9.2/etc/hadoop/workers
47.95.216.255 # slave-ip
master节点格式化HDFS文件系统:
hdfs namenode -format
查看集群状态:
主节点刷新:hdfs dfsadmin -refreshNodes
负载均衡:sbin/start-balancer.sh
查看存活节点:hdfs dfsadmin -report
集群添加删除
添加:
修改所有slaves文件:添加slave3
启动DataNode进程:hadoop-daemon.sh start datanode
启动nodemanage:yarn-daemon.sh start nodemanager
删除:
hadoop-daemon.sh stop datanode
hadoop-daemon.sh stop tasktracker