1.CentOS7 搭建 Hadoop 单节点模式

本文旨在搭建一个简单的 Hadoop 环境用于编程，依据官方文档而成，补充了几个搭建过程中可能会遇到的问题。官方文档地址：
https://hadoop.apache.org/docs/stable/hadoop-project-dist/hadoop-common/SingleCluster.html

1.前期准备 1.1.硬件要求

CentOS7 服务器一台，虚拟机也可以，保证网络畅通

[root@localhost ~]# ping www.baidu.com
PING www.wshifen.com (103.235.46.39) 56(84) bytes of data.
64 bytes from 103.235.46.39 (103.235.46.39): icmp_seq=2 ttl=128 time=249 ms

1.2.jdk8 环境

jdk8 环境搭建参见笔者的另一篇文章：
CentOS7 安装 jdk8

1.3.ssh 命令

直接执行安装命令

[root@localhost ~]# yum install ssh

检查是否能够免密登录本机

[root@localhost ~]# ssh localhost

开启本机的免密访问权限

[root@localhost ~]# ssh-keygen -t rsa -P '' -f ~/.ssh/id_rsa
[root@localhost ~]# cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
[root@localhost ~]# chmod 0600 ~/.ssh/authorized_keys

1.4.关闭防火墙

[root@localhost hadoop]# systemctl stop firewalld.service
[root@localhost hadoop]# systemctl disable firewalld.service

1.5.下载 Hadoop

选择合适的版本下载，笔者的版本是3.3.2
下载地址

2.安装 Hadoop 2.1.上传安装包并解压

[root@localhost usr]# pwd
/usr
[root@localhost usr]# tar -xvf hadoop-3.3.2.tar.gz
// 给文件夹改名
[root@localhost usr]# mv hadoop-3.3.2/ hadoop/
// 删除安装包
[root@localhost usr]# rm -rf hadoop-3.3.2.tar.gz

2.2.配置 JAVA_HOME

[root@localhost usr]# cd hadoop/
[root@localhost hadoop]# pwd
/usr/hadoop
[root@localhost hadoop]# vi etc/hadoop/hadoop-env.sh 

修改如下内容：

# The java implementation to use. By default, this environment
# variable is REQUIRED on ALL platforms except OS X!
export JAVA_HOME=/usr/java/jdk1.8.0_333

2.3.伪分布模式启动 Hadoop 2.3.1.配置

etc/hadoop/core-site.xml:

    
        fs.defaultFS
        hdfs://localhost:9000
    

etc/hadoop/hdfs-site.xml:

    
        dfs.replication
        1
    

2.3.2.配置 root 用户运行

hadoop中没有默认root的用户，需要手动添加。
sbin/start-dfs.sh 和 sbin/stop-dfs.sh

HDFS_DATANODE_USER=root
HADOOP_SECURE_DN_USER=hdfs
HDFS_NAMENODE_USER=root
HDFS_SECONDARYNAMENODE_USER=root

关闭CentOS中SELinux
/etc/selinux/config

SELINUX=disabled

2.3.3运行 Hadoop

1. 格式化文件系统

[root@localhost hadoop]# bin/hdfs namenode -format

2. 运行NameNode和DataNode

[root@localhost hadoop]# sbin/start-dfs.sh

3. 打开浏览器访问：
   http://ip:9870/
4. 创建执行 MepReduce 任务所需的文件夹

[root@localhost hadoop]# bin/hdfs dfs -mkdir /user
[root@localhost hadoop]# bin/hdfs dfs -mkdir /user/root

5. 将输入文件拷贝至 Hdfs

[root@localhost hadoop]# bin/hdfs dfs -mkdir input
[root@localhost hadoop]# bin/hdfs dfs -put etc/hadoop/*.xml input 

6. 运行示例

[root@localhost hadoop]# bin/hadoop jar share/hadoop/mapreduce/hadoop-mapreduce-examples-3.3.2.jar grep input output 'dfs[a-z.]+'

7. 查看运算结果

[root@localhost hadoop]# bin/hdfs dfs -cat output/*

8. 关闭 Hadoop

[root@localhost hadoop]# sbin/stop-dfs.sh

3.安装 Yarn 3.1.配置

etc/hadoop/mapred-site.xml:

    
        mapreduce.framework.name
        yarn
    
    
        mapreduce.application.classpath
        $HADOOP_MAPRED_HOME/share/hadoop/mapreduce/*:$HADOOP_MAPRED_HOME/share/hadoop/mapreduce/lib/*
    

etc/hadoop/yarn-site.xml:

    
        yarn.nodemanager.aux-services
        mapreduce_shuffle
    
    
        yarn.nodemanager.env-whitelist
        JAVA_HOME,HADOOP_COMMON_HOME,HADOOP_HDFS_HOME,HADOOP_CONF_DIR,CLASSPATH_PREPEND_DISTCACHE,HADOOP_YARN_HOME,HADOOP_HOME,PATH,LANG,TZ,HADOOP_MAPRED_HOME
    

3.2.配置 root 用户运行

sbin/start-yarn.sh 和 sbin/stop-yarn.sh 添加

YARN_RESOURCEMANAGER_USER=root
HADOOP_SECURE_DN_USER=yarn
YARN_NODEMANAGER_USER=root

3.3.运行 Yarn

1. 启动 ResourceManager 和 NodeManager

[root@localhost hadoop]# sbin/start-yarn.sh

2. 打开浏览器访问：
   http://ip:8088/
3. 运行任务
4. 关闭 Yarn

[root@localhost hadoop]# sbin/stop-yarn.sh