准备工作

关闭防火墙 (也可放行)

# 停止防火墙
systemctl stop firewalld

# 关闭防火墙开机自启动
systemctl disable firewalld

修改 hosts 文件，让 hadoop 对应本机 IP 地址 (非 127.0.0.1)

vim /etc/hosts

127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6

xxx.xxx.xxx.xxx    hadoop

安装 JDK

https://www.oracle.com/technetwork/java/javase/downloads/jdk8-downloads-2133151.html

# 解压
tar -zxf /opt/jdk-8u202-linux-x64.tar.gz -C /opt/

# 配置环境变量
vim /etc/profile

# JAVA_HOME
export JAVA_HOME=/opt/jdk1.8.0_202/
export CLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar:$CLASSPATH
export PATH=$JAVA_HOME/bin:$JAVA_HOME/jre/bin:$PATH

# 刷新环境变量
source /etc/profile

# 验证
java -version

# java version "1.8.0_202"
# Java(TM) SE Runtime Environment (build 1.8.0_202-b08)
# Java HotSpot(TM) 64-Bit Server VM (build 25.202-b08, mixed mode)

安装Hadoop

# 解压
tar -zxf /opt/hadoop-2.9.2-snappy-64.tar.gz -C /opt/

# 配置环境变量
vim /etc/profile

# HADOOP_HOME
export HADOOP_HOME=/opt/hadoop-2.9.2
export PATH=$PATH:$HADOOP_HOME/bin
export PATH=$PATH:$HADOOP_HOME/sbin

# 刷新环境变量
source /etc/profile

# 验证
hadoop version

# 自己编译的，显示可能不一样
# Hadoop 2.9.2
# Subversion Unknown -r Unknown
# Compiled by root on 2018-12-16T09:39Z
# Compiled with protoc 2.5.0
# From source with checksum 3a9939967262218aa556c684d107985
# This command was run using /opt/hadoop-2.9.2/share/hadoop/common/hadoop-common-2.9.2.jar

配置 Hadoop 伪分布式

一、配置 HDFS

hadoop-env.sh

vim /opt/hadoop-2.9.2/etc/hadoop/hadoop-env.sh

# 配置 JDK 路径
# The java implementation to use.
export JAVA_HOME=/opt/jdk1.8.0_202/

core-site.xml

<configuration>
    <!-- 指定HDFS中NameNode的地址 -->
    <property>
        <name>fs.defaultFS</name>
        <value>hdfs://hadoop:9000</value>
    </property>
    <!-- 指定Hadoop运行时产生文件的存储目录 -->
    <property>
        <name>hadoop.tmp.dir</name>
        <value>/opt/hadoopTmp</value>
    </property>
</configuration>

hdfs-site.xml

<configuration>
    <!-- 指定HDFS副本的数量 -->
    <property>
        <name>dfs.replication</name>
        <value>1</value>
    </property>
    <!-- 默认为true，namenode 连接 datanode 时会进行 host 解析查询 -->
    <property>
        <name>dfs.namenode.datanode.registration.ip-hostname-check</name>
        <value>true</value>
    </property>
</configuration>

启动 hdfs

# 第一次使用需要先格式化一次。之前若格式化过请先停止进程，然后删除文件再执行格式化操作
hdfs namenode -format

# 启动 namenode
hadoop-daemon.sh start namenode

# 启动 datanode
hadoop-daemon.sh start datanode

# 验证，查看 jvm 进程
jps

# 84609 Jps
# 84242 NameNode
# 84471 DataNode

浏览器访问 CentOS 的 IP 地址加端口号 (默认50070) 即可看到 web 端

二、配置 YARN

yarn-env.sh

vim /opt/hadoop-2.9.2/etc/hadoop/yarn-env.sh

# 配置 JDK 路径
# some Java parameters
export JAVA_HOME=/opt/jdk1.8.0_202/

yarn-site.xml

<configuration>
<!-- Site specific YARN configuration properties -->
    <!-- Reducer获取数据的方式 -->
    <property>
        <name>yarn.nodemanager.aux-services</name>
        <value>mapreduce_shuffle</value>
    </property>
    <!-- 指定YARN的ResourceManager的地址 -->
    <property>
        <name>yarn.resourcemanager.hostname</name>
        <value>hadoop</value>
    </property>
</configuration>

启动 yarn，需保证 hdfs 已启动

# 启动 resourcemanager
yarn-daemon.sh start resourcemanager

# 启动 nodemanager
yarn-daemon.sh start nodemanager

# 查看 JVM 进程
jps

# 1604 DataNode
# 1877 ResourceManager
# 3223 Jps
# 1468 NameNode
# 2172 NodeManager

浏览器访问 CentOS 的 IP 地址加端口号 (默认8088) 即可看到 web 端

三、配置 MapReduce

mapred-env.sh

vim /opt/hadoop-2.9.2/etc/hadoop/mapred-env.sh

# 配置 JDK 路径
export JAVA_HOME=/opt/jdk1.8.0_202/

# when HADOOP_JOB_HISTORYSERVER_HEAPSIZE is not defined, set it.

mapred-site.xml

# 复制一份
cp /opt/hadoop-2.9.2/etc/hadoop/mapred-site.xml.template /opt/hadoop-2.9.2/etc/hadoop/mapred-site.xml

# 编辑
vim /opt/hadoop-2.9.2/etc/hadoop/mapred-site.xml

<configuration>
    <!-- 指定MR运行在YARN上 -->
    <property>
        <name>mapreduce.framework.name</name>
        <value>yarn</value>
    </property>
</configuration>

运行一个 MapReduce 任务

# 计算圆周率
hadoop jar /opt/hadoop-2.9.2/share/hadoop/mapreduce/hadoop-mapreduce-examples-2.9.2.jar pi 10 100

# Job Finished in 26.542 seconds
# Estimated value of Pi is 3.14800000000000000000

浏览器访问 CentOS 的 IP 地址加端口号 (默认8088) 可以查看记录

其他配置

四、配置 jobhistory，打开历史记录

mapred-site.xml

<configuration>
    <!-- 历史服务器端地址 -->
    <property>
        <name>mapreduce.jobhistory.address</name>
        <value>hadoop:10020</value>
    </property>
    <!-- 历史服务器web端地址 -->
    <property>
        <name>mapreduce.jobhistory.webapp.address</name>
        <value>hadoop:19888</value>
    </property>
    <property>
        <name>yarn.log.server.url</name>
        <value>http://hadoop:19888/jobhistory/logs</value>
    </property>
</configuration>

# 启动 jobhistory
mr-jobhistory-daemon.sh start historyserver

# JVM 进程
jps

# 7376 NodeManager
# 6903 DataNode
# 18345 Jps
# 6797 NameNode
# 7086 ResourceManager
# 18254 JobHistoryServer

浏览器访问 CentOS 的 IP 地址加端口号 (默认19888) 即可看到 web 端

五、配置 log-aggregation，打开日志聚集，在 web 端可以查看运行详情

yarn-site.xml

<configuration>
    <!-- 开启日志聚集功能 -->
    <property>
        <name>yarn.log-aggregation-enable</name>
        <value>true</value>
    </property>
    <!-- 设置日志保留时间(7天) -->
    <property>
        <name>yarn.log-aggregation.retain-seconds</name>
        <value>604800</value>
    </property>
</configuration>

# 需要重启一遍服务

hadoop-daemon.sh stop namenode
hadoop-daemon.sh stop datanode
yarn-daemon.sh stop resourcemanager
yarn-daemon.sh stop nodemanager
mr-jobhistory-daemon.sh stop historyserver

hadoop-daemon.sh start namenode
hadoop-daemon.sh start datanode
yarn-daemon.sh start resourcemanager
yarn-daemon.sh start nodemanager
mr-jobhistory-daemon.sh start historyserver

# 再运行一个任务，就可以看到详情
hadoop jar /opt/hadoop-2.9.2/share/hadoop/mapreduce/hadoop-mapreduce-examples-2.9.2.jar pi 10 100

查看刚刚运行的任务详情，未开启日志聚集之前运行的任务无法查看详情

http://hadoop.apache.org/docs/current/hadoop-project-dist/hadoop-common/SingleCluster.html