Debian下Hadoop 3.12 集群搭建

Debian系统配置

我这里在Vmware里面虚拟4个Debian系统，一个master，三个solver。hostname分别是master、solver1、solver2、solver3。对了，下面的JDK和hadoop安装配置操作都是使用hadoop用户权限来执行，并非root权限。

1. 静态网络的配置

编辑/etc/network/interfaces文件，注释自动获取IP，并添加下面内容

# The primary network interface
#allow-hotplug ens33
#iface ens33 inet dhcp

# static IP address
auto ens33
iface ens33 inet static
address 192.168.20.101
netmask 255.255.255.0
gateway 192.168.20.2
dns-nameservers 192.168.20.2
dns-nameservers 114.114.114.114

2. 修改/etc/hosts文件，添加如下内容

# Hadoop
192.168.20.101  master
192.168.20.102  solver1
192.168.20.103  solver2
192.168.20.104  solver3

3. openssh-server安装和vim的安装

sudo apt-get install openssh-server vim

4. 生成ssh密钥

# 分别在不同的主机上执行`ssh-keygen`命令

# master
ssh-keygen -t rsa -C "master"

# solver1
ssh-keygen -t rsa -C "solver1"

# solver2
ssh-keygen -t rsa -C "solver2"

# solver3
ssh-keygen -t rsa -C "solver3"

5. 免密码登录

# 在每台主机上执行：
ssh-copy-id -i ~/.ssh/id_rsa.pub master
ssh-copy-id -i ~/.ssh/id_rsa.pub solver1
ssh-copy-id -i ~/.ssh/id_rsa.pub solver2
ssh-copy-id -i ~/.ssh/id_rsa.pub solver3

6. 创建用户和用户组

# 在每台主机上执行：
useradd -m -s /bin/bash hadoop

JDK 安装与配置

1. 手动安装JDK

解压jdk安装包到/usr/lib/jvm/，然后创建jdk软链接:

sudo ln -sf /usr/lib/jvm/jdk1.8.0_202 /usr/lib/jvm/jdk

2. JDK环境变量的配置

• 新建jdk.sh文件

vi /etc/profile.d/jdk.sh

• 添加如下内容：

# JDK environment settings
export JAVA_HOME=/usr/lib/jvm/jdk
export JRE_HOME=${JAVA_HOME}/jre
export CLASSPATh=.:${JAVA_HOME}/lib:${JRE_HOME}/lib
export PATH=${JAVA_HOME}/bin:$PATH

• JAVA环境的验证

$ 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)

把jdk安装包和jdk.sh分别scp到每台主机上，重复上面的操作。

Hadoop 安装与配置

Hadoop 安装

1. 解压hadoop安装包到/opt，修改hadoop-3.1.2的拥有者:

sudo chown -R hadoop:hadoop /opt/hadoop-3.1.2

2. 然后创建hadoop软链接

sudo ln -sf /opt/hadoop-3.1.2 /opt/hadoop

3. 在hadoop下创建logs、hdfs/name、hdfs/data文件夹

mkdir /opt/hadoop/logs
mkdir -p /opt/hadoop/hdfs/name
mkdir -p /opt/hadoop/hdfs/data

4. hadoop环境变量的配置

• 新建文件hadoop.sh

vi /etc/profile.d/hadoop.sh

• 添加如下内容：

# Hadoop environment settings
export HADOOP_HOME=/opt/hadoop
export PATH=$PATH:$HADOOP_HOME/bin:$HADOOP_HOME/sbin

• 刷新profile变量

# 使profile生效
source /etc/profile

Hadoop文件配置

配置文件都在etc/hadoop/文件夹下

1. hadoop-env.sh

# jdk环境变量 （因为要远程调用 ${java_home}找不到变量）
export JAVA_HOME=/usr/lib/jvm/jdk

2. workers

# 添加所有solver机器的hostname
solver1
solver2
solver3

3. core-site.xml

<configuration> 

  <!-- hdfs的位置 -->
  <property>
    <name>fs.defaultFS</name>
    <value>hdfs://master:9000</value>
  </property>  

  <!-- hadoop运行时产生的缓冲文件存储位置 -->
  <property>
    <name>hadoop.tmp.dir</name>
    <value>/opt/hadoop/tmp</value>
  </property> 

</configuration>

4. hdfs-site.xml

<configuration> 

  <!-- hdfs 数据备份数量 -->
  <property>
    <name>dfs.replication</name>
    <value>1</value>
  </property>  

  <!-- hdfs namenode上存储hdfs名字空间元数据 -->
  <property>
    <name>dfs.namenode.name.dir</name>
    <value>/opt/hadoop/hdfs/name</value>
  </property>  

  <!-- hdfs datanode上数据块的物理存储位置 -->
  <property>
    <name>dfs.datanode.data.dir</name>
    <value>/opt/hadoop/hdfs/data</value>
  </property> 

</configuration>

5. mapred-site.xml

<configuration> 

  <!--  mapreduce运行的平台 默认local本地模式 -->
  <property>
    <name>mapreduce.framework.name</name>
    <value>yarn</value>
  </property>  

  <!--  mapreduce web UI address -->
  <property>
    <name>mapreduce.jobhistory.webapp.address</name>
    <value>master:19888</value>
  </property> 

</configuration>

6. yarn-site.xml

<configuration>
  <!-- Site specific YARN configuration properties --> 

  <!--  yarn 的 hostname -->
  <property>
    <name>yarn.resourcemanager.hostname</name>
    <value>master</value>
  </property>  

  <!--  yarn Web UI address -->
  <property>
    <name>yarn.resourcemanager.webapp.address</name>
    <value>${yarn.resourcemanager.hostname}:8088</value>
  </property>  

  <!--  reducer 获取数据的方式 -->
  <property>
    <name>yarn.nodemanager.aux-services</name>
    <value>mapreduce_shuffle</value>
  </property> 

</configuration>

把/opt/hadoop-3.1.2和hadoop.sh打包scp到每台电脑上，然后重复Hadoop安装步骤

Hadoop 的验证

• 首先格式化 hdfs

hdfs namenode -format

• 启动与关闭 jobhistoryserver

mr-jobhistory-daemon.sh start historyserver
mr-jobhistory-daemon.sh stop historyserver

• 启动与关闭 yarn

start-yarn.sh
stop-yarn.sh

• 启动与关闭 hdfs

start-dfs.sh
stop-dfs.sh

• 一键启动与关闭

start-all.sh
stop-all.sh

• 验证

$ jps
13074 SecondaryNameNode
14485 Jps
10441 JobHistoryServer
12876 NameNode
13341 ResourceManager

访问Web UI

Daemon	Web Interface	Notes
NameNode	https://192.168.20.101:9870	Default HTTP port is 9870.
Resourcemanager	http://192.168.20.101:8088	Default HTTP port is 8088.
MapReduce JobHistory Server	http://192.168.20.101:19888	Default HTTP port is 19888.