259 lines
5.8 KiB
Markdown
259 lines
5.8 KiB
Markdown
# Hadoop单机版环境搭建
|
||
|
||
|
||
|
||
>系统环境：centos 7.6
|
||
>
|
||
>JDK版本：jdk 1.8.0_20
|
||
>
|
||
>Hadoop版本：hadoop-2.6.0-cdh5.15.2
|
||
|
||
|
||
|
||
<nav>
|
||
<a href="#一安装JDK">一、安装JDK</a><br/>
|
||
<a href="#二配置-SSH-免密登录">二、配置 SSH 免密登录</a><br/>
|
||
<a href="#三HadoopHDFS环境搭建">三、Hadoop(HDFS)环境搭建</a><br/>
|
||
<a href="#四HadoopYARN环境搭建">四、Hadoop(YARN)环境搭建</a><br/>
|
||
</nav>
|
||
|
||
|
||
|
||
## 一、安装JDK
|
||
|
||
Hadoop 需要在java环境下运行，所以需要先安装Jdk,安装步骤见[Linux下JDK的安装](https://github.com/heibaiying/BigData-Notes/blob/master/notes/installation/JDK%E5%AE%89%E8%A3%85.md)
|
||
|
||
|
||
|
||
## 二、配置 SSH 免密登录
|
||
|
||
Hadoop 组件之间的各个节点需要进行通讯，所以需要配置SSH免密登录。
|
||
|
||
#### 2.1 配置ip地址和主机名映射，在配置文件末尾添加ip地址和主机名映射
|
||
|
||
```shell
|
||
vim /etc/hosts
|
||
# 文件末尾增加
|
||
192.168.43.202 hadoop001
|
||
```
|
||
|
||
#### 2.2 执行下面命令行，一路回车，生成公匙和私匙
|
||
|
||
```
|
||
ssh-keygen -t rsa
|
||
```
|
||
|
||
#### 3.3 进入`~/.ssh`目录下，查看生成的公匙和私匙，并将公匙写入到授权文件
|
||
|
||
```shell
|
||
[root@@hadoop001 sbin]# cd ~/.ssh
|
||
[root@@hadoop001 .ssh]# ll
|
||
-rw-------. 1 root root 1675 3月 15 09:48 id_rsa
|
||
-rw-r--r--. 1 root root 388 3月 15 09:48 id_rsa.pub
|
||
```
|
||
|
||
```shell
|
||
# 写入公匙到授权文件
|
||
[root@hadoop001 .ssh]# cat id_rsa.pub >> authorized_keys
|
||
[root@hadoop001 .ssh]# chmod 600 authorized_keys
|
||
```
|
||
|
||
|
||
|
||
## 三、Hadoop(HDFS)环境搭建
|
||
|
||
|
||
|
||
#### 3.1 下载CDH 版本的Hadoop
|
||
|
||
从[CDH官方下载地址](http://archive.cloudera.com/cdh5/cdh/5/)下载所需版本的Hadoop（本用例下载的版本为hadoop-2.6.0-cdh5.15.2.tar.gz ）,上传至服务器对应文件夹（这里我传至新建的/usr/app/ 目录）；
|
||
|
||
|
||
|
||
#### 3.2 解压软件压缩包
|
||
|
||
```shell
|
||
# tar -zvxf hadoop-2.6.0-cdh5.15.2.tar.gz
|
||
```
|
||
|
||
|
||
|
||
#### 3.3 把路径添加到环境变量
|
||
|
||
```shell
|
||
# vi /etc/profile
|
||
```
|
||
|
||
```
|
||
export HADOOP_HOME=/usr/app/hadoop-2.6.0-cdh5.15.2
|
||
export PATH=${HADOOP_HOME}/bin:$PATH
|
||
```
|
||
|
||
```shell
|
||
# 执行source命令,使得配置立即生效
|
||
# source /etc/profile
|
||
```
|
||
|
||
|
||
|
||
#### 3.4 修改Hadoop相关配置文件
|
||
|
||
cd 到 安装目录的/etc/hadoop/ 文件夹下：
|
||
|
||
```shell
|
||
[root@hadoop001 hadoop-2.6.0-cdh5.15.2]# cd etc/hadoop
|
||
```
|
||
|
||
1. 修改 `hadoop-env.sh` , 指定jdk 安装路径
|
||
|
||
```shell
|
||
# The java implementation to use.
|
||
export JAVA_HOME=/usr/java/jdk1.8.0_201/
|
||
```
|
||
|
||
2. 修改`core-site.xml`，添加如下配置，指定hdfs地址：
|
||
|
||
```xml
|
||
<configuration>
|
||
<property>
|
||
<name>fs.defaultFS</name>
|
||
<value>hdfs://hadoop001:8020</value>
|
||
</property>
|
||
</configuration>
|
||
```
|
||
|
||
3. 修改`hdfs-site.xml`添加如下配置，指定副本系数和临时文件存储位置，由于这里我们搭建是单机版本，所以指定dfs的副本系数为1。
|
||
|
||
```xml
|
||
<configuration>
|
||
<property>
|
||
<name>dfs.replication</name>
|
||
<value>1</value>
|
||
</property>
|
||
<property>
|
||
<name>hadoop.tmp.dir</name>
|
||
<value>/usr/app/tmp</value>
|
||
</property>
|
||
</configuration>
|
||
```
|
||
|
||
4. 修改`slaves`文件，由于是单机版本，所以指定本机为从节点，修改后`slaves`文件内容如下：
|
||
|
||
```shell
|
||
[root@hadoop001 hadoop]# cat slaves
|
||
hadoop001
|
||
```
|
||
|
||
|
||
|
||
#### 3.5 关闭防火墙
|
||
|
||
由于防火墙可能会影响节点间通讯，所以建议关闭，执行命令：
|
||
|
||
```shell
|
||
# 查看防火墙状态
|
||
sudo firewall-cmd --state
|
||
# 关闭防火墙:
|
||
sudo systemctl stop firewalld.service
|
||
```
|
||
|
||
|
||
|
||
#### 3.6 启动HDFS
|
||
|
||
1. 第一次执行的时候一定要格式化文件系统，执行以下命令格式化文件系统（hdfs命令位于安装目录的bin目录下）。
|
||
|
||
```shell
|
||
[root@hadoop001 bin]# ./hdfs namenode -format
|
||
```
|
||
|
||
2. 执行sbin目录下的`start-dfs.sh`脚本，启动hdfs
|
||
|
||
```shell
|
||
[root@hadoop001 sbin]# ./start-dfs.sh
|
||
```
|
||
|
||
|
||
|
||
#### 3.7 验证是否启动成功
|
||
|
||
方式一：执行jps 查看NameNode和DataNode的进程是否已经存在
|
||
|
||
```shell
|
||
[root@hadoop001 hadoop-2.6.0-cdh5.15.2]# jps
|
||
9137 DataNode
|
||
9026 NameNode
|
||
9390 SecondaryNameNode
|
||
```
|
||
|
||
|
||
|
||
方式二：访问50070端口 http://ip地址:50070 。如果jps查看进程均以启动，但是无法访问页面，则需要关闭防火墙。
|
||
|
||
<div align="center"> <img width="700px" src="https://github.com/heibaiying/BigData-Notes/blob/master/pictures/hadoop安装验证.png"/> </div>
|
||
|
||
|
||
|
||
## 四、Hadoop(YARN)环境搭建
|
||
|
||
#### 4.1 修改Hadoop配置文件，指明mapreduce运行在YARN上
|
||
|
||
etc/hadoop/mapred-site.xml:
|
||
|
||
```shell
|
||
# 如果没有mapred-site.xml，则拷贝一份样例文件后再修改
|
||
cp ./mapred-site.xml.template ./mapred-site.xml
|
||
```
|
||
|
||
```xml
|
||
<configuration>
|
||
<property>
|
||
<name>mapreduce.framework.name</name>
|
||
<value>yarn</value>
|
||
</property>
|
||
</configuration>
|
||
```
|
||
|
||
etc/hadoop/yarn-site.xml:
|
||
|
||
```xml
|
||
<configuration>
|
||
<property>
|
||
<name>yarn.nodemanager.aux-services</name>
|
||
<value>mapreduce_shuffle</value>
|
||
</property>
|
||
<!--这个目录需要和上面配置的hadoop.tmp.dir的路径保持一致-->
|
||
<property>
|
||
<name>yarn.nodemanager.local-dirs</name>
|
||
<value>/usr/app/tmp/nm-local-dir</value>
|
||
</property>
|
||
</configuration>
|
||
```
|
||
|
||
|
||
|
||
#### 4.2 在sbin目录下启动YARN
|
||
|
||
```shell
|
||
./start-yarn.sh
|
||
```
|
||
|
||
|
||
|
||
#### 4.3 验证是否启动成功
|
||
|
||
方式一：执行jps查看NodeManager和ResourceManager的进程是否已经存在
|
||
|
||
```shell
|
||
[root@hadoop001 hadoop-2.6.0-cdh5.15.2]# jps
|
||
9137 DataNode
|
||
9026 NameNode
|
||
12294 NodeManager
|
||
12185 ResourceManager
|
||
9390 SecondaryNameNode
|
||
```
|
||
|
||
方式二：访问8088端口 http://ip地址:8088 。如果jps查看进程均以启动，但是无法访问页面，则需要关闭防火墙。
|
||
|
||
<div align="center"> <img width="700px" src="https://github.com/heibaiying/BigData-Notes/blob/master/pictures/hadoop-yarn安装验证.png"/> </div>
|