0. Hadoop分布式环境搭建之环境介绍

我这里准备了三台机器，IP地址如下：

192.168.153.113 
192.168.153.186 
192.168.153.223

1. 网络配置

首先在这三台机器上修改主机名以及配置其他机器的主机名

1.1 修改主机名(三台机器都要修改)

[[email protected] ~]# hostnamectl set-hostname bigdata100

1.2 编辑/etc/hosts配置文件(三台机器都要修改)

[[email protected] ~]# vim /etc/hosts

修改为如下

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

192.168.153.113 bigdata100
192.168.153.186 bigdata101
192.168.153.223 bigdata102

1.3 配置ssh免密码登录(三台机器都要修改)

[[email protected] ~]# ssh-keygen -t rsa  # 三台机器都需要执行这个命令生成密钥对

[[email protected] ~]# ls .ssh/
authorized_keys  id_rsa  id_rsa.pub  known_hosts

[[email protected] ~]# ssh-copy-id -i ~/.ssh/id_rsa.pub bigdata100
[[email protected] ~]# ssh-copy-id -i ~/.ssh/id_rsa.pub bigdata101
[[email protected] ~]# ssh-copy-id -i ~/.ssh/id_rsa.pub bigdata102

1.4 测试

[[email protected] ~]# ping bigdata100
[[email protected] ~]# ping bigdata101
[[email protected] ~]# ping bigdata102
[[email protected] ~]# ssh bigdata101
[[email protected] ~]# ssh bigdata102

2. JDK安装与配置

下载压缩包 jdk-8u151-linux-x64.tar.gz

2.1 解压

[[email protected] ~]# tar -zxvf jdk-8u151-linux-x64.tar.gz

2.2 移动到自己目录(这里我的目录是bigDataSoftware)

[[email protected] ~]# mv ./jdk1.8.0_151 /home/bigDataSoftware/JDK_1.8

2.3 配置环境变量

[[email protected] ~]# vim /etc/profile

在最后添加

export JAVA_HOME=/home/bigDataSoftware/JDK_1.8
export JRE_HOME=/home/bigDataSoftware/JDK_1.8/jre
export CLASSPATH=.:$JAVA_HOME/lib:$JRE_HOME/lib:$CLASSPATH
export PATH=$JAVA_HOME/bin:$JRE_HOME/bin:$JAVA_HOME:$PATH

加载配置文件，让其生效

[[email protected] ~]# source /etc/profile

2.4 测试

[[email protected] ~]# java -version
[[email protected] ~]# javac -version

2.5 JDK及环境变量分发

[[email protected] ~]# rsync -av /home/bigDataSoftware/JDK_1.8 bigdata101:/home/bigDataSoftware
[[email protected] ~]# rsync -av /home/bigDataSoftware/JDK_1.8 bigdata101:/home/bigDataSoftware
[[email protected] ~]# rsync -av /etc/profile bigdata101:/etc/profile
[[email protected] ~]# rsync -av /etc/profile bigdata101:/etc/profile

3. Hadoop安装与配置

3.1 将Hadoop压缩包拖入解压

[[email protected] /home/bigDataSoftware]# tar -zxvf hadoop-2.7.7.tar.gz

3.2 在hadoop-2.7.7文件夹里面先创建4个文件夹：

hadoop-2.7.7/hdfs
hadoop-2.7.7/hdfs/tmp
hadoop-2.7.7/hdfs/name
hadoop-2.7.7/hdfs/data

3.3 配置Hadoop文件

进入/home/bigDataSoftware/hadoop-2.7.7/etc/hadoop

3.3.1 配置core-site.xml文件

vim core-site.xml

在中加入以下代码：

<property>
    <name>hadoop.tmp.dir</name>
    <value>file:/home/bigDataSoftware/hadoop-2.7.7/hdfs/tmp</value>
    <description>A base for other temporary directories.</description>
</property>
<property>
    <name>io.file.buffer.size</name>
    <value>131072</value>
</property>
<property>
    <name>fs.defaultFS</name>
    <value>hdfs://bigdata100:9000</value>
</property>

注意：第一个属性中的value和我们之前创建的/hadoop-2.7.7/hdfs/tmp路径要一致。

3.3.2 配置 hadoop-env.sh文件

将JAVA_HOME文件配置为本机JAVA_HOME路径

# The java implementation to use.
export JAVA_HOME=/home/bigDataSoftware/JDK_1.8

3.3.3 配置 yarn-env.sh

将其中的JAVA_HOME修改为本机JAVA_HOME路径(先把这一行的#去掉)

# some Java parameters
export JAVA_HOME=/home/bigDataSoftware/JDK_1.8
if [ "$JAVA_HOME" != "" ]; then

3.3.4 配置hdfs-site.xml

在中加入以下代码

<property>
 <name>dfs.replication</name>
   <value>2</value>
 </property>
 <property>
   <name>dfs.namenode.name.dir</name>
   <value>file:/home/bigDataSoftware/hadoop-2.7.7/hdfs/name</value>
   <final>true</final>
</property>
 <property>
   <name>dfs.datanode.data.dir</name>
   <value>file:/home/bigDataSoftware/hadoop-2.7.7/hdfs/data</value>
   <final>true</final>
 </property>
 <property>
  <name>dfs.namenode.secondary.http-address</name>
   <value>bigdata100:9001</value>
 </property>
 <property>
   <name>dfs.webhdfs.enabled</name>
   <value>true</value>
 </property>
 <property>
   <name>dfs.permissions</name>
   <value>false</value>
 </property>

因为这里有两个从主机，所以dfs.replication设置为2

3.3.5 复制mapred-site.xml.template文件，并命名为mapred-site.xml

配置 mapred-site.xml,在标签中添加以下代码

<property>
   <name>mapreduce.framework.name</name>
   <value>yarn</value>
</property>

3.3.6 配置yarn-site.xml

在标签中添加以下代码

<property>
 <name>yarn.resourcemanager.address</name>
   <value>bigdata100:18040</value>
 </property>
 <property>
   <name>yarn.resourcemanager.scheduler.address</name>
   <value>bigdata100:18030</value>
 </property>
 <property>
   <name>yarn.resourcemanager.webapp.address</name>
   <value>bigdata100:18088</value>
 </property>
 <property>
   <name>yarn.resourcemanager.resource-tracker.address</name>
   <value>bigdata100:18025</value>
 </property>
 <property>
   <name>yarn.resourcemanager.admin.address</name>
   <value>bigdata100:18141</value>
 </property>
 <property>
    <name>yarn.nodemanager.aux-services</name>
    <value>mapreduce_shuffle</value>
 </property>
 <property>
     <name>yarn.nodemanager.auxservices.mapreduce.shuffle.class</name>
     <value>org.apache.hadoop.mapred.ShuffleHandler</value>
 </property>

3.3.7 配置slaves 文件

修改为如下

bigdata101
bigdata102

3.3.8 配置hadoop的环境，就像配置jdk一样

在/etc/profile文件最后添加如下

export HADOOP_HOME=/home/bigDataSoftware/hadoop-2.7.7
export PATH=$HADOOP_HOME/bin:$HADOOP_HOME/sbin:$PATH
export HADOOP_CONF_DIR=$HADOOP_HOME/etc/hadoop

3.3.9 Hadoop分发

rsync -av /home/bigDataSoftware/hadoop-2.7.7/ bigdata101:/home/bigDataSoftware/hadoop-2.7.7/
rsync -av /home/bigDataSoftware/hadoop-2.7.7/ bigdata102:/home/bigDataSoftware/hadoop-2.7.7/

环境变量配置文件分发

rsync -av /etc/profile bigdata101:/etc/profile
rsync -av /etc/profile bigdata102:/etc/profile

分别在其他两台主机上source

3.4 Hadoop格式化及启停

hdfs namenode -format

hadoop启动

start-all.sh 

查看是否有一下进程启动

[[email protected] ~]# jps
6256 Jps
5843 ResourceManager
5413 NameNode
5702 SecondaryNameNode

[[email protected] ~]# jps
3425 DataNode
3538 NodeManager
3833 Jps

[[email protected] ~]# jps
3425 DataNode
3538 NodeManager
3833 Jps

hadoop关闭

stop-all.sh

3.5 web管理

各机器的进程检查完成，并且确定没有问题后，在浏览器上访问主节点的50070端口，例如：192.168.153.115:50070。会访问到如下页面：

查看存活的节点：

如上，可以访问50070端口就代表集群中的HDFS是正常的。

接下来我们还需要访问主节点的8088端口，这是YARN的web服务端口，例如：192.168.153.115:18088。如下：

参考：
https://cloud.tencent.com/developer/article/1703113
https://blog.csdn.net/fanxin_i/article/details/80425461

问题：
hadoop启动start-all.sh,slave节点没有datanode的问题 https://blog.csdn.net/long085/article/details/78535873