Hadoop2.8 安装

一、下载Hadoop与java

jdk-8u221-linux-x64.tar.gz  Oracle官网下载

hadoop-2.8.5.tar.gz   Hadoop官网下载

二、配置服务期间ssh免密互通

使用如下互信安装脚本

https://www.cnblogs.com/xibuhaohao/p/11772047.html

三、配置服务期间时钟同步服务

略

四、解压安装Hadoop文件与Java

1、解压缩（每个结点都要做）

1）安装Java

2）安装Hadoop

新建Hadoop用户

2、配置结点环境变量

cat .bash_profile

添加如下：

export JAVA_HOME=/usr/java/jdk1.8.0_221
export CLASSPATH=.:$JAVA_HOME/jre/lib/rt.jar:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar
export PATH=$JAVA_HOME/jre/bin:$PATH

export HADOOP_HOME=/home/hadoop/hadoop-2.8.5
export PATH=$PATH:$JAVA_HOME/bin:$HADOOP_HOME/bin

五、修改配置文件

cd /home/hadoop/hadoop-2.8.5/etc/hadoop

1、core-site.xml

<configuration>

<!--配置hdfs文件系统的命名空间-->
  <property>
    <name>fs.defaultFS</name>
    <value>hdfs://data0:9000</value>
  </property>

<!-- 配置操作hdfs的存冲大小 -->
  <property>
    <name>io.file.buffer.size</name>
    <value>4096</value>
  </property>
<!-- 配置临时数据存储目录 -->
  <property>
    <name>hadoop.tmp.dir</name>
    <value>/home/hadoop/hadoop-2.8.5/tmp</value>
  </property>

</configuration>

2、hdfs-site.xml

<!--配置副本数-->
  <property>
    <name>dfs.replication</name>
      <value>3</value>
        </property>
        <!--hdfs的元数据存储位置-->
  <property>
    <name>dfs.namenode.name.dir</name>
      <value>/home/hadoop/hadoop-2.8.5/hdfs/name</value>
       </property>
       <!--hdfs的数据存储位置-->
 <property>
   <name>dfs.datanode.data.dir</name>
     <value>/home/hadoop/hadoop-2.8.5/hdfs/data</value>
      </property>
      <!--hdfs的namenode的web ui 地址-->
 <property>
   <name>dfs.http.address</name>
    <value>data0:50070</value>
      </property>
      <!--hdfs的snn的web ui 地址-->
  <!-- <property>
   <name>dfs.secondary.http.address</name>
     <value>data0:50090</value>
       </property>

-->
       <!--是否开启web操作hdfs-->
 <property>
   <name>dfs.webhdfs.enabled</name>
    <value>true</value>
      </property>
      <!--是否启用hdfs权限（acl）-->
  <property>
    <name>dfs.permissions</name>
      <value>false</value>
        </property>
</configuration>

3、mapred-site.xml

<configuration>

<!--指定maoreduce运行框架-->
        <property>
                        <name>mapreduce.framework.name</name>
                                        <value>yarn</value> </property>
                                        <!--历史服务的通信地址-->
        <property>
                        <name>mapreduce.jobhistory.address</name>
                                        <value>data0:10020</value>
                                                </property>
                                                <!--历史服务的web ui地址-->
        <property>
                        <name>mapreduce.jobhistory.webapp.address</name>
                                        <value>data0:19888</value>
                                                </property>

</configuration>

4、yarn-site.xml

<configuration>

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

<!--指定resourcemanager所启动的服务器主机名-->
        <property>
                        <name>yarn.resourcemanager.hostname</name>
                                        <value>data0</value>
                                                </property>
                                                <!--指定mapreduce的shuffle-->
        <property>
                        <name>yarn.nodemanager.aux-services</name>
                                        <value>mapreduce_shuffle</value>
                                                </property>
                                                <!--指定resourcemanager的内部通讯地址-->
        <property>
                        <name>yarn.resourcemanager.address</name>
                                        <value>data0:8032</value>
                                                </property>
                                                <!--指定scheduler的内部通讯地址-->
        <property>
                        <name>yarn.resourcemanager.scheduler.address</name>
                                        <value>data0:8030</value>
                                                </property>
                                                <!--指定resource-tracker的内部通讯地址-->
        <property>
                        <name>yarn.resourcemanager.resource-tracker.address</name>
                                        <value>data0:8031</value>
                                                </property>
                                                <!--指定resourcemanager.admin的内部通讯地址-->
        <property>
                        <name>yarn.resourcemanager.admin.address</name>
                                        <value>data0:8033</value>
                                                </property>
                                                <!--指定resourcemanager.webapp的ui监控地址-->
        <property>
                        <name>yarn.resourcemanager.webapp.address</name>
                                        <value>data0:8088</value>
                                                </property>

</configuration>

六、启动Hadoop

1、初始化Namenode

 hadoop namenode -format

2、开启namenode

./sbin/hadoop-daemon.sh start namenode

开启后使用jps验证

 

 

3、启动集群

./sbin/start-all.sh

4、查看集群个资源是否启动

jps

1）namenode

2)datanode1

3)datanode2

七、网页登录

1、namenode

data0:50070

2、Hadoop

data0:8088