安装Hadoop 2.9

image.png

因为要搞一波flink，再加上之前的hadoop版本有点旧了所以重写一下

环境：Centeros7 jdk8 hadoop2.9

1.配置环境变量和主机和ip对应

vi /etc/profile
JAVA_HOME=/home/java/jdk8
JRE_HOME=/home/java/jdk8/jre
HADOOP_HOME=/home/bigdata/hadoop
CLASS_PATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar:$JRE_HOME/lib
PATH=$PATH:$JAVA_HOME/bin:$JRE_HOME/bin:$HADOOP_HOME/bin
export JAVA_HOME JRE_HOME CLASS_PATH PATH HADOOP_HOME

source /etc/profile

vi etc/hosts

192.168.229.133 master
192.168.229.138 slave1
192.168.229.139 slave2

2.解压hadoop 并改名

tar -zxvf hadoop-2.9.2.tar.gz
mv hadoop-2.9.2 hadoop

进入hadoop的配置目录

cd hadoop/etc/hadoop

image.png

3.vi hadoop-env.sh

#第27行
export JAVA_HOME=/home/java/jdk8

4.vi core-site.xml

<configuration>
    <property>
        <name>fs.defaultFS</name>
        <value>hdfs://master:9000</value>
    </property>
    <property>
        <name>hadoop.tmp.dir</name>
        <value>/home/hadoop/tmp</value>
    </property>
</configuration>

**5.vi hdfs-site.xml **

    <configuration>
    <property>
        <name>dfs.namenode.secondary.http-address</name>
        <value>master:50090</value>
    </property>
    <property>
        <name>dfs.replication</name>
        <value>1</value>
    </property>
</configuration>

6.mv mapred-site.xml.template mapred-site.xml
vi mapred-site.xml

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

7.vi yarn-site.xml

        <configuration>
    <property>
        <name>yarn.resourcemanager.hostname</name>
        <value>master</value>
    </property>
    <property>
        <name>yarn.nodemanager.aux-services</name>
        <value>mapreduce_shuffle</value>
    </property>
</configuration>

8.集群搭建
vi slaves

slave1
slave2

9.clone两台机器
10.配置免密登陆这个自行百度
11.启动namenode在主节点上

hadoop namenode -format

12.启动hadoop
cd bin目录

./start-all.sh

13.检查各个节点是否正确
主节点

image.png
两个从节点
image.png

至此安装hadoop成功

QuorumPeerMain是zk的进程