1.下載hadoop

http://hadoop.apache.org/releases.html

2.解壓下載的hadoop文件

tar -zxvf /HADOOP-VERSION-FULL-PATH.tar.gz

3.配置hadoop環(huán)境

# vim /etc/profile
##添加HADOOP_HOME路徑
# export HADOOP_HOME=/HADOOP-VERSION-FULL-PATH
# export PATH=$PATH:$HADOOP_HOME/bin:$HADOOP_HOME/sbin
# wq
# source /etc/profile
## 創(chuàng)建hadoop數(shù)據(jù)目錄
# sudo mkdir /var/hadoop
# sudo chmod 777 /var/hadoop

4.hadoop需要java環(huán)境的支持买喧，配置java環(huán)境略

5.修改hadoop配置文件，主要包括以下幾個文件

• core.site.xml

# vim /HADOOP-VERSION-FULL-PATH/etc/hadoop/core.site.xml

在configuration節(jié)點(diǎn)中添加以下內(nèi)容

    <property>
        <name>fs.defaultFS</name>
        <value>hdfs://hadoop:9000</value>
    </property>

• hdfs.site.xml文件

# vim /HADOOP-VERSION-FULL-PATH/etc/hadoop/hdfs.site.xml

在configuration節(jié)點(diǎn)中添加以下內(nèi)容

        <!-- 單節(jié)點(diǎn)hadoop不需要復(fù)制 -->
        <property>
                <name>dfs.replication</name>
                <value>1</value>
        </property>
        <property>
                <name>dfs.namenode.name.dir</name>
                <value>file:/var/hadoop/data/hdfs/nn</value>
        </property>
         <property>
                <name>fs.checkpoint.dir</name>
                <value>file:/var/hadoop/data/hdfs/snn</value>
        </property>
        <property>
                <name>fs.checkpoint.edits.dir</name>
                <value>file:/var/hadoop/data/doop/hdfs/snn</value>
        </property>
        <property>
                <name>dfs.datanode.data.dir</name>
                <value>file:/var/hadoop/data/hdfs/dn</value>
        </property>

• mapred-site.xml

# cp /HADOOP-VERSION-FULL-PATH/etc/hadoop/mapred-site.xml.template
/HADOOP-VERSION-FULL-PATH/etc/hadoop/mapred-site.xml

在configuration節(jié)點(diǎn)中添加以下內(nèi)容

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

• yarn-site.xml

# vim /HADOOP-VERSION-FULL-PATH/etc/hadoop/yarn-site.xml

在configuration節(jié)點(diǎn)中添加以下內(nèi)容

        <property>
                <name>yarn.nodemanager.aux-services</name>
                <value>mapreduce_shuffle</value>
        </property>
        <property>
                <name>yarn.nodemanager.aux-services.shuffle.class</name>
                <value>org.apache.hadoop.mapred.ShuffleHandler</value>
        </property>

• hadoop-env.sh

# vim /HADOOP-VERSION-FULL-PATH/etc/hadoop/hadoop-env.sh

覆蓋正確的JAVA_HOME

export JAVA_HOME=JAVA_HOME_PATH

6.格式化hdfs文件系統(tǒng)

# /HADOOP-VERSION-FULL-PATH/bin/hdfs namenode -format

7.啟動hdfs服務(wù)

# /HADOOP-VERSION-FULL-PATH/sbin/start-dfs.sh

8.啟動yran服務(wù)

# /HADOOP-VERSION-FULL-PATH/sbin/start-yarn.sh

9.通過web接口驗(yàn)證正在運(yùn)行的服務(wù)

http://127.0.0.1:50070
http://127.0.0.1:8088

10.配置ssh免秘鑰登錄

# cd ~/.ssh/                     # 若沒有該目錄，請先執(zhí)行一次ssh localhost  
# ssh-keygen -t rsa              # 會有提示受神，都按回車就可以  
# cat id_rsa.pub >> authorized_keys  # 加入授權(quán)  
# chmod 600 ./authorized_keys    # 修改文件權(quán)限