02大数据学习之hadoop搭建——hadoop-2.6.1

前言——hadoop搭建2.x和1.x有区别，这里不做阐述。介绍2.6.1搭建的过程，主要是配置文件中参数的配置需要的工具包——jdk1.7,hadoop-2.6.1,集群环境是上篇博客中自己搭建的虚拟集群：

                

1、安装包下载：方法自行get

2、先安装jdk,

将压缩包解压：tar zxvf jdk-7u67-linux-x64.tar.gz配置jdk环境变量：master、slave1、slave2：vim ~/.bashrc     将JDK分别拷贝到从节点上：         scp -r /usr/local/src/jdk1.7.0_67 root@slave1:/usr/local/src/jdk1.7.0_67         scp -r /usr/local/src/jdk1.7.0_67 root@slave2:/usr/local/src/jdk1.7.0_67

3、hadoop安装：

解压安装包到指定目录下：tar zxvf hadoop-2.6.1.tar.gz    修改Hadoop配置文件：master :cd hadoop-2.6.1/etc/hadoop   vim hadoop-env.sh                      export JAVA_HOME=/usr/local/src/jdk1.7.0_67           vim yarn-env.sh                     export JAVA_HOME=/usr/local/src/ jdk1.7.0_67                                vim slaves                     slave1                     slave2           vim core-site.xml                         <configuration>                     <property>                     <name>fs.defaultFS</name>                     <value>hdfs://192.168.142.10:9000</value>                     </property>                     <property>                     <name>hadoop.tmp.dir</name>                     <value>file:/usr/local/src/hadoop-2.6.1/tmp</value>                     </property>                     </configuration>         vim hdfs-site.xml                         <configuration>                         <property>                         <name>dfs.namenode.secondary.http-address</name>                         <value>master:9001</value>                         </property>                         <property>                         <name>dfs.namenode.name.dir</name>                         <value>file:/usr/local/src/hadoop-2.6.1/dfs/name</value>                         </property>                         <property>                         <name>dfs.datanode.data.dir</name>                         <value>file:/usr/local/src/hadoop-2.6.1/dfs/data</value>                         </property>                         <property>                         <name>dfs.repliction</name>                         <value>3</value>                         </property>                         </configuration>         vim mapred-site.xml                     <configuration>                     <property>                     <name>mapreduce.framework.name</name>                     <value>yarn</value>                     </property>                     </configuration>         vim yarn-site.xml                     <configuration>                 <property>                 <name>yarn.nodemanager.aux-services</name>                 <value>mapreduce_shuffle</value>                 </property>            <property>                 <name>yarn.nodemanager.aux-services.mapreduce.shuffle.class</name>                 <value>org.apache.hadoop.mapred.ShuffleHandler</value>                  </property>            <property>                 <name>yarn.resourcemanager.address</name>                 <value>master:8032</value>                  </property>             <property>                 <name>yarn.resourcemanager.scheduler.address</name>                 <value>master:8030</value>                   </property>              <property>                 <name>yarn.resourcemanager.resource-tracker.address</name>                 <value>master:8035</value>                  </property>            <property>                 <name>yarn.resourcemanager.admin.address</name>                 <value>master:8033</value>                  </property>             <property>                 <name>yarn.resourcemanager.webapp.address</name>                 <value>master:8088</value>                   </property>             </configuration> #创建临时目录和文件目录         mkdir /usr/local/src/hadoop-2.6.1/tmp         mkdir -p /usr/local/src/hadoop-2.6.1/dfs/name         mkdir -p /usr/local/src/hadoop-2.6.1/dfs/data

配置环境变量：master、slave1、slave2:vim ~/.bashrc  HADOOP_HOME=/usr/local/src/hadoop-2.6.1 export PATH=$PATH:$HADOOP_HOME/bin

    拷贝安装包：

            scp -r /usr/local/src/hadoop-2.6.1 root@slave1:/usr/local/src/hadoop-2.6.1             scp -r /usr/local/src/hadoop-2.6.1 root@slave2:/usr/local/src/hadoop-2.6.1

启动集群         #Master         #初始化Namenode         hadoop namenode -format //只能初始化一次，如果安装的有问题，需要将三个节点中dfs下的文件都删掉。 rm -rf /home/hadoop/hadoop/dfs        启动hadoop集群：./sbin/start-all.sh

        

    查看master进程状态：jps

    查看slave1、slave2进程状态：jps

    关闭集群：./sbin/stop-all.sh

总结：在这个过程中一定要细致，先配置主节点，然后再拷贝到从节点，自己在配置过程中就是遇到了ssh互信的问题，导致有的节点无法启动，后来从新互信以后就解决了connection time out，已及refused的问题，在启动集群以后试试能不能put上文件：hadoop fs -put filename /filedir

这里就是整个hadoop集群搭建的过程！