Hadoop之——CentOS + hadoop2.5.2分布式环境配置

转载请注明出处:http://blog.csdn.net/l1028386804/article/details/51536051

一、基础环境准备

系统：(VMWare) CentOS-6.5-x86_64-bin-DVD1.iso
hadoop版本：hadoop-2.5.2
jdk版本：jdk-7u72-linux-x64.tar.gz

1.集群机器

三台测试集群，一个master(liuyazhuang-01),两个slave(liuyazhuang-02,liuyazhuang-03)

/etc/hosts192.168.1.112 liuyazhuang-01192.168.1.113 liuyazhuang-02192.168.1.114 liuyazhuang-03

注意不要保留127.0.0.1 localhost

配置同步到其他两台机器

scp /etc/hosts root@192.168.1.113:/etc/hostsscp /etc/hosts root@192.168.1.114:/etc/hosts

2. 设置linux上ssh是用户可以自动登录

$ ssh-keygen -t dsa -P '' -f ~/.ssh/id_dsa$ cat ~/.ssh/id_dsa.pub >> ~/.ssh/authorized_keys

3.java环境配置

1)JAVA_HOME为/usr/local/java/jdk1.7.0_72
2)/etc/profile文件追加以下内容

JAVA_HOME=/usr/local/jdk1.7.0_72CLASS_PATH=$JAVA_HOME/libPATH=$JAVA_HOME/bin:$PATHexport PATH JAVA_HOME CLASS_PATH

二、下载解压hadoop-2.5.2.tar.gz

hadoop@liuyazhuang-01:~/data$ pwd/home/hadoop/datahadoop@liuyazhuang-01:~/data$ wget http://mirror.bit.edu.cn/apache/hadoop/common/hadoop-2.5.2/hadoop-2.5.2.tar.gzhadoop@liuyazhuang-01:~/data$tar zxvf hadoop-2.5.2.tar.gz

三、配置环境变量

hadoop@liuyazhuang-01:~/data$gedit /etc/profile
追加内容如下：

#HADOOP VARIABLES STARTexport HADOOP_INSTALL=/home/hadoop/data/hadoop-2.5.2export PATH=$PATH:$HADOOP_INSTALL/binexport PATH=$PATH:$HADOOP_INSTALL/sbinexport HADOOP_MAPRED_HOME=$HADOOP_INSTALLexport HADOOP_COMMON_HOME=$HADOOP_INSTALLexport HADOOP_HDFS_HOME=$HADOOP_INSTALLexport YARN_HOME=$HADOOP_INSTALLexport HADOOP_COMMON_LIB_NATIVE_DIR=$HADOOP_INSTALL/lib/nativeexport HADOOP_OPTS="-Djava.library.path=$HADOOP_INSTALL/lib"#HADOOP VARIABLES END

使配置生效

hadoop@liuyazhuang-01:~/data$source /etc/profile

同时需要修改$HADOOP_HOME/etc/hadoop/hadoop-env.sh

export JAVA_HOME=/usr/local/java/jdk1.7.0_72

四、修改$HADOOP_HOME/etc/hadoop/core-site.xml

添加如下内容：

<property><name>fs.default.name</name><value>hdfs://liuyazhuang-01:9000</value></property><property><name>hadoop.tmp.dir</name><value>/home/hadoop/data/hadoop-2.5.2/hadoop-${user.name}</value></property>

五、修改$HADOOP_HOME/etc/hadoop/yarn-site.xml

添加如下内容：

<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.hostname</name><value>liuyazhuang-01</value></property>

更多yarn-site.xml参数配置可参考：
http://hadoop.apache.org/docs/r2.5.2/hadoop-yarn/hadoop-yarn-common/yarn-default.xml

六、修改$HADOOP_HOME/etc/hadoop/mapred-site.xml

默认没有mapred-site.xml文件，copy mapred-site.xml.template 一份为 mapred-site.xml即可

#cp etc/hadoop/mapred-site.xml.template ./etc/hadoop/mapred-site.xml

添加如下内容：

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

七、配置hdfs-site.xml (这里可以不配，采用默认参数)

/usr/local/hadoop/etc/hadoop/hdfs-site.xml
用来配置集群中每台主机都可用，指定主机上作为namenode和datanode的目录。

<property><name>dfs.name.dir</name><value>/home/hadoop/data/hadoop-2.5.2/name1,/home/hadoop/data/hadoop-2.5.2/name2</value></property><property><name>dfs.data.dir</name><value>/home/hadoop/data/hadoop-2.5.2/data1,/home/hadoop/data/hadoop-2.5.2/data2</value></property><property><name>dfs.replication</name><value>2</value></property>

八、配置salves

告诉hadoop 其他从节点，这样，只要主节点启动，他会自动启动其他机器上的nameNode dataNode 等等
编辑 $HADOOP_HOME/etc/hadoop/slaves
内容如下：

liuyazhuang-02liuyazhuang-03

九、同步同步该文件夹 到其他各个从主机上即可

因为我们使用ssh免登陆 不需要使用密码

hadoop@liuyazhuang-01:~/data/hadoop-2.5.2$scp -r /home/hadoop/data/hadoop-2.5.2 hadoop@192.168.1.113:/home/hadoop/data/hadoop-2.5.2hadoop@liuyazhuang-01:~/data/hadoop-2.5.2$scp -r /home/hadoop/data/hadoop-2.5.2 hadoop@192.168.1.114:/home/hadoop/data/hadoop-2.5.2

十、格式化hdfs

hadoop@liuyazhuang-01:~/data/hadoop-2.5.2$./bin/hdfs namenode -format

或者

hadoop@liuyazhuang-01:~ hadoop namenode -format

十一、启动hadoop集群

hadoop@liuyazhuang-01:~/data/hadoop-2.5.2$./sbin/start-dfs.shhadoop@liuyazhuang-01:~/data/hadoop-2.5.2$./sbin/start-yarn.sh

十二、浏览器查看

浏览器打开 http://liuyazhuang-01:50070/，会看到hdfs管理页面
浏览器打开 http://liuyazhuang-01:8088/，会看到hadoop进程管理页面
浏览器打开 http://liuyazhuang-01:8088/cluster 查看cluster情况

十三、验证（WordCount验证）

1.dfs上创建input目录

hadoop@liuyazhuang-01:~/data/hadoop-2.5.2$bin/hadoop fs -mkdir -p input

2.把hadoop目录下的README.txt拷贝到dfs新建的input里

hadoop@liuyazhuang-01:~/data/hadoop-2.5.2$bin/hadoop fs -copyFromLocal README.txt input

3.运行WordCount

hadoop@liuyazhuang-01:~/data/hadoop-2.5.2$bin/hadoop jar share/hadoop/mapreduce/sources/hadoop-mapreduce-examples-2.5.2-sources.jar org.apache.hadoop.examples.WordCount input output

4.运行完毕后，查看单词统计结果

hadoop@liuyazhuang-01:~/data/hadoop-2.5.2$bin/hadoop fs -cat output/*

假如程序的输出路径为output，如果该文件夹已经存在,先删除

hadoop@liuyazhuang-01:~/data/hadoop-2.5.2$bin/hadoop dfs -rmr output

参考资料：

Ubuntu14.04下安装Hadoop2.4.0 （单机模式）
http://www.cnblogs.com/kinglau/p/3794433.html

Ubuntu14.04下安装Hadoop2.4.0 （伪分布模式）
http://www.cnblogs.com/kinglau/p/3796164.html

伪分布模式下执行wordcount实例时报错解决办法
http://www.cnblogs.com/kinglau/p/3364928.html

Eclipse下搭建Hadoop2.4.0开发环境
http://www.cnblogs.com/kinglau/p/3802705.html

Hadoop学习三十：Win7 Eclipse调试Centos Hadoop2.2-Mapreduce
http://zy19982004.iteye.com/blog/2024467

hadoop2.5.0 centOS系列 分布式的安装 部署
http://my.oschina.net/yilian/blog/310189

Centos6.5源码编译安装Hadoop2.5.1
http://www.myhack58.com/Article/sort099/sort0102/2014/54025.htm

Hadoop MapReduce两种常见的容错场景分析
http://www.chinacloud.cn/show.aspx?id=15793&cid=17

hadoop 2.2.0集群安装
http://blog.csdn.net/bluishglc/article/details/24591185

Apache Hadoop 2.2.0 HDFS HA + YARN多机部署
http://blog.csdn.net/u010967382/article/details/20380387

Hadoop集群配置（最全面总结）
http://blog.csdn.net/hguisu/article/details/7237395

Hadoop hdfs-site.xml 配置项清单
http://he.iori.blog.163.com/blog/static/6955953520138107638208/
http://slaytanic.blog.51cto.com/2057708/1101111


Hadoop三种安装模式
http://blog.csdn.net/liumm0000/article/details/13408855