CentOS 6.7安装Hadoop 2.7.3
VMware建立虚拟机
配置Java环境
网上同样很多教程
克隆虚拟机
在修改master的hosts,
192.168.197.132 master-01 192.168.197.133 slave-01 192.168.197.134 slave-02
然后克隆matser,分别命名为slave1,slave2。
现在有三台虚拟机
IP 虚拟机名称 用户
192.168.197.132 master yang
192.168.197.133 slave1 yang
192.168.197.134 slave2 yang
ssh免登陆
(1)CentOS默认没有启动ssh无密登录，去掉/etc/ssh/sshd_config其中2行的注释，每台服务器都要设置，
#RSAAuthentication yes
#PubkeyAuthentication yes
安装ssh
在master-01的机器上进入 yang用户 的 .ssh 目录
使用 ssh-keygen -t rsa 来生成公钥和私钥(连续回车，不设置密码)
把公钥文件复制到要访问的机器的yang的用户目录下的.ssh 目录
scp ~/.ssh/id_rsa.pub yang@master-01:/home/yang/.ssh/authorized_keys
scp ~/.ssh/id_rsa.pub yang@slave-01:/home/yang/.ssh/authorized_keys
scp ~/.ssh/id_rsa.pub yang@slave-02:/home/yang/.ssh/authorized_keys
检测是否可以不需要密码登陆
ssh localhost
ssh yang@master-01
ssh yang@slave-01
ssh yang@slave-02
这里只有master-01是master，如果有多个namenode，或者rm的话则需要打通所有master到其他剩余节点的免密码登陆。（将master-01的authorized_keys追加到02和03的authorized_keys）
配置安装Hadoop 2.7.3
下载Hadoop-2.7.3
下载Hadoop 2.7.3并解压到/usr/software目录下，在hadoop-2.7.3目录下新建hdfs，hdfs/data，hdfs/name，hdfs/temp目录。
配置core-site.xml
<configuration>
<property>
<name>fs.defaultFS</name>
<value>hdfs://master-01:9000</value>
</property>
<property>
<name>hadoop.tmp.dir</name>
<value>file:/usr/software/hadoop-2.7.3/tmp</value>
</property>
<property>
<name>io.file.buffer.size</name>
<value>131702</value>
</property>
</configuration>配置mapred-site.xml
<configuration>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
<property>
<name>mapreduce.jobhistory.address</name>
<value>master-01:10020</value>
</property>
<property>
<name>mapreduce.jobhistory.webapp.address</name>
<value>master-01:19888</value>
</property>
</configuration>配置yarn-site.xml
<configuration>
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
<property>
<name>yarn.nodemanager.auxservices.mapreduce.shuffle.class</name>
<value>org.apache.hadoop.mapred.ShuffleHandler</value>
</property>
<property>
<name>yarn.resourcemanager.address</name>
<value>master-01:8032</value>
</property>
<property>
<name>yarn.resourcemanager.scheduler.address</name>
<value>master-01:8030</value>
</property>
<property>
<name>yarn.resourcemanager.resource-tracker.address</name>
<value>master-01:8031</value>
</property>
<property>
<name>yarn.resourcemanager.admin.address</name>
<value>master-01:8033</value>
</property>
<property>
<name>yarn.resourcemanager.webapp.address</name>
<value>master-01:8088</value>
</property>
<!--
<property>
<name>yarn.nodemanager.resource.memory-mb</name>
<value>768</value>
</property>-->
</configuration>设置slaves
修改hadoop-2.7.3/etc/hadoop下的slaves文件，添加我们之前建立好的两个slave
slave-01 slave-02
网上很多地方说需要设置hadoop-env.sh和yarn-env.sh的Java环境，我看了这两个文件的内容，已经帮我们配置好了，所以不用管。
配置完成
然后分别复制master下面的已经配置好的Hadoop-2.7.3到yang@slave-01和yang@slave02的/usr/software目录下。
启动
在Master服务器启动hadoop，从节点会自动启动，进入/usr/software/hadoop-2.7.3目录
(1)初始化，输入命令，bin/hdfs namenode -format
(2)启动sbin/start-dfs.sh,输出如下内容，则成功
Starting namenodes on [master-01] master-01: starting namenode, logging to /usr/software/hadoop-2.7.3/logs/hadoop-yang-namenode-master-01.out slave-01: starting datanode, logging to /usr/software/hadoop-2.7.3/logs/hadoop-yang-datanode-slave-01.out slave-02: starting datanode, logging to /usr/software/hadoop-2.7.3/logs/hadoop-yang-datanode-slave-02.out Starting secondary namenodes [master-01] master-01: starting secondarynamenode, logging to /usr/software/hadoop-2.7.3/logs/hadoop-yang-secondarynamenode-master-01.out
(3)sbin/start-yarn.sh,如下则成功
[yang@master-01 hadoop-2.7.3]$ ./sbin/start-yarn.sh starting yarn daemons starting resourcemanager, logging to /usr/software/hadoop-2.7.3/logs/yarn-yang-resourcemanager-master-01.out slave-02: starting nodemanager, logging to /usr/software/hadoop-2.7.3/logs/yarn-yang-nodemanager-slave-02.out slave-01: starting nodemanager, logging to /usr/software/hadoop-2.7.3/logs/yarn-yang-nodemanager-slave-01.out
(4)停止的话，输入命令，sbin/stop-dfs.sh,sbin/stop-yarn.sh
(5)输入命令，jps，可以看到相关信息
yang@master-01 hadoop-2.7.3]$ jps 6932 SecondaryNameNode 7384 Jps 6729 NameNode 7118 ResourceManager [yang@master-01 hadoop-2.7.3]$ ./bin/hdfs dfsadmin -report Configured Capacity: 75404550144 (70.23 GB) Present Capacity: 54191501312 (50.47 GB) DFS Remaining: 54191452160 (50.47 GB) DFS Used: 49152 (48 KB) DFS Used%: 0.00% Under replicated blocks: 0 Blocks with corrupt replicas: 0 Missing blocks: 0 Missing blocks (with replication factor 1): 0 ------------------------------------------------- Live datanodes (2): Name: 192.168.197.133:50010 (slave-01) Hostname: slave-01 Decommission Status : Normal Configured Capacity: 37702275072 (35.11 GB) DFS Used: 24576 (24 KB) Non DFS Used: 10606755840 (9.88 GB) DFS Remaining: 27095494656 (25.23 GB) DFS Used%: 0.00% DFS Remaining%: 71.87% Configured Cache Capacity: 0 (0 B) Cache Used: 0 (0 B) Cache Remaining: 0 (0 B) Cache Used%: 100.00% Cache Remaining%: 0.00% Xceivers: 1 Last contact: Tue Sep 27 17:18:44 CST 2016 Name: 192.168.197.134:50010 (slave-02) Hostname: slave-02 Decommission Status : Normal Configured Capacity: 37702275072 (35.11 GB) DFS Used: 24576 (24 KB) Non DFS Used: 10606292992 (9.88 GB) DFS Remaining: 27095957504 (25.24 GB) DFS Used%: 0.00% DFS Remaining%: 71.87% Configured Cache Capacity: 0 (0 B) Cache Used: 0 (0 B) Cache Remaining: 0 (0 B) Cache Used%: 100.00% Cache Remaining%: 0.00% Xceivers: 1 Last contact: Tue Sep 27 17:18:44 CST 2016
Web访问直接关闭防火墙
(1)浏览器打开http://192.168.197.132:8088/
(2)浏览器打开http://192.168.197.132:50070/
有如下信息：
Configured Capacity: 35.11 GB
DFS Used: 28 KB (0%)
Non DFS Used: 9.88 GB
DFS Remaining: 25.23 GB (71.87%)
Block Pool Used: 28 KB (0%)
DataNodes usages% (Min/Median/Max/stdDev): 0.00% / 0.00% / 0.00% / 0.00%
Live Nodes 1 (Decommissioned: 0)
Dead Nodes 1 (Decommissioned: 0)
Decommissioning Nodes 0
Total Datanode Volume Failures 0 (0 B)
Number of Under-Replicated Blocks 0
Number of Blocks Pending Deletion 0
Block Deletion Start Time 9/27/2016, 5:15:33 PM
总结
我在启动的时候总是出现错误，提示权限问题，后来发现我之前的操作是用root用户，然后hadoop-2.7.3的用户组也是root，yang这个用户根本没有权限，那么问题找到了就修改呗，chrown修改为yang，问题解决。当然配置的过程中出现各种问题，都是参照网上的办法和logs解决了。就不一一指出了，如果大家按照这个配置，还是有些问题，那么请多多百度，google吧。