基于hadoop2.2的namenode与SecondaryNameNode分开配置在不同的计算机

前段时间由于配置hadoop做实验，在网上看了许多有关hadoop的配置，但是这些配置多数是将namenode和secondaryNameNode配置在同一台计算机上，这种配置方法如果是做做实验的还可以，如果应用到实际中，存在较大风险，
如果存放namenode的主机出现问题，整个文件系统将被破坏，严重的情况是所有文件都丢失。现在来配置hadoop2.2,将namenode和secondaryNameNode配置在不同的机器上，这样的实用价值更大。


1.假设我们已经配置好了hadoop2.2,这时namenode和secondaryNameNode在同一台机器上，如cloud001;


2.单独配置secondaryNameNode文件，指明secondaryNameNode所在的计算机地址，需要masters文件（类似slaves文件指明从节点），而我下载的hadoop2.2配置文件中不包含masters,
所以我们新建一个masters文件，编辑内容为：


sudo vim masters  


cloud002  


此处将cloud002作为secondaryNameNode的主机。
3.修改hdfs-site.xml的内容


<!--  
 <property>  
  <name>dfs.namenode.secondary.http-address</name>  
  <value>cloud001:9001</value>  
 </property>  
-->  


注释掉之前使namenode与SecondaryNameNode在同一台主机的代码，并新添如下代码


<property>  
<name>dfs.http.address</name>  
<value>cloud001:50070</value>  
<description>  
The address and the base port where the dfs namenode web ui will listen on.  
If the port is 0 then the server will start on a free port.  
</description>  
</property>  
<property>  
<name>dfs.namenode.secondary.http-address</name>  
<value>cloud002:50090</value>  
</property>  


网上也有说要修改core-site.xml的代码，而我之前配置过，所以现在就不需要再改了。修改的内容如下：


<property>  
  <name>fs.checkpoint.period</name>  
  <value>60</value>  
  <description>The number of seconds between two periodic checkpoints.  
  </description>  
</property>  
<property>  
  <name>fs.checkpoint.size</name>  
  <value>67108864</value>  
</property>  
4.以上修改只是在namenode上修改，在集群中的其他节点也要作相应修改，简便方法便是将需要修改的文件覆盖掉。
5.启动hdfs,在hadoop目录下执行：




sbin/start-dfs.sh  




xuhui@cloud001:~/hadoop-2.2.0$ sbin/start-all.sh   
This script is Deprecated. Instead use start-dfs.sh and start-yarn.sh  
Starting namenodes on [cloud001]  
cloud001: starting namenode, logging to /home/xuhui/hadoop-2.2.0/logs/hadoop-xuhui-namenode-cloud001.out  
cloud004: starting datanode, logging to /home/xuhui/hadoop-2.2.0/logs/hadoop-xuhui-datanode-cloud004.out  
cloud002: starting datanode, logging to /home/xuhui/hadoop-2.2.0/logs/hadoop-xuhui-datanode-cloud002.out  
Starting secondary namenodes [cloud002]  
cloud002: starting secondarynamenode, logging to /home/xuhui/hadoop-2.2.0/logs/hadoop-xuhui-secondarynamenode-cloud002.out  
starting yarn daemons  
starting resourcemanager, logging to /home/xuhui/hadoop-2.2.0/logs/yarn-xuhui-resourcemanager-cloud001.out  
cloud004: starting nodemanager, logging to /home/xuhui/hadoop-2.2.0/logs/yarn-xuhui-nodemanager-cloud004.out  
cloud002: starting nodemanager, logging to /home/xuhui/hadoop-2.2.0/logs/yarn-xuhui-nodemanager-cloud002.out  
xuhui@cloud001:~/hadoop-2.2.0$ ls  


6.检查配置成功与否：


xuhui@cloud001:~/hadoop-2.2.0$ jps  
3490 ResourceManager  
3123 NameNode  
4784 Jps  
xuhui@cloud001:~/hadoop-2.2.0$   




xuhui@cloud002:~$ jps  
7162 SecondaryNameNode  
7412 NodeManager  
6624 DataNode  
7698 Jps  
xuhui@cloud002:~$   






xuhui@cloud004:~$ jps  
6287 NodeManager  
6409 Jps  
5836 DataNode  
xuhui@cloud004:~$   


验证结果很显然是配置成功的！
7.关闭hdfs:


xuhui@cloud001:~/hadoop-2.2.0$ sbin/stop-dfs.sh   
Stopping namenodes on [cloud001]  
cloud001: stopping namenode  
cloud004: stopping datanode  
cloud002: stopping datanode  
Stopping secondary namenodes [cloud002]  
cloud002: stopping secondarynamenode  
xuhui@cloud001:~/hadoop-2.2.0$