hadoop高可用集群规划

Hadoop集群部署前置规划：

1. 利用上海汉水信息技术服务有限公司提供的的HanWate_BigData以及HanWate_HDP2.6.1.0镜像作为底层实现。
2. 服务器共有九台，每台服务器有3G内存，1个CPU内核，2个20G硬盘，网络状态为桥接模式。
3. 网址和主机名对应如下：

   192.168.1.201 kmaster1
   192.168.1.202 kmaster2
   192.168.1.203 kslave1
   192.168.1.204 kslave2
   192.168.1.205 kslave3
   192.168.1.206 kslave4
   192.168.1.207 kslave5
   192.168.1.208 kslave6
   192.168.1.209 kslave7

4. 集群规划　　　　　　　　　　　　 
   kmaster1:==》namenode,HMaster,historyserver,resourcemanager,zkfc(zookeeper控制器)
   kmaster2:==》namenode,HMaster,resourcemanager,zkfc
   kslave1:===》datanode,nodemanager,HRegionserver,
   kslave2:===》datanode,nodemanager,HRegionserver,
   kslave3:===》datanode,nodemanager,HRegionserver,
   kslave4:===》datanode,nodemanager,HRegionserver,
   kslave5:===》datanode,nodemanager,zookeeper,journalnode,HRegionserver,
   kslave6:===》datanode,nodemanager,zookeeper,journalnode,HRegionserver,
   kslave7:===》datanode,nodemanager,zookeeper,journalnode,HRegionserver,