(1) Start the log nodes of the three machines: xiaoye@ubuntu:~$ hadoop/sbin/hadoop-daemons.sh start jorunalnode
(2) Start the namenode of ubuntu separately, pay attention to the daemon of the command without s. xiaoye@ubuntu:~$ hadoop/sbin/hadoop-daemon.sh start namenode
(3) Ubuntu2 node execution command: xiaoye@ubuntu2:~$ ./hadoop/bin/hdfs namenode -bootstrapStandby
(4) The ubuntu2 node starts namenode alone; xiaoye@ubuntu2:~$ ./hadoop/sbin/hadoop-daemon.sh start namemode
(5) ubuntu mobile owned datanode: xiaoye @ ubuntu: ~ $ ./hadoop/sbin/hadoop-daemons.sh start datanode
(6) ubuntu3 starts yarn resource management: xiaoye@ubuntu3:~$ ./hadoop/sbin/start-yarn.sh
(7) ubuntu starts all zkfc: xiaoye@ubuntu:~$ ./hadoop/sbin/hadoop-daemons.sh start zkfc