脚本分析start-all.sh
1）首先进入/soft/hadoop/sbin目录
2）nano start-all.sh
sbin/start-all.sh主要做的事情：
1) libexec/hadoop-config.sh
2) start-dfs.sh
3) start-yarn.sh
3)cat libexec/hadoop-config.sh hadoop_conf_dir =etc/hadoop
4)cat sbin/start-dfs.sh
sbin/start-dfs.sh的作用：
1)libexec/hadoop-config.sh
2)sbin/hadoop-daemons.sh --config .. --hostname .. start namenode ...
3)sbin/hadoop-daemons.sh --config .. --hostname .. start datanode ...
4)sbin/hadoop-daemons.sh --config .. --hostname .. start sescondarynamenode ...
5)sbin/hadoop-daemons.sh --config .. --hostname .. start zkfc ...//
5)cat sbin/start-yarn.sh
libexec/yarn-config.sh作用
sbin/yarn-daemon.sh start resourcemanager
sbin/yarn-daemons.sh start nodemanager
6）cat sbin/hadoop-daemons.sh
sbin/hadoop-daemons.sh的作用：
libexec/hadoop-config.sh
slaves
hadoop-daemon.sh
7）cat sbin/hadoop-daemon.sh
sbin/hadoop-daemon.sh的作用
libexec/hadoop-config.sh
bin/hdfs ...
8）sbin/yarn-daemon.sh
sbin/yarn-daemon.sh作用
libexec/yarn-config.sh
bin/yarn
Start-all.sh ===config.sh----start-dfs.sh 和start-yarn.sh
Start-dfs.sh -----config.sh----(namenode,send,datanode)
Stop-dfs.sh------config---------------------------------------
hadoop-daemon.sh ----------->