Question

cygwin 的$ bin/start-all.sh 有问题啊，求解，我在conf/hadoop-env.sh中设了JAVA_HOME变量。

$ bin/start-all.sh
starting namenode, logging to /home/WPS/hadoop/bin/../logs/hadoop-WPS-namenode-WPS-PC.out
localhost: starting datanode, logging to /home/WPS/hadoop/bin/../logs/hadoop-WPS-datanode-WPS-PC.out
localhost: starting secondarynamenode, logging to /home/WPS/hadoop/bin/../logs/hadoop-WPS-secondarynamenode-WPS-PC.out
localhost: Error: JAVA_HOME is not set.
starting jobtracker, logging to /home/WPS/hadoop/bin/../logs/hadoop-WPS-jobtracker-WPS-PC.out
localhost: starting tasktracker, logging to /home/WPS/hadoop/bin/../logs/hadoop-WPS-tasktracker-WPS-PC.out
localhost: Error: JAVA_HOME is not set.

Answer 1

你linux环境配置了java_home 没问题以外，你还需要给在biin/hadoop 这个启动脚本中也配置 java_home,具体你看下这个脚本的注释就很清楚了！

Answer 2

export 一下java_home吧，可能ssh脚本里面有些命令在新的shell里运行

Answer 3

那你在命令行输入java -version有显示java 版本吗？

Answer 4

有