将JAVA_HOME设置为Java安装的根目录

Question

I'm currently using hadoop and in the config part there's "define JAVA_HOME at least to be the root of your java installation" I'm actually confused about it. Say, my JAVA_HOME when using: echo $JAVA_HOME is: /System/Library/Java/JavaVirtualMachines/1.6.0.jdk/Contents/Home

and my java installation when using: which java is: /usr/bin/java

What change should I make to set JAVA_HOME as the java root?

Thanks!

And for a follow up, I think the tutorial mentioned to make the change inside the conf/en-sh part. I tried the combinations of the below answers proposed, but still got the error of 1. invalid identifier for the export JAVA_HOME = .. part 2. unable to load hadoop library

Please help..... Any suggestion?

Answer 1

For GNU/Linux operation system users: in your bash profile set your Java Home variable export JAVA_HOME=/usr/bin/java then export PATH=$PATH:/usr/bin/java . Here is the complete tutorial.

Answer 2

On my system, I have JAVA_HOME set to /Library/Java/Home on Mountain Lion. Fair enough... add export JAVA_HOME=/Library/Java/Home line to the .bash_profile, .zshrc file in your home directory. If you're using tcsh, the line you want is setenv JAVA_HOME /Library/Java/Home . The default shell on recent versions of OS X for users in bash. You can find out what you're using using echo $SHELL .

Answer 3

Set you JAVA_HOME to /System/Library/Java/JavaVirtualMachines/1.6.0.jdk. That's enough.