【hadoop】hadoop-0.20.2安装找不到JAVA_HOME解决办法

转载：http://p-x1984.iteye.com/blog/710337

在安装hadoop-0.20.2时出现了找不到JAVA_HOME,配置方法跟官方的版本一致，具体请参考前边的文章或者网上的文档。配置好以后，运行start-all.sh时，确出现如下的报错： 

======================================================================+ 
|       Error: JAVA_HOME is not set and Java could not be found          | 
+----------------------------------------------------------------------+ 
| Please download the latest Sun JDK from the Sun Java web site        | 
|       > http://java.sun.com/javase/downloads/ <                      | 
|                                                                      | 
| Hadoop requires Java 1.6 or later.                                   | 
| NOTE: This script will find Sun Java whether you install using the   | 
|       binary or the RPM based installer.                             | 
+======================================================================+ 
明明在hadoop-env.sh设置好了，为了保险，又在/etc/profile设置了一遍。以hadoop用户登陆系统，在任意路径执行 echo $JAVA_HOME ,显示"/usr/local/jdk"，是我服务器jdk正确的路径啊！再查一下文档，云：需要jdk1.6以上的版本，我使用的是 jdk1.6.0_16，应该符合要求的。 

那问题又会在哪里呢？grep一把，看那些文件包含字符串JAVA_HOME.先搜hadoop/conf，只有hadoop-env.sh包含这个字串。再换一个目录hadoop/bin,查看一下,其输出如下： 
[root@hadoops2 hadoop]# grep JAVA_HOME bin/*bin/hadoop:export JAVA_HOME=/usr/local/jdkbin/hadoop:#   JAVA_HOME        The java implementation to use.  Overrides JAVA_HOME. 
bin/hadoop:if [ "$JAVA_HOME" != "" ]; then 
bin/hadoop:  #echo "run java in $JAVA_HOME" 
bin/hadoop:  JAVA_HOME=$JAVA_HOME 
bin/hadoop:if [ "$JAVA_HOME" = "" ]; then 
bin/hadoop:  echo "Error: JAVA_HOME is not set." 
bin/hadoop:JAVA=$JAVA_HOME/bin/java 
bin/hadoop:CLASSPATH=${CLASSPATH}:$JAVA_HOME/lib/tools.jar 
bin/hadoop-config.sh:if [ -z "$JAVA_HOME" ]; then 
bin/hadoop-config.sh:      export JAVA_HOME=$candidate 
bin/hadoop-config.sh:  if [ -z "$JAVA_HOME" ]; then 
bin/hadoop-config.sh:|      Error: JAVA_HOME is not set and Java could not be found         | 
bin/rcc:#   JAVA_HOME        The java implementation to use.  Overrides JAVA_HOME. 
bin/rcc:if [ "$JAVA_HOME" != "" ]; then 
bin/rcc:  #echo "run java in $JAVA_HOME" 
bin/rcc:  JAVA_HOME=$JAVA_HOME 
bin/rcc:if [ "$JAVA_HOME" = "" ]; then 
bin/rcc:  echo "Error: JAVA_HOME is not set." 
bin/rcc:JAVA=$JAVA_HOME/bin/java 
bin/rcc:CLASSPATH=${CLASSPATH}:$JAVA_HOME/lib/tools.jar 

嘿！好几个文件都与JAVA_HOME相关呢。接着挨个查看文件，感觉hadoop-config.sh这个文件嫌疑最大，其内容刚好有一段与运行hadoop报错内容相一致，我把程序片段贴在下面： 
# attempt to find java 
if [ -z "$JAVA_HOME" ]; then 
  for candidate in \ 
    /usr/lib/jvm/java-6-sun \ 
    /usr/lib/j2sdk1.6-sun \ 
    /usr/local/jdk \ 
    /usr/java/jdk1.6* \ 
    /usr/java/jre1.6* \ 
    /Library/Java/Home ; do 
    if [ -e $candidate/bin/java ]; then 
      export JAVA_HOME=$candidate 
      break 
    fi 
  done 
  # if we didn't set it 
  if [ -z "$JAVA_HOME" ]; then 
    cat 1>&2 <<EOF 
+======================================================================+ 
|      Error: JAVA_HOME is not set and Java could not be found         | 
+----------------------------------------------------------------------+ 
| Please download the latest Sun JDK from the Sun Java web site        | 
|       > http://java.sun.com/javase/downloads/ <                      | 
|                                                                      | 
| Hadoop requires Java 1.6 or later.                                   | 
| NOTE: This script will find Sun Java whether you install using the   | 
|       binary or the RPM based installer.                             | 
+===================================================================== 
我在这段中间加如行  " /usr/local/jdk \" ,然后在运行，一切就正常了！ 
希望对大家有所帮助！