Hi, As far as I understand HADOOP ( by following its scripts like start-all.sh etc .. a& my experience ) it does not care about the rest of the environmental variable as long as you provide right values in its environment file.
ie .. conf/hadoop-env.sh set some thing similar to get around your problem. # The java implementation to use. Required. export JAVA_HOME="/cygdrive/c/Program Files/Java/jre1.5.0_11" #/usr/lib/j2sdk1.5-sun Regards, -Vikas. On Sat, May 24, 2008 at 6:11 AM, vatsan <[EMAIL PROTECTED]> wrote: > > I have installed hadoop on cygwin , I am running windows XP. > > My Java directory is C:\Program Files\Java\jre1.6.0_06 > > I am not able to run hadoop as it complains of "no such file or directory > error". > > I did some searching and found out someone had proposed a solution of doing > > SET JAVA_HOME=C:\Program Files\Java\jre1.6.0_06 > > in the Cygwin.bat file, > > but that doesn't work for me. > > Neither does using the absolute path name "\cygwin\c\Program Files\Java" OR > using \cygwin\c\"Program Files"\Java > > Can someone guide me here? > > (I understand that the problem is because of the path convention conflicts > in windows and Cygwin, I found some stuff on fixes for the path issues that > spoke of using cygpath.exe as a fix ... for example while running a java > program on cygwin, but could not find anything that addressed my problem.) > > -- > View this message in context: > http://www.nabble.com/JAVA_HOME-Cygwin-problem-%28solution-doesn%27t-work%29-tp17443172p17443172.html > Sent from the Hadoop core-user mailing list archive at Nabble.com. > >