Look into /tmp for a logs folder. That contains the scripts execution logs and you should be able to see why it failed - I assume the download failed for the Hadoop tarball.
On Wed, Feb 22, 2012 at 9:56 AM, Frank Scholten <[email protected]>wrote: > Hi all, > > I see this message several times in the log when I start a Hadoop cluster. > > No directory, logging in with HOME=/ > , error=-su: /usr/local/hadoop/bin/hadoop-daemon.sh: No such file or > directory > > I can SSH into the nodes but Hadoop isn't started obviously. I > navigated to /user/local/hadoop and found the bin directory to be > empty. Do you know what might be causing this and also: how do you > generally debug these kind of problems? > > Cheers, > > Frank >
