Hi Folks, I am running a map only (no reduce) job on Hadoop 2.3.0 and the /tmp/hadoop-df/nm-local-dir (df being my user name) directory is growing exponentially causing me run out of my 80GB disk space.
I would like to know if there is any mechanism(automatic or manual) by which I can free up the space in /tmp/hadoop-df/nm-local-dir while the job is still running. Thanks.
