If the job's task isn't doing any HDFS operations, it should be able
to continue running.
On Sun, Apr 29, 2012 at 3:41 AM, Aisling O' Driscoll
<[EMAIL PROTECTED]> wrote:
> If the name node fails, I understand new jobs can't be accepted but do
> active jobs still continue to run (even if the results can' be retrieved
> from HDFS)?