On Mon, Apr 19, 2010 at 12:45 PM, Tiago Veloso <[email protected]> wrote:
> Hi, > > I have recently started working with hadoop. I will be doing a project with > it. > > One of the goals of the project is to compare how long it would take to > execute a given task in a purely sequential mode vs a parallelized one. > > My problem is having a sequential execution to serve as a baseline > comparison. > > Would the Standalone Operation mode of hadoop be appropriate for this? > > Tiago Veloso > [email protected] > > > > Would the Standalone Operation mode of hadoop be appropriate for this? Yes. I think so. You can run on a local machine with no configuration: Running 'hadoop jar' uses local jobtracker and your local filesystem On a distributed cluster: 1) set mapred.map.tasks=1 and mapred.reduce.tasks=1 --or-- 2) set mapred.job.tracker=local Uses 1 "process" but uses the distributed filesystem.
