If all you're trying to analyze is 20 gigs you don't need Hadoop. I suggest loading it into an SQL database and crunching on a desktop box. If the point is to setup and use Hadoop, on Digital Ocean you can rent a requisite cluster for about the price of a latte.


Also -- don't use raw Hadoop, use Spark/Shark.




For another fun tool -- google BigQuery. 


—
Sent from Mailbox

On Sat, Oct 25, 2014 at 5:23 PM, Sachin Garg <[log in to unmask]>
wrote:

> Hi,
> We are looking for a Hadoop Cluster to analyze data related to airlines 
> ontime arrival for Dr. Borne's CS695 class.
> Wondering if anyone can point us to a Hadoop cluster on campus to do 
> this? I remember that we have a shared computing resource in the 
> university that we can use.
> Best,
> Sachin
> -- 
> Sachin Garg <[log in to unmask]>
> Doctoral Student
> School of Policy, Government, and International Affairs
> George Mason University, Arlington, VA 22201
> Phone: +1-703-993-8647  Cell: +1-703-996-9445
> SSRN page: http://ssrn.com/author=690016