MapReduce is best optimized for HDFS, but you can run MapReduce jobs over data stored in other file-systems e.g. local file system (e.g. ext3, ext4, xfs), S3, CloudStore (formerly Kosmos) to name a few.
2013/11/25 Nitin Pawar <[email protected]> > you don't necessarily have to have to hdfs to run mapreduce. > > But its recommended :) > > > > > On Mon, Nov 25, 2013 at 3:25 PM, RajBasha S <[email protected]>wrote: > >> can Map Reduce will run on HDFS or any other file system ? HDFS is >> Mandatory >> > > > > -- > Nitin Pawar >
