Please use the hadoop wiki for infrmation: http://wiki.apache.org/hadoop/
There is a HDFS design document at : http://hadoop.apache.org/core/docs/current/hdfs_design.html sure, it i possible to run only hdfs. You have to use bin/start-dfs.sh to start the cluster. hope it helps, thanks, dhruba On Thu, Feb 19, 2009 at 8:42 AM, Ajit Ratnaparkhi < [email protected]> wrote: > Hi all, > > I am a student of IIT, Bombay India. > We are doing research on Distributed File System, in which we are > experimenting some design modifications to HDFS. > We have done some modification to the HDFS code, and we want to test the > results. > > We are new to hadoop development, can you please help us in following > matter, > > 1. Can you tell the exact compilation procedure of Hadoop so as to reflect > any modification to the code in behaviour of system. > 2. Is there any debug tool or procedure? > 3. As we are interested in only HDFS part of hadoop(not map-reduce), Is it > possible to separate HDFS module from hadoop and how to do that? We want > to write programs which will directly use apis provided by HDFS. > > thanks a lot, > -Ajit. >
