Following any hadoop dfs related commands to see what code path it takes would be a good start. Bulk of hdfs code in org.apache.hadoop.dfs package. org.apache.hadoop.fs has generic interface/apis, especially FileSystem.java You could follow that and check DistributedFileSystem.java which implements/overrides FileSystem apis DFSClient.java should be a good start for hdfs client end code. On the server side NameNode.java would be a good start. Similarly DataNode.java for Data nodes.
Thanks, Lohit ----- Original Message ---- From: Sangmin Lee <[EMAIL PROTECTED]> To: [email protected] Sent: Monday, June 23, 2008 12:32:17 PM Subject: Start point for source code Hi all, I'm a novice for hadoop code and want to hack hdfs code. Can any body point me to the startpoint ? I appreciate your help in advance. Cheers, Sangmin
