Graduate School Project

p2pMapReduce

##log: 2015/4/19 (Lingfeng)

• 1, program framework
• 2, start up client program with login, logout
• 3, start up server program with table, table check, keepalive

###Next Step:

• 1, file transfer
• 2, file segmentation
• 3, file run
• 4, client job submit
• 5, ...more

##log: 2015/4/21 (Lingfeng)

• 1, file transfer
• 2, file segment
• 3, run script in program
• 4, job submit, job allocation, job result return

###How To Use (temporarily)

• client: login, logout, submit, upload, download
• server: checktable, basic unit test

###Next Step:

• 1, reducer related
• 2, fault tolerance
• 3, performance opimization

##log: 2015/4/22 (Lingfeng)

• 1, Sort output before send back to server. This feature helps reducer, just like Hadoop.
• 2, Merge File on server.
• 3, If sub task failed, now the server will find a new available node to redo the task.
• 4, File segmentation program now will not split line with the same key into two different sub files.
• 5, Clean all the temp files on both client and server.
• 6, Build example space and run the demo.

###Next Step:

• 1, Report
• 2, Demo