question about hadoop-mapreduce project

[Samaneh Shokuhi <sa...@gmail.com>]

Hi ,
I am new to hadoop framework.My master thesis is about finding  weaknesses
of  map-reduce section of hadoop project.
1)  what do you suggest to start with? And any known weaknesses to start
investigating on that ?

I have downloaded the Release 1.0.3  and ran some samples successfully by
hadoop-core-1.0.3.jar to be familier with hadoop.
I have got the source code from Jit and built successfully.

2)How can i generate hadoop-core-1.0.3.jar from source code? Since i want
to change the source and then test it need to generate this jar file after
changing the code.
As i saw in the pom.xml,  after building the source code, in each project
we will have .tar file in target folder.  Is hadoop-mapreduce.tar.gz
equivalent to hadoop-core.jar ?!

3) I want to look into the sort and shuffling  mechansim in source code
,what project should i look at ? hadoop-mapreduce project?


Best Regards,
Samaneh