mapreduce for proxy log file analysis

[Bright D L]

Hi all,
        I am doing a simple project to analyze http proxy server logs by hadoop 
mapreduce approach (in Java). The log file contains logs for a week or some 
times more than that.  
        I  have following requirements:
                1) Find the top 50 bandwidth consumers (IPs) for each day
                2) Find the hour of the day where there is maximum bandwidth 
utilization
        Please help me out with some directions. Sample code is highly 
appreciated.
Thank you all,
Bright