Assignment 5: MapReduce

Get familiar with Google's MapReduce [1] and its programming model. 

Implement a sample application with an open source framework of your choice; e.g. Hadoop [2], Twister [3], etc. 
In case you cannot come up with an own idea, you may base your application on the sample applications of the tutorials. 

Take a look at Amazon web services, such as the Amazon Elastic Cloud (EC2) or Amazon Elastic MapReduce.
Browse the web for news/articles about applications implemented with MapReduce (and Amazon). E.g., the New York Times created some million PDF documents from several TB of scanned documents.

Test your application locally and on a (at least pseudo-) distributed system and document performance.
Please submit your implementation as zip without binaries but including a readme (what is it, how does it work, how to compile/run it, ...) and your performance observations.

Explain MapReduce (and be prepared to answer some questions about the paper!!) and present/run your implementation in the PS.

Deadline: Monday 30.05. --> E-Mail! Subject: [VS]
Presentation/Discussion: in the PS: Wednesday 01.06.

[1] http://labs.google.com/papers/mapreduce-osdi04.pdf
[2] http://hadoop.apache.org/mapreduce/
[3] http://www.iterativemapreduce.org/