I have seen Ganglia monitoring being implemented and analyzed on grid computing projects, but haven't read about any procedure for Amazon Elastic Mapreduce programs. Ganglia has a lot of metrics, but what are the important ones to focus on if we want to write a report analyzing the performance of our elastic mapreduce program? Is there any good research paper or article to reference to? What would be important concepts to discuss in a thesis paper in relation to elastic mapreduce and ganglia metrics?
I have written my program in python and want to run it on elastic mapreduce and then want to monitor it using ganglia, but I want to make some conclusions from the ganglia monitoring results and hopefully able to compare and reference to any article or previously done research similar to this type of usage.
