1

I know the rule of thumb of it: Big data, non-iterative, fault tolerant => MapReduce ; Speed, small data, iterative, non-Mapper-Reducer type => MPI (Hadoop MapReduce vs MPI (vs Spark vs Mahout vs Mesos) - When to use one over the other?).

What i want to know is what benchmarking method should i use to prove MapReduce is good for Big data, non-iterative, fault tolerant case. And what benchmarking method should i use to prove MPI is excel in Speed, small data, and iterative case.

Really appreciate any help you can provide

Community
  • 1
  • 1
kurniawan dwi septian
  • 11
  • 3

1 Answers1

1

You could have a look at BigDataBench. It has a range of different workloads and it also works with MPI. Maybe checking out their existing publications could be helpful as well.

js84
  • 3,676
  • 2
  • 19
  • 23