Hello, I'm wondering how does work hadoop on a cluster of few machine with different hardware.
For instance, i have a cluster of 2 machine with the same hardware (same CPU p4 2.8 , same memory RAM 2Go). I've configured hadoop to use 1Go of memory for the MapReduce process, with 2 tasktracker child of 400 Mo. I would like to add another machine which is an AMD Athlon with only 512Mo of RAM. is it possible to do it ? is there any script to change to limit the number of process to run on this machine to have only 1 tasktracker of 400Mo running on this machine ? if not, does it mean that we need to have the same size of RAM on each machine ? Cheers E
