If I am not mistaken the code for both a distributed and a non-distributed server is the same.    The only difference is in the results. For a distributed you get a record per distributed node. If you group these records as suggested in the documentation you will get the total memory usage. I am using this code now on my current project. As Juan stated this is a very interesting topic, especially when you start digging into data stored in HDFS and mapped to LASR. Some cool stuff here.    Richard ... View more