On Mon, Oct 5, 2015 at 4:38 PM, Andy Zelinski <[email protected]> wrote:
> > thoughts so far: translate Spark map transformation, map reduce algorithm > to a fork-join(use cores!). partition cluster so users distributed evenly. > > I think this is a very good use-case for Ignite. You will get sub-second latencies for your job, as well as guaranteed fault-tolerance in case of any kind of failures D.
