Choose any Hadoop distribution, and run it wherever you’d like, on-premises or in the cloud. It doesn’t matter to us, our big data platform will just run with it (and on it, of course).
Do you know which execution engine will process your data most efficiently? The answer is, “it depends on the job.” Every. Single. Time. Sometimes it’s Spark, sometimes it’s Tez, sometimes it’s in-memory; let Smart Execution figure it out every time, so you don’t have to.