I think the author's point was that the example in question was orders of magnitude smaller than "big data" and that it was more efficient to process it on a si

I think the author's point was that the example in question was orders of magnitude smaller than "big data" and that it was more efficient to process it on a single machine, not that Hadoop and friends aren't easy to use.