Get the latest tech news
Are efficiency and horizontal scalability at odds?
Why are scalable systems locally-inefficent, and locally-efficient systems unscalable? Plus, new book release!
But at what COST, Frank McSherry shows that his single-threaded laptop could outperform 128-node "big data systems" on PageRank and graph connectivity (via label propagation). … The union-find algorithm is fundamentally incompatible with the graph computation approaches Giraph, GraphLab, and GraphX put forward (the so-called “think like a vertex” model). But I think now that we both missed something obvious: with the "scalable" solution, the data scientists didn't have to write an optimized script for every single query.
Or read this on Hacker News