Independent SQL-On-Hadoop Benchmark of SparkSQL, Impala, and Hive

50 days ago on the "Announcing Spark 1.6 (databricks.com)" thread I mentioned that we were doing a 3rd party SQL-on-Hadoop benchmark (https://news.ycombinator.com/item?id=10837758) and a bunch of folks sent me notes asking for a heads up when the results were in.

Well, we are done with our analysis and have a blog post offering up up the bulk of the results - http://blog.atscale.com/how-different-sql-on-hadoop-engines-...

The blog has the majority of the results, and additionally there is a registration link for the full 17 page whitepaper if you are really keen on SQL-on-Hadoop. www.atscale.com/benchmark

Trystan, the engineer that did the bulk of the benchmark work, would be happy to answer questions regarding the methodology, hardware, etc.

Due to how fast these engines are evolving, we plan on doing an update to this benchmark on a quarterly basis.