Comment by zhousun

people tried to run spark (better hadoop) and failed lol. https://github.com/ClickHouse/ClickBench/pull/139