1 comment

[ 6.6 ms ] story [ 20.1 ms ] thread
Just use bigquery. The last thing I want these days is to manage a cluster. Next to last thing I want is to worry about the zillion tuneables on hadoop/spark.