Hive is too slow for some of my use case and trying to look at different options. I read that Hive can run on Spark. Has anybody run Hive on Spark ? If yes what is the experience. Is it possible to run Hive on both MapReduce and Spark ? In other words, if I want to run Hive in batch, then I can run it against MapReduce and when I am running adhoc query then I like to run Hive against Spark.