Hive is a data warehouse built on top of Hadoop which provides a SQL-like query and data definition language. Hive converts queries into MapReduce jobs that can be run across the entire Hadoop cluster. Like all useful abstractions, Hive does not try to deny the existence of the underlying mechanics of Hadoop and supports custom map-reduce operations as a powerful extension mechanism. Despite the superficial similarities to SQL, Hive does not try to be a replacement for low-latency, real-time query engines found on relational database systems. We strongly advise against using Hive for online ad-hoc querying purposes.