We can run Hive in following modes:
1. Local mode: In Hive local mode, Map Reduce jobs related to Hive run locally on a user machine. This is the default mode in which Hadoop uses local file system.
2. Distributed Mode: In this mode, Hive as well as Hadoop is running in a fully distributed mode.
NameNode, DataNode, JobTracker, TaskTracker etc run on different machines in this mode.
3. Pseudo-distributed Mode: This is the mode used by developers to test the code before deploying to production. In this mode, all the daemons run on same virtual machine. With this mode, we can quickly write
scripts and test on limited data sets.