There are four data units in HIVE.
Databases - It is a namespace to avoid naming conflicts for tables, views, partitions, columns, etc. It supports multiple databases.
Tables - They are schemas in the namespace. Tables can be either internal, where Hive manages the lifecycle of data, or they can be external when files are used outside of Hive.
Partition
Tables can be partitioned to have more managed data.
Buckets (cluster)
Partitions can be divided more into buckets, to support data sampling.
