This only applies to tables which have partitions, i.e. just those in the Hive metastore (unless some RDBMS tables are sharded but... let's not go there). Exposing this information would make it easier to say to people "filter on partitions!" rather than have that be intimate knowledge for each dataset.
This only applies to tables which have partitions, i.e. just those in the Hive metastore (unless some RDBMS tables are sharded but... let's not go there). Exposing this information would make it easier to say to people "filter on partitions!" rather than have that be intimate knowledge for each dataset.