Query metadata tables feature
From Hive and Impala, you can query Iceberg metadata tables as you would query a Hive table. For example, you can use projections, joins, filters, and so on.
The
following Iceberg metadata tables are available from Hive and Impala:
- entries
- files
- manifests
- partitions
- snapshots
The following additional Iceberg metadata tables are available from Impala:
- history
- all_data_files
- all_entries
- all_manifests
For more information about querying Iceberg metadata, see the Apache Iceberg Spark documentation.
Hive and Impala Syntax
To reference a metadata table, use the full name of the table as shown in the following
syntax:
<DATABASE_NAME>.<TABLE_NAME>.<METADATA_TABLE_NAME>
Hive and Impala Example
SELECT * FROM default.table_a.files;
Limitations
- The following types are not supported: ARRAY, MAP, BINARY
- Nested types are not expanded by default when the you execute a SELECT *.
- You cannot explore the tables. For example, you cannot list the available metadata tables or describe the tables.
