Table details
Read a table's schema, properties, and data preview.
Select a table in the catalog tree to open its detail pane. The breadcrumb
spells out its identity — warehouse, namespace, table. The header shows the Iceberg format version
(for example iceberg v2) and when the table last changed. An Iceberg table has four tabs:
Preview, Schema, Properties, and Permissions.
Preview
The first 20 rows of the table’s current snapshot — a sample, not an analysis. The preview runs as you, with your own grants, so it shows exactly what your queries would:
- A badge counts the rows shown. A second badge (currently labelled truncated to first parquet file) means the table holds more rows than the preview shows.
- An empty table says Table is empty.
To read a table for real — every row, filtered, or joined — use a notebook or SQL Lab.
Schema
The table’s columns: name, type, whether it is Required, whether it is an Identifier field, and a description if it has one.
Properties
The table’s metadata: storage location, table UUID, Iceberg format version, current snapshot, and last-updated time, plus any table properties. The current snapshot is the version every engine reads by default — see Iceberg tables.
Permissions
Who is granted what on this table, directly and inherited from further up the tree. See data permissions.
Tables that are not Iceberg
A generic table — registered in the catalog but not Iceberg, such as a Zarr array or a set of files — has no snapshots or schema, so there is nothing to preview. Its header shows its format, and it has two tabs instead: Dataset (format, location, and any properties) and Permissions. See files, blobs and Zarr arrays for reading and writing them.