Iceberg tables
What Apache Iceberg gives you and how Databaas stores your tables.
Databaas stores your tables as Apache Iceberg tables. You never manage Iceberg yourself, but what it gives you explains several things you see in the catalog and in queries.
Snapshots and time travel
Every write creates a new snapshot — a complete, consistent version of the table at that moment. The table has one current snapshot, shown in its Properties tab, and older ones are kept. Because past snapshots survive, you can read a table as it was at an earlier point — time travel — which is what makes a preview or a query reproducible even while a pipeline keeps writing.
Schema evolution
An Iceberg table has a real schema — named columns with types, some required. The schema can change (a column added, say) without rewriting existing data, so a long-lived table keeps working as it grows. The current columns are in its Schema tab.
One table, many engines
The same Iceberg table is read by every engine — a notebook, SQL Lab, a pipeline — through the shared catalog. None of them owns the table; they all ask the catalog where the data is and read it directly. This is why a table you load in Python and one you query in SQL are the same table, with the same rows and permissions.
Not everything is Iceberg
The catalog can also list generic tables — registered but not Iceberg, such as a folder of files or a Delta table. They carry a format badge and have no snapshots, schema, or preview. See table details.
Going deeper
To work with Iceberg from Python the platform uses pyiceberg; see querying data for the Databaas helpers that load a catalog for you. For the format itself, read the Apache Iceberg documentation. The terms used here are collected in core concepts.