Skip to content
BENIBIX Learn

Table Catalog

Also known as Iceberg Catalog, Iceberg REST Catalog, Lakehouse Catalog, Metastore

A list that tells tools where each lake table is right now.

Think of it like

It is like the sign by a lift that says which floor each shop is on. Every visitor reads the same sign.

Example

Spark writes an orders table and Trino reads it. Both ask the same Iceberg REST catalog, so both see the new rows at the same moment.

Why it matters

Without one shared catalog, each engine keeps its own idea of a table. Then two tools give two answers.

Definition

A table catalog is the service that lets query engines find the tables in a data lake. It keeps the list of table names. For each one, it points at the file that describes the table right now. When a write changes a table, that pointer is swapped in one step. So every engine, from Spark to Trino, sees the same version. Hive Metastore, Unity Catalog and the Iceberg REST catalog are all examples.
The catalog is the box in the middle. Both engines follow its pointer to the same version.

Common mistake

Confusing it with a data catalog. One is for engines, and the other helps people find data.

Key takeaways

  1. 1
    Names to files. Points each table at its current state.
  2. 2
    One swap. A change lands all at once.
  3. 3
    Shared by engines. Spark, Trino and others agree.

Compare