Query one Apache Iceberg table from Trino, Spark, and DuckDB
Iceberg is useful because it lets multiple engines read the same table without copying data into engine-specific storage. The table stays in object storage. The engines differ only in how they reach the metadata.
That is the part most posts skip. They explain Trino, Spark, or DuckDB in isolation, then leave you to infer how the same table behaves across all three. If you are working on a real lakehouse, that missing comparison is the whole problem.
The short version:
Trino usually reads Iceberg through a catalog.
