Skip to content
Lucas Lessa
Voltar para as notícias
DatabricksGAImpacto: Médio

Share foreign Iceberg tables with OpenSharing

You can now share and read Iceberg tables federated via Lakehouse Federation using OpenSharing, including with external Iceberg clients.

O que mudou

Sharing and reading foreign Iceberg tables using OpenSharing is now generally available (GA). Providers can include in a share Iceberg tables federated from external catalogs via Lakehouse Federation and make them available to recipients, including those using external Iceberg clients.

Use cases

Imagine you have centralized several external Iceberg tables using Lakehouse Federation and need to expose them to external partners or systems without copying the data. With this feature, you can:

  • Share federated Iceberg tables directly via OpenSharing.
  • Allow external clients using the Iceberg protocol to consume the shared data.
  • Manage access using Unity Catalog security controls.

How to use

You can create a share object and add foreign tables using SQL commands in a notebook or SQL editor.

CREATE SHARE IF NOT EXISTS my_share;

ALTER SHARE my_share ADD TABLE my_catalog.my_schema.my_foreign_iceberg_table;

What happens in this code:

  1. Creates a new share object named my_share in Unity Catalog.
  2. Adds the federated foreign Iceberg table to the share.

Limitations and requirements

  • This feature is generally available (GA).
  • Requires appropriate privileges in Unity Catalog, such as CREATE SHARE and read permissions on the target table and schema.
  • Using SQL commands to add entire schemas requires a SQL warehouse or compute running Databricks Runtime 13.3 LTS or above.
Ver nota oficial do Databricks

Sharing foreign Iceberg tables with OpenSharing is now generally available

Resumo e explicação gerados por IA a partir da release note oficial. Em caso de dúvida, a documentação oficial da plataforma prevalece.

Relacionadas