Snowflake makes Open Data Sharing generally available, with regional access limits
Snowflake’s November 2 release notes make Open Data Sharing generally available, ending its preview status. The feature lets providers share live, read-only data in Snowflake-managed Iceberg tables with consumers outside Snowflake through standard Iceberg REST Catalog APIs.
The release adds cross-region external sharing in commercial Snowflake regions on AWS and Azure. A provider selects one region when creating each external consumer, and Snowflake automatically fulfills the listing to that region. That serving region is distinct from the location where a consumer’s client connects.
Snowflake describes open sharing as avoiding data movement or separate ETL pipelines for client access. The same release explicitly describes regional automatic fulfillment. The announcement should therefore not be read as a promise that cross-region distribution never involves replication or that no storage or compute costs can arise.
External consumers can now authenticate through workload identity federation as an alternative to programmatic access tokens. The listed identity providers are AWS, Microsoft Azure, Google Cloud and OpenID Connect. Identity-provider support is a separate question from the supported regions for cross-region external consumers.
Providers can also add external consumers to an existing listing, allowing that listing to serve Snowflake consumers and open clients. This changes how providers can distribute a listed data product; it does not turn the read-only sharing capability into permission for an external client to modify the source tables.
Teams evaluating the release should check the chosen serving region, supported identity provider and listing setup before deployment. The short release note does not establish a universal price, unrestricted availability in Google Cloud regions or a guarantee of better security for every configuration.