MAY 20, 2025|1 min read
OCT 13, 2025|6 min read
Extending Snowflake Data Sharing to Open Table Formats
Blog updated in February 2026 to clarify product details.
We're excited to announce a major expansion of Snowflake’s zero-ETL data sharing: Now, you can easily and securely share open table formats, including Apache Iceberg and Delta Lake, across regions and clouds. This release allows source providers to securely share data stored in open table formats across any cloud and region without complex pipelines and without incurring exponential per-query egress charges.
Snowflake data sharing already leads in the market when it comes to collaboration capabilities and adoption. With 2.5x the data sharing ecosystem compared to leading competitors, Snowflake collaboration provides increasing value to a growing, global user base. With thousands of customers already sharing data, businesses can use open table format sharing to get the best of both worlds: open-format data sharing and all of the core benefits of Snowflake sharing. These benefits include:
Near real-time data access
Snowflake Horizon Catalog’s policy-based governance controls for shared data
Cross-Cloud Auto-Fulfillment and Egress Cost Optimizer for simplified and economical cross-region and cross-cloud sharing
Delivery into regulated regions including the U.S. government cloud and Virtual Private Snowflake
Why is this important?
For many years, Snowflake customers have been able to securely share data and collaborate with a vast ecosystem of customers and partners. Data sharing is a cornerstone of the Snowflake platform, with many customers using it to develop connections and build strong data ecosystems in the Snowflake AI Data Cloud.
With the introduction of open table format sharing, Snowflake collaboration capabilities have expanded to:
Data stored outside of Snowflake in customers' own cloud storage (AWS S3, Azure Storage, Google GCS)
Data in open table formats, including:
Apache Iceberg managed by Horizon Catalog or managed by external catalogs (AWS Glue, Apache Polaris)
Delta Lake managed by external catalogs (Databricks Unity, Hive Metastore)
This means businesses that choose open table formats are also now connected into the AI Data Cloud and can enjoy the benefits of a thriving data ecosystem.
What challenge does this solve?
Most organizations frequently need to share diverse data formats, both internally and externally, yet they often encounter obstacles such as:
Security and compliance: Enforcing fine-grained data-access policies on shared data is crucial for maintaining security and compliance.
Geographic and cloud dispersion: Collaboration among business units (LOBs, vendors, customers) that are often spread across different regions and clouds, including commercial and government cloud environments.
Varied data formats: Data exists in diverse formats, such as Apache Iceberg and Delta Lake.
Snowflake data sharing, and specifically open table format sharing, directly addresses the key obstacles that data engineers and data architects face when collaborating on data. It reduces geographic and cloud barriers with the ability to share open-format tables, it extends core governance capabilities, and it gives customers the flexibility to standardize on a data format such as Iceberg, all while allowing global organizations to share with any business unit, vendor and/or customer.
How does it work?
Open table format sharing is enabled by Cross-Cloud Auto-Fulfillment (supported on commercial, Virtual Private Snowflake and U.S. government clouds), which simplifies data sharing for Apache Iceberg and Delta Lake directly from your cloud storage. You can share this data with a Snowflake consumer in any region or cloud without needing to manage the underlying infrastructure or requiring you to maintain extract, transform and load (ETL) jobs. In addition, open table format sharing optimizes data transfer costs through Egress Cost Optimizer, helping to avoid unpredictable and astronomical per-query egress charges.
Snowflake Horizon Catalog provides comprehensive policy-based governance controls that can be applied to open table format data shared with consumers across different regions or clouds. This capability ensures data residency and facilitates the compliance necessary for collaboration, particularly within or in conjunction with regulated sectors, such as public sectors, financial services, healthcare and life sciences.
When combined with Delta Direct or catalog federation* (leveraging Unity Catalog and Uniform's IRC API), Snowflake Cross-Cloud Auto-Fulfillment extends this capability to a Delta Lake table residing in your cloud storage and written by Delta engines such as Microsoft Fabric or Databricks or managed by catalogs such as Databricks Unity or Hive Metastore. This extends open table sharing to Delta Lake tables with Snowflake consumers in any region or cloud, again, without the need to manage underlying infrastructure.
Example of sharing managed Apache IcebergTM tables on Snowflake
Open table format sharing allows Apache Iceberg and Delta Lake tables to be shared across regions or clouds and is available only through the Iceberg Cross-Cloud Auto-Fulfillment feature. For details on how auto-fulfillment works, see “Auto-fulfillment for listings.” For more information on how open table format sharing works, see the “Using auto-fulfillment with open format tables” documentation.
Here is how to get started with sharing open table formats, using the example of sharing Iceberg tables.
Step 1: Create a table with an open table format (for instance, Iceberg)
Sign in to the provider account and create an open table format (such as Iceberg) following the SQL commands listed below. For more information on how to create Iceberg tables or managed by external catalogs, see the CREATE ICEBERG TABLE documentation here.
SQL command for creating an Iceberg table in AWS S3
create database sample_db_laf_<your alias>;
create schema sample_sc;
create warehouse xsmall;
use warehouse xsmall;
-- Create an external volume for managed table
create or replace external volume managed_exvol
STORAGE_LOCATIONS = (
(
NAME = 'my-s3-us-west-2'
STORAGE_PROVIDER = 'S3'
STORAGE_BASE_URL = 's3://datalake-storage-team/iceberg/table_replication_west'
STORAGE_AWS_ROLE_ARN = 'arn:aws:iam::631484165566:role/datalake-storage-integration-role-2'
)
);
-- create iceberg managed table
create or replace iceberg table sample_iceberg_table_for_laf_testing (
symbol string,
description string
) EXTERNAL_VOLUME = 'managed_exvol'
catalog=snowflake
BASE_LOCATION='sample_iceberg_table_for_laf_testing';
insert into sample_iceberg_table_for_laf_testing values ('A', 'Adam'), ('E', 'Eve'), ('B', 'Bob');
insert into sample_iceberg_table_for_laf_testing values ('D', 'Dan'), ('C', 'Cat'), ('F', 'Fan');
select * from sample_iceberg_table_for_laf_testing;Step 1b (optional): Protect data with policy-based governance controls
Before sharing, leverage Snowflake Horizon Catalog’s governance features’ full support on your Horizon-managed and externally managed Apache Iceberg or Delta tables, including:
Fine-grained governance policies: row access, masking policies
Privacy-enhancing policies: aggregation policy, join policy, projection policy
Data classification and tagging
Additionally, you can audit and monitor the sharing with a complete audit using account usage and access history.
For more information, review examples in “Share data protected by policy.”
Step 2: Create a listing targeting consumers within/across regions with a few clicks
In your provider account, create a private listing through Snowsight interface by clicking +Listing and selecting Only Specified Customers, then doing the following:
Specify a listing name
Attach the Iceberg table that you created
Add a listing description
Specify the account alias of the consumer account
Publish the listing
Detailed steps are available here.
Status: Using the ACCOUNTADMIN role, you can check the status of the listing in the provider account by going to Data Sharing -> Provider Studio -> Listings.
Cross-Cloud Auto-Fulfillment ensures that your data is securely and automatically delivered to consumers across any region or cloud.
Step 3: Consumers within or across organizations can access your listing
Consumers can go to Data Products -> Private Sharing to find the private listing, and then click Get. The consumer can be within the same account (internal sharing), within the same organization (internal marketplace), or across organizations (external sharing on Snowflake Marketplace).
Wrapping up
Snowflake is extending its zero-ETL data-sharing leadership by integrating open table formats to enable secure, governed and cost-optimized data sharing across any cloud and region. To see these powerful new capabilities in action and learn how to get started, check out this solutions page featuring a demo and quickstart!
*Limitation: Currently, catalog federation does not support catalog linked databases. See limitation in documentation here.
Authors
Just For You
APR 02, 2025|5 min read
The Apache Iceberg Avalanche: How the Open Table Format Changes the Face of Data Lakes
Learn how Apache Iceberg™ enables open, high-performance data lakehouses—combining warehouse reliability with lake flexibility and vendor-neutral architecture.
Make Your Lakehouse AI-Ready
Snowflake announces new data engineering capabilities for open table formats, empowering you to build a connected, high-performance lakehouse on Iceberg and Delta Lake.
How Snowflake Delivers a Single Data Experience Across Multiple Clouds and Regions
Snowflake was built on top of public cloud infrastructure & can deliver a single date experience across multiple clouds & regions.
Expand to New Regions with Zero Additional Egress Costs
Snowflake Egress Cost Optimizer enables cost-efficient, secure data sharing across clouds and regions—cutting egress costs without added latency.
Unifying Iceberg Tables on Snowflake
Learn about our new Apache Iceberg features offering seamless integration for unified table management, improved performance and hassle-free catalog interaction.
Unmatched Collaboration for Data & AI Products: What’s New
Snowflake’s latest features enable seamless data and AI collaboration across teams and clouds, supporting secure sharing, scaling, and advanced app integration.
Salesforce and Snowflake Bidirectional Data Sharing in GA
Bidirectional data sharing between Snowflake and Salesforce Data Cloud is now available to streamline processes and accelerate decision-making across industries.
Snowflake and Workday Accelerate the AI-Driven Enterprise
Snowflake and Workday announce a zero-copy, bidirectional partnership, enabling customers to securely share data and interoperate AI agents for faster, unified insights.
Snowflake Cortex AI Achieves FedRAMP Moderate on AWS Region US East (N. Virginia)
Snowflake Cortex AI is now FedRAMP Moderate on AWS Region US East (N. Virginia), allowing U.S. federal agencies to use generative AI securely on sensitive data.
