Data became valuable the moment we started generating it at scale. As organizations began storing it by region — each with its own compliance rules, protocols, and security boundaries — the challenge shifted to: how do we share and consume data across regions securely, efficiently, and with minimal friction?
Enter Delta Sharing: a modern, open, and cost-effective way to share data across boundaries in real time, without sacrificing governance or performance. In this blog, we’ll walk through how Delta Sharing makes cross-region data collaboration effortless — hello world, data liberated.
Delta Sharing is an open protocol for secure real-time exchange of large datasets, which enables organizations to share data in real time regardless of which computing platforms they use. In Simple terms, you can share a huge amount of data, real-time hassle hassle-free. It works on the REST protocol, REST protocol is a powerful protocol that helps to connect both ends, the server and receiver. REST protocol uses HTTP methods to share data. Now, to set up the Delta sharing, we require certain things!! which helps to work with the open protocol with ease. Let’s have a look at it.
Share Identifier — As Delta sharing uses the REST protocol, HTTPS requests are used. HTTP header used to authorize, we have to provide the Bearer token to authorize. In Deta, sharing to get the Bearer token, we have to navigate to Catalog → Delta sharing → shared with me → name of demo → copy sharing identifier
Share — A share is a read-only collection of tables and table partitions that a provider wants to share with one or more recipients. It helps in organizing and categorizing shared data assets, making it easier to manage permissions and access controls. In short, Share is the basket in which we add the objects or assets that we want to share.
Recipient — A recipient identifies an organization with which you want to share any number of shares. There are 2 types of recipients in delta sharing.
- Databricks-to-Databricks Recipients: These recipients have access to a Databricks workspace enabled with Unity Catalog. They authenticate using a sharing identifier, a unique string that identifies their Unity Catalog metastore. Share Identifier — As Delta sharing uses the REST protocol, HTTPS requests are used, and the HTTP header is used to authorize. We have to provide the Bearer token. To get the Bearer token, we have to navigate to Catalog → Delta sharing → shared with me → name of demo → copy sharing identifier
- Open Sharing Recipients: These recipients do not use Databricks or Unity Catalog. They authenticate using token-based methods, such as bearer tokens or OAuth federation. This approach enables data sharing with external users or systems outside of the Databricks environment.
Shared Catalog — When we create share from provider end, in recipient end we must create the catalog, were we store our shared views and tables or any AI assets. In future, whenever we run the provider side script to add the asset in share, then it gets dumped into the shared catalog.
Now if want to share the assets in much controlled way, such that which view or table to share, which table or view not to share on recipient side, we can do it, using the script. we can play around the customized script and add only those objects which we want to add and grant only those objects we want to grant access to users or group.
we can create an excel in which we can mention object name, AAD group name to grant access and read that excel and then add that object in share and attach that share to recipient and in another side, we can create catalog, and shared object can we visible in that shared catalog.
Delta Sharing with Iceberg Tables and Materialized Views: Modern Cross-Region Data Collaboration
When you need to share data across regions — each with its own compliance, security boundary, and protocol — moving or copying data manually becomes brittle, expensive, and slow. Delta Sharing is the modern answer: it’s an open, real-time, secure protocol for sharing governed data without replication. What makes it even more powerful today is that you can share:
- Managed Iceberg tables (open, versioned, multi-engine native tables)
- Materialized views (pre-computed query results for performance)
All of these can be included directly in a Delta Share, so recipients get live, governed access with minimal friction.
Customized Script — If we want to share the data with much controlled way, then we have to create customized script, in customized script we take excel file as input
