AWS re:Invent 2025 - Seamless data sharing in Amazon Redshift (ANT348)

Seamless Data Sharing in Amazon Redshift: Improving Agility and Governance

Introduction to Amazon Redshift Data Sharing

• Amazon Redshift data sharing is a mature capability that allows organizations to share live data securely and easily across Redshift warehouses.
• It enables data sharing within the same AWS account, across different accounts, and even across regions, for both read and write workloads.
• Data sharing eliminates the need for manual data copying and movement, complex ETL pipelines, and the resulting data staleness and governance challenges.

Technical Overview of Redshift Data Sharing Architecture

• Redshift's storage and compute isolation powers the data sharing architecture.
• Multiple producer warehouses can share data through Redshift-managed storage to multiple consumer warehouses.
• Consumers can also share their data with other warehouses, enabling a flexible, multi-directional data sharing topology.
• Users and applications can access the most up-to-date and consistent data without the need for manual data copying.

Data Share Management Options

1. Redshift-Managed Data Shares: Data shares are created and permissions managed within the producer Redshift warehouse, then shared directly with consumer warehouses.
2. Lake Formation-Managed Data Shares: Data shares are created in the producer warehouse but permissions are centrally managed through AWS Lake Formation.
3. AWS Data Exchange for Amazon Redshift: Customers can share their data as a product through the AWS Data Exchange, allowing subscribers to access the data.

Common Use Cases for Redshift Data Sharing

1. Workload Isolation: Splitting a monolithic Redshift warehouse into specialized, purpose-built warehouses for different workloads (ETL, BI, data science, etc.), connected via data sharing.
   • Example: Peloton implemented this to save $300,000 annually.
2. Cross-Group Collaboration: Enabling seamless data access and sharing across business teams (finance, sales, marketing, R&D, etc.) for broader analytics and data science.
   • Example: Fenmai built a central data marketplace using Redshift data sharing.
3. Delivering Data as a Service: Securely sharing live data with internal and external parties as a data product, with usage monitoring and control.
4. Sharing Data between Environments: Enabling agile data access between development, test, and production Redshift environments without the need for data copying.

Enhancing Governance with AWS Lake Formation

• For complex data sharing topologies with multiple producers and consumers, Lake Formation can simplify permission management by centralizing access controls.
• Lake Formation allows centralized definition and enforcement of data sharing policies, eliminating the need for manual scripting and complex coordination.

Best Practices and Considerations

• Pricing: Producer warehouses are charged for data storage, while consumers pay for compute usage to access shared data.
• Cross-region data sharing incurs additional data transfer fees.
• Query performance depends on the consumer cluster's compute capacity, not the producer's.
• Redshift supports concurrency scaling to handle unpredictable workloads.
• Data is encrypted in transit and at rest, with options for separate KMS keys for producers and consumers.
• Redshift system tables can be used to audit data share usage and changes.

Getting Started with Redshift Data Sharing

• The presentation provides a QR code linking to resources, including blog posts and best practice documents, to help customers get started with Redshift data sharing.