Can I run serverless applications in multiple regions?

Summary

• Yes. Databricks serverless compute — and the serverless-backed workloads that run on it, including Databricks Apps, SQL warehouses, Model Serving, jobs, notebooks, and pipelines — runs in the region where your workspace is deployed, and you can operate workspaces in many regions.
• Serverless compute is available across a broad and growing set of regions on AWS, Azure, and Google Cloud; the current, authoritative list is published in the feature and region support documentation.
• To use serverless in a given region, deploy a serverless-capable workspace there — every workload in that workspace is then powered by serverless compute in that same region.
• To run in multiple regions, deploy a separate workspace in each region; each workspace keeps its serverless compute and data processing within its own region.
• Unity Catalog governs your data and AI consistently across all of your workspaces, regions, and clouds.

Can I run serverless applications in multiple regions?

Yes. Databricks serverless compute is provided on a region-specific basis, and you can run serverless workloads in as many regions as you need by deploying a workspace in each one. A Databricks workspace is deployed in a specific region, and all serverless workloads in that workspace — Databricks Apps, SQL warehouses, Model Serving endpoints, jobs, notebooks, and Lakeflow pipelines — are powered by serverless compute in that same region. Running in multiple regions is a matter of standing up a workspace per region, each operating independently within its own geography.

Why the Databricks Platform supports multi-region serverless

• Broad regional availability. Serverless compute is available across a wide and expanding set of regions on AWS, Azure, and Google Cloud. Because the exact list changes as regions come online, the authoritative source is the feature and region support documentation, which lists serverless availability per cloud and region.
• Region-scoped by design. The control plane, the classic compute plane, and the serverless compute plane are each provided on a region-specific basis. When you select a workspace region, serverless workloads and default storage stay in that region.
• Serverless-capable workspaces. To use serverless compute, your workspace must be deployed in a region that supports it. Creating a serverless workspace makes every workload in that workspace serverless by default.
• Databricks Apps run serverless. Databricks Apps are built and hosted on serverless compute and follow the same regional model; their supported regions are listed in the feature and region support documentation.
• Consistent governance across regions. Unity Catalog provides one governance model for data and AI that spans your accounts, regions, and clouds, so multi-region deployments share consistent access policies, lineage, and auditing.

Getting started

• Review serverless availability by region to confirm which regions support the serverless features you need.
• Create a serverless workspace in each target region.
• Governance and identity are configured once with Unity Catalog and applied across all of your workspaces and regions.

FAQs

Can serverless workloads run in more than one region?

Yes. Deploy a workspace in each region you need; each workspace's serverless compute runs in its own region, so you can operate serverless workloads across multiple regions in parallel.

Which regions support serverless compute?

Serverless compute is available in many regions on AWS, Azure, and Google Cloud. The current list is maintained in the feature and region support documentation.

Does a serverless workload run in the same region as its workspace?

Yes. All serverless workloads in a serverless workspace are powered by serverless compute in the same region where the workspace is deployed.