databricks Platform · Updated Jun 12, 2026

Managed Disaster Recovery

Now in Public Preview for enhanced resilience.

Public Preview Open docs ↗release note ↗

What it does

Managed disaster recovery (DR) replicates your Databricks deployment to a secondary region, allowing for rapid recovery from regional outages. It continuously replicates Unity Catalog metadata, managed table data, and workspace assets, while providing a stable URL that remains consistent through failover events. Users can trigger failovers as needed, ensuring minimal disruption during outages.

Why it matters

This feature significantly enhances business continuity by minimizing downtime during regional outages, thus protecting critical data and applications. Organizations can maintain operational resilience, reduce recovery time objectives (RTO), and ensure that downstream clients remain connected without reconfiguration, ultimately safeguarding productivity and reducing risk.

Key capabilities

Continuous Replication

Replicates Unity Catalog metadata and workspace assets continuously to a secondary region.

Stable URL Support

Provides a stable URL that points to the current primary, ensuring seamless client connectivity post-failover.

Manual Failover Control

Allows users to manually trigger failovers for both planned tests and unplanned outages.

Asset ID Preservation

Maintains workspace asset IDs across regions, ensuring URLs referencing assets remain valid after failover.

Prerequisites

Mission Critical workspace add-on enabled on both workspaces.

Serverless compute enabled on both workspaces.

Account admin role with ALL PRIVILEGES on external locations used by catalogs.

Account-level SSO with all workspaces enabled and identities synced via SCIM.

A secondary workspace and Unity Catalog metastore in the secondary region.

Recommended use cases

Rapid recovery from regional outages to maintain business operations.Testing disaster recovery plans to ensure readiness for actual outages.Maintaining client connectivity without reconfiguration during failovers.Safeguarding critical data and applications in multi-region deployments.

Limitations · remember

!Only specific categories of data and assets are replicated; anything not listed is not replicated.

!Initial replication can take up to two weeks for large workspaces.

!Secondary in-scope catalogs are read-only during the initial replication phase.

!Existing in-scope assets in the secondary workspace are deleted upon initial replication.

Architecture

Primary Databricks Workspace → Secondary Databricks Workspace → Unity Catalog → Replication Pipeline → Stable URL

Get started

Enable Mission Critical

Turn on the Mission Critical add-on for both primary and secondary workspaces.

Create Stable URL

In the account console, navigate to Resilience and create a stable URL for client connections.

Create Failover Group

Define a failover group by specifying primary and secondary workspaces and replication settings.

Map Storage Locations

Add storage mappings for external locations referenced by the primary catalogs.

Monitor Replication

Use the Failover groups tab to check the state and health of the replication process.

Key takeaway

Managed disaster recovery enhances operational resilience by enabling rapid recovery from regional outages with minimal disruption.

2026-06-12-managed-disaster-recovery-is-now-in-public-preview Generated by the Databricks Updates Agent · browse all updates