Databricks Azure experienced a minor incident on May 21, 2026 affecting AI/BI and Compute and 1 more component, lasting 3h. The incident has been resolved; the full update timeline is below.
Affected components
Update timeline
- investigating May 21, 2026, 06:55 AM UTC
We are actively investigating an issue with the Databricks service. Next Update We will provide another update within 1 hour, or sooner if there is a material change in service status.
- investigating May 21, 2026, 06:56 AM UTC
We are actively investigating an issue with the Databricks service. Next Update We will provide another update within 1 hour, or sooner if there is a material change in service status.
- investigating May 21, 2026, 06:56 AM UTC
We are actively investigating an issue with the Databricks service. Next Update We will provide another update within 1 hour, or sooner if there is a material change in service status.
- identified May 21, 2026, 08:08 AM UTC
Starting at approximately 06:07 UTC on May 21, 2026, a subset of customers using Azure Databricks may be experiencing errors or failures when launching or running workloads on Classic Compute, Serverless Compute, and Jobs. Affected customers may observe cluster start failures, job run failures, or degraded availability across these services. Symptoms • Cluster launches failing with errors when attempting to start Classic Compute or Serverless Compute workloads • Job runs failing or not completing due to compute provisioning errors • Degraded availability or errors when starting or scaling compute resources Recommendations No customer action is required at this time. Current Status We have identified the cause of degraded availability affecting Classic Compute, Serverless Compute, and Jobs, and are monitoring recovery across all impacted services. Next Update We will provide another update within 1 hour, or sooner if there is a material change in service status.
- monitoring May 21, 2026, 08:19 AM UTC
Between 06:07 UTC and 07:05 UTC on May 21, 2026, a subset of customers using Azure Databricks may have experienced degraded availability, cluster launch failures, or job run failures affecting Classic Compute, Serverless Compute, and Jobs. Serverless Compute began recovering earlier, with availability confirmed restored by approximately 06:35 UTC, while Classic Compute was fully mitigated by 07:05 UTC. Symptoms • Cluster launches failed or returned errors when attempting to start Classic Compute or Serverless Compute workloads • Job runs failed or did not complete successfully due to compute provisioning errors • Degraded availability or timeouts when attempting to start or scale compute resources Current Status The issue has been mitigated and services have been restored. Next Update We will provide another update within 1 hour, or sooner if there is a material change in service status.
- monitoring May 21, 2026, 08:19 AM UTC
Between 06:07 UTC and 07:05 UTC on May 21, 2026, a subset of customers using Azure Databricks may have experienced degraded availability, cluster launch failures, or job run failures affecting Classic Compute, Serverless Compute, and Jobs. Serverless Compute began recovering earlier, with availability confirmed restored by approximately 06:35 UTC, while Classic Compute was fully mitigated by 07:05 UTC. Symptoms • Cluster launches failed or returned errors when attempting to start Classic Compute or Serverless Compute workloads • Job runs failed or did not complete successfully due to compute provisioning errors • Degraded availability or timeouts when attempting to start or scale compute resources Current Status The issue has been mitigated and services have been restored. Next Update We will provide another update within 1 hour, or sooner if there is a material change in service status.
- resolved May 21, 2026, 09:56 AM UTC
Between 06:07 UTC and 07:05 UTC on May 21, 2026, a subset of customers using Azure Databricks may have experienced degraded availability, cluster launch failures, or job run failures affecting Classic Compute, Serverless Compute, and Jobs. Serverless Compute began recovering earlier, with availability confirmed restored by approximately 06:35 UTC, while Classic Compute was fully mitigated by 07:05 UTC. Symptoms • Cluster launches failed or returned errors when attempting to start Classic Compute or Serverless Compute workloads • Job runs failed or did not complete successfully due to compute provisioning errors • Degraded availability or timeouts when attempting to start or scale compute resources Recommendations No customer action is required at this time. Current Status This incident has been fully resolved as of 21 May 2026 07:05 UTC We apologize for the disruption. If you require further details about this incident, please submit an Azure support ticket. Back to current status Status History Subscribe to receive status updates by email Close Subscribe Manage Subscription Manage Existing Subscription Create New Subscription Subscribe to receive status updates by webhook Each status update will POST a JSON payload to this URL Email address for managing webhook Close Subscribe Manage Subscription Manage Existing Subscription Create New Subscription Subscribe to receive status updates in Microsoft Teams Enter your Microsoft Teams webhook. View Instructions Email address for managing subscription Close Subscribe Manage Subscription Manage Existing Subscription Create New Subscription Subscribe to receive status updates in Slack Slack channel ID Find the channel ID: Select the channel in your Slack workspace. The channel ID is displayed in the browser URL. Example: https://app.slack.com/client/T04SJBK1C/ C03SKGJ1P Email address Close Manage Subscription Manage Existing Subscription Create New Subscription