dbt Cloud Outage History

dbt Cloud's Azure US multi-cell environment experienced a complete service outage for 2.8 hours, preventing users from accessing the platform entirely. The incident was caused by a routine database vertical scaling operation that failed silently due to Azure capacity constraints. The service was restored after direct intervention to recover the database server, with full functionality returning at 7:43 AM EST.

dbt Cloud's Copilot/AI functionality experienced issues for 1.4 hours due to an upstream OpenAI service outage. The incident was classified as minor and affected the AI-powered features within the dbt Cloud platform. The service was restored once the underlying OpenAI outage was resolved.

dbt Cloud experienced orchestration job failures in the US multi-tenant cell C2, causing source freshness and compare steps to fail for affected customers. The issue was identified and a fix was deployed within 1.6 hours, returning the orchestration jobs to normal operation.

dbt Cloud experienced delays in audit log event updates for 3.8 hours, preventing users from seeing recent audit log activity in real-time. The engineering team identified the root cause and deployed a fix to restore normal audit log processing. Audit log events are now updating normally with continued monitoring in place.

dbt Cloud experienced an issue where Snowflake customers using dbt Fusion with OAuth authentication encountered "Configuration Error: Snowflake OAuth authentication does not support 'user'" errors when opening the dbt Studio IDE. The problem affected users' ability to access the Studio interface and prevented them from working with their dbt projects. A fix was deployed after 50 minutes and the issue was resolved with continued monitoring to ensure full effectiveness.

A runtime error related to 'table_subtype' in the Redshift adapter caused scheduled dbt Cloud jobs to fail for 5.4 hours. The team identified the root cause and deployed a fix, restoring normal operations. All affected components returned to normal state with continued monitoring.

dbt Studio in Azure US was unavailable for 24 minutes due to a service issue. The outage prevented users from accessing dbt Studio functionality during this period. The team deployed a fix that resolved the problem and restored normal service operations.

The dbt Cloud application became completely unavailable in the US GCP instance (us3) for 27 hours due to an unspecified system issue. The outage affected all users on that instance, preventing access to the dbt Cloud application entirely. Engineering teams deployed a fix to restore functionality and continued monitoring the system for stability.

dbt Cloud CLI and Studio experienced elevated 5xx errors globally for 3.1 hours, preventing users from running commands and causing "session occupied" errors. The incident occurred in two phases: initial 5xx errors were resolved after 30 minutes, followed by session occupied errors in Studio that persisted for another 2.5 hours. Both issues were identified, fixed through deployments, and services returned to normal operation.

The dbt Cloud Metadata API experienced an issue causing all GraphQL queries to return HTTP 500 errors for 21 minutes. The problem affected all API calls to the Metadata API service. A fix was deployed and the API returned to normal operation with continued monitoring.