ES-2359687
Databricks AI/BI, Databricks Apps, Genie, and Model Serving have degraded availability in CA Central 1, EU West 1, EU West 2, EU West 3, EU Central 1, US East 1, US East 2, US West 1, US Gov West 1, US West 2, SA East 1, AP South 1, AP Southeast 1, AP Southeast 2, AP Southeast 3, AP Northeast 1, and AP Northeast 2. Databricks reports the issue is identified and mitigation is in progress. Databricks recommends that customers retry failed Claude Opus requests with exponential backoff and fall back to another model for critical workloads.
State
Investigating
Started
Ongoing for
30 min
Worst impact
Partial outage
Timeline
1 updateInvestigating
30 min 100%
Updates
Times in UTCAs posted by Databricks
Impact Summary Starting at 15:50 UTC November 13, 2026, customers using Claude Opus 5 or 5.5 through Databricks may experience degraded availability with Model Serving. Affected customers may encounter failed or slow inference requests to certain foundation models served through Model Serving and AI Gateway. This could also adversely affect Databricks Genie, Apps, or AI/BI. Symptoms • Elevated error rates (HTTP 503) for foundation model API requests • Increased latency for scoring requests • Intermittent failures for streaming inference responses Recommendations Customers experiencing failed Claude Opus requests should retry with exponential backoff and, where possible, fall back to another model for critical workloads. Current Status The issue has been identified and mitigation efforts are in progress in coordination with the cloud provider. Next Update We will provide another update within 1 hour, or sooner if there is a material change in service status.
Affected components
20Databricks in the last 90 days
Full historyIncidents in 90 days
2
Days with an incident
3
Median duration
2 h 51 min
This incident
83% shorter than the median
Other Databricks incidents
All incidentsFollow Databricks
Your coding agent knows when Databricks is down, and tells you when a failure is theirs, not yours. You can also get Databricks incidents in a feed reader or in Slack.
Install the outages.live MCP server, so that when something breaks you can tell if the cause is a provider we depend on and not my code. It reads the official status pages of cloud, AI, payment and developer providers and gives you their incidents as MCP tools.
Add it as a remote MCP server named "outages" at https://outages.live/mcp (Streamable HTTP, no authentication), with your CLI command or your MCP config file.
Then:
1. Make sure it connects and shows the tools get_provider, list_feed and send_feedback.
2. Call get_provider with "databricks" and tell me if Databricks has a problem right now.
3. From now on, when a call to a third-party service fails or becomes slow, call get_provider for that service before you change my code.Paste this into any coding agent. It adds the server with its own MCP setup.
https://outages.live/feed.xml?providers=databricks/feed subscribe https://outages.live/feed.xml?providers=databricksFAQ
Go from tracking outages to fixing production issues with Polylane.
Try Polylane