Forge Intermittent Web Page Unavailability
Weights & Biases reported that Forge web pages and the inference API were intermittently unavailable. The Frontend and Inference Endpoint components were affected. Weights & Biases resolved the incident after 2 hours and 48 minutes.
State
Resolved
Started
Lasted
2 h 48 min
Worst impact
Partial outage
Timeline
8 updatesInvestigating
17 min 10%
Identified
1 h 42 min 61%
Monitoring
49 min 29%
Updates
Times in UTCAs posted by Weights & Biases
We are seeing a return to nominal performance for the last 30 mins along with corresponding internal metrics, and will mark this incident as resolved.
We are seeing stability return back to normal for the affected services, and are continuing to monitor uptime.
We have deployed the comprehensive patch, and are seeing improvements on our internal metrics and external error rates. We will continue monitoring the results to validate the impact.
We have not seen a significant improvement from our initial mitigations, but we are finalizing testing on the more comprehensive patch that will be rolled out shortly pending verification. We will update at that time, and monitor the incident from then on.
We are deploying a temporary configuration change around the integration rate limits to mitigate the impact while we are working on a larger architectural change to resolve the larger root cause. We will monitor for the impact of the mitigating fix and will update when we are able to push out the more comprehensive patch.
We have identified the technical root cause and are looking into a patch to resolve the rate limit errors that are blocking user flows.
We are investigating an issue where the forge site web pages and the inference API are intermittently unavailable due to underlying rate limit issues.
We are investigating an issue where the forge site web pages are intermittently unavailable due to underlying rate limit issues.
Affected components
2Weights & Biases in the last 90 days
Full historyIncidents in 90 days
1
Days with an incident
1
Median duration
2 h 48 min
This incident
About the median
Follow Weights & Biases
Your coding agent knows when Weights & Biases is down, and tells you when a failure is theirs, not yours. You can also get Weights & Biases 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.cc/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 "wandb" and tell me if Weights & Biases 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.cc/feed.xml?providers=wandb/feed subscribe https://outages.cc/feed.xml?providers=wandbFAQ
Go from tracking outages to fixing production issues with Polylane.
Try Polylane