Resolved
The incident now has been resolved.
The model DeepSeek-V4-Pro is now fully operational.
Monitoring
The node has been replaced.
The model is currently starting. Expect delay 10-30 mins
Identified
The provider still hasn't fixed the node issue.
Our team is finishing the node switch
Identified
The provider still working on this issue
Our team now is working to change the affected node
Identified
We've identified the issue and it is associated with an up-stream DC provider.
The team now is working with to bring the GPU node back online.
Investigating
We are experiencing a timeouts with DeepSeek-V4-Pro vLLM model. The team are investigating. Updates will follow.
