←View full DeepEval status
DeepEval: Model Registry / Artifact Store Error
Current Status: Operational
Last checked: 8m ago
What We're Seeing Right Now
No recent issues reported. If you're experiencing problems with DeepEval, report below to help the community.
What is this error?
You cannot push, pull, or access model artifacts in DeepEval's model registry. This blocks model deployment and versioning workflows. Registry errors are often caused by storage backend issues or permission problems.
Error Signatures
Failed to push artifactArtifact not foundAccess denied to model registryStorage quota exceededConnection timeout to registryModel version not foundRegistry authentication failedCommon Causes
- Object storage backend (S3, GCS, Azure Blob) experiencing issues
- Permission or IAM misconfiguration blocking access to artifacts
- Storage quota exceeded on your account or workspace
- Network connectivity issues between the registry and your environment
- Platform-side registry service degradation
✓ How to Fix It
- Verify your credentials and permissions for the registry
- Check your storage quota and usage in DeepEval's dashboard
- Test connectivity to the underlying storage bucket directly
- Check DeepEval's status page for registry-specific incidents
- Try accessing the registry from a different environment or region
- Contact support with the artifact path and error message
Live Signals
Service Components
DeepEval Web
OperationalRecent Incidents
No incidents in the past 30 days
Frequently Asked Questions
Is my model data safe during a registry outage?
Yes. Registry outages typically affect access, not data integrity. Your model artifacts stored in the underlying object storage (S3/GCS) are durable and should be intact once the service recovers.
Can I deploy a model if the registry is down?
If you have the model weights cached locally or in your own storage, you may be able to deploy directly. Otherwise, you'll need to wait for the registry to recover.
How do I back up my model artifacts independently of DeepEval?
Always store a copy of critical model weights in your own cloud storage bucket. Configure DeepEval to use your own S3/GCS bucket as the artifact backend so you retain full ownership.