SUSE AI Factory
SUSE AI Factory standardizes AI platform management, replacing manual integration with version-controlled, reproducible blueprints to streamline enterprise workloads. Deploying SUSE AI Factory in air-gapped environments allows organizations to install and run artificial intelligence (AI) workloads securely within restricted networks without external internet connectivity.
Publication Date: 2026-11-20
- 1 What should I consider before an air-gapped deployment?
- 2 What are the requirements for an air-gapped deployment?
- 3 How do I deploy SUSE AI Factory in an air-gapped environment?
- 3.1 Downloading artifacts
- 3.2 Make downloaded data available to the air-gapped environment
- 3.3 Set up OCI-compliant registries in the air-gapped environment
- 3.4 Upload Helm charts and containers
- 3.5 Create a PVC for the model cache
- 3.6 Transfer model cache to the PVC
- 3.7 Configure Helm charts to use the model cache PVC
- 3.8 Configure RKE2 private registries
- 4 How do I air-gap Ollama and blueprints?
- 4.1 Placeholders used in this topic
- 4.2 Blueprint architecture
- 4.3 Prerequisites
- 4.4 Identifying blueprint artifacts
- 4.5 Downloading artifacts on the connected host
- 4.6 Downloading Ollama models
- 4.7 Transferring and uploading artifacts
- 4.8 Pointing chart repositories at the mirror
- 4.9 Configuring RKE2 private registries
- 4.10 Loading Ollama models into a PVC
- 4.11 Configuring the Ollama component to use the PVC
- 4.12 Applying and verifying the deployment
- 4.13 Considerations