Home/Docs/Local Continuity

Local Continuity & Offline Fallbacks

For mission-critical operations, relying entirely on public cloud APIs introduces single-point-of-failure vulnerabilities. Selixes features an automated Continuity Mode that shifts workloads to local inference engines (such as Ollama or vLLM containers) when WAN routes degrade or global cloud blackouts occur.

1. Setting Up Local Inference (Ollama)

Run a local continuity node using Docker. Ensure the port 11434 is exposed so that the Selixes gateway can route requests to it:

# Pull and run the official Ollama Docker container
docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama

# Download a compatible model checkpoint (Llama 3.1 8B recommended)
docker exec -it ollama ollama run llama3.1:8b

2. Configure Gateway Continuity Routes

Instruct the gateway where your local inference endpoint is located by configuring the environment variables or passing matching headers:

# Environment variables for Selixes gateway container
LOCAL_CONTINUITY_ENABLED=true
LOCAL_CONTINUITY_URL=http://localhost:11434
LOCAL_CONTINUITY_MODEL=llama3.1:8b

Autonomic Engagement

When the gateway trips its cloud failover circuit (all public routes returned errors or timed out), it engages Continuity Mode:

  • The gateway interceptor intercepts the outgoing request.
  • It maps the query format to match the local model parameters.
  • The request is processed locally at $0.00 token cost, guaranteeing system availability during complete ISP or backbone network outages.