Skip to main content

Tutorials

Tutorial: Run the control plane locally​

This walkthrough starts the Edge Fleet Control Plane on your laptop for development and smoke testing.

Prerequisites​

  • Python 3.12+
  • git
  • (Optional) Docker, for building images or running edge model containers

Steps​

  1. Clone the repository

    git clone https://github.com/ICICLE-ai/intelligent-edge-management-service.git
    cd intelligent-edge-management-service/edge_fleet_control_plane

  2. Create a virtual environment and install dependencies

    python3 -m venv .venv
    source .venv/bin/activate
    pip install -r requirements.txt

  3. Configure local development

    cp .env.example .env

    For local dev, keep LOCAL_DEV_AUTH=true (default in .env.example). This bypasses Tapis OAuth and logs you in as local_tapis_user.

  4. Start the server

    uvicorn app.main:app --reload --host 0.0.0.0 --port 8000

  5. Open the dashboard

    Visit http://localhost:8000. You should see the ICICLE Edge Control Plane dashboard.

  6. Run the smoke test (optional)

    python tests/smoke_test.py

Expected result​

The dashboard loads, you can browse devices, model cards, and deployments. SQLite stores data under data/edge_control_plane.db.