Skip to content

Running Locally

First local run

From a fresh clone, create an isolated environment and install the dashboard and mobile API dependencies:

python3 -m venv .venv
source .venv/bin/activate
python -m pip install -r requirements-dashboard.txt -r requirements-mobile-api.txt

Run the fast checks before serving the app:

python -m unittest discover -p 'test_*.py'
python check_docs.py

Basic local serve

cd aurora_cloud_dashboard
source .venv/bin/activate
panel serve app.py --address 127.0.0.1 --port 5006 --allow-websocket-origin=<host>

For local AURORACam image display, add a matching static route when you have a local raw mirror:

panel serve app.py \
  --address 127.0.0.1 \
  --port 5006 \
  --allow-websocket-origin=<host> \
  --static-dirs=auroracam-media=/project/aurora/raw/auroracam

What you need locally

  • the Python environment created above
  • access to the configured raw and product roots if you want realistic data
  • a websocket origin that matches your browser host

Notes

  • the deployed application expects /etc/aurora-dashboard.env; do not copy production credentials into a local checkout
  • without real Zarr and quicklook trees, the app should still import and show its empty-data states, but it cannot validate live instruments
  • local development is easiest with a read-only mirror of the deployed data
  • quicklook generators fall back to repo-local quicklooks/ when AURORA_QUICKLOOK_ROOT is not set; source /etc/aurora-dashboard.env before manual deployed runs if the goal is to update the live dashboard products