Installation

Prerequisites

The Scheduler supports both 3.10 and 3.11 Python version. Newer versions are not tested yet. For the list of dependencies check the pyproject.toml in the backend folder.

For the Docker installation the Docker daemon must be running on the local machine.

Env variables

The system relies heavily on environment variables to function. In the source code we provide a .env.exmaple to fill up necessary variables. You can read more about what each variable represent in the Configuration page.

Runtime calculations and ephemerides files

The Scheduler can be configured to work with both visibility calculations done in runtime and with Sight. If the first option is selected, due to the amount of data we store all the ephemerides files for 2018B Semester that are needed to create the whole semester. Is possible to skip this step but the performance of the Scheduler would be severely hindered. The files are in a .bz2 compressed file in /scheduler/scheduler/services/horizons/data/ and it needs git-lfs to be cloned from the repo. You can install it from here

To unzip you can run:

$ tar -xjf /scheduler/scheduler/services/horizons/data/ephemerides.tar.bz2

Installation

First clone the monorepo.

$ git clone https://github.com/gemini-hlsw/scheduler.git

:bulb: Remember that if you use runtime visibility you need to use Git LFS

git lfs install
git lfs pull

After that you have two options using Docker or using a local setting.

Local installation

To run a local development version first make sure your environment have the appropriate variables already set. This will change depending on the shell you are using. In bash and zsh you can copy the .env.example file to a new .env file, fill the secrets in this new .env file and run

source .env

Also, the PYTHONPATH environment variable should be properly set to add the backend directory of this repository. Add the following line to your ~/.bash_profile or add it to the console session:

$ export PYTHONPATH=$PYTHONPATH:{path-to-project-base} # should end in scheduler/

Weather

If REALTIME mode wants to be used a properly running weather service is required. Check external-services/weather for detailed instructions.

Backend

cd backend
uv sync --group gpp-dev --no-group gpp-prod
uv run python scheduler/main.py

Frontend

pnpm install
pnpm frontend dev

Docker installation

You also need to source the .env file here so docker can pickup the correct variables. Then build and start all services:

docker compose build
docker compose up

Service	URL	Mode
Frontend	http://localhost:80	—
Backend (Validation)	http://localhost:8001/graphql	VALIDATION
Backend (Realtime)	http://localhost:8000/graphql	REALTIME
Weather	http://localhost:4000	—

Updating the GraphQL schema

When the backend schema changes, regenerate backend/scheduler.graphql before rebuilding:

uv run python scripts/export_graphql_schema.py
docker compose build frontend