Web interface to SIM-CITY

Contains a Python JSON webservice for a SIM-CITY webserver. It uses the API from SIM-CITY client to create simulation tasks, start new pilot jobs on computing infrastructure and query the status of the simulations and jobs. Documentation of the API is in docs/apiary.html.

Building and deployment

Following the SIM-CITY client API, a ./config.ini file should be created containing the configuration of the CouchDB service that will keep the administration of the tasks and jobs, and the type of simulations that can be served. The config.ini.dist file can be used as a template for this.

Install by calling

pip install -r requirements.txt .

and serve on port 9090 using make serve.

Building and deployment with Docker

With docker, services are compartmentalized and installations standardized. First, configuration files need to be put in place. First run

./configure-docker.py

1. Osmium: edit docker/osmium/joblauncher.yml. See also docker-osmium repository.
2. SIM-CITY webservice: edit docker/webservice/config.ini.

To run the docker components, run

. couchdb.env
make docker-run

The web-service is accessible from port 9090, the CouchDB server on port 5984.

Before running again, run

make docker-clean

Testing

Install test dependencies with

pip install -U ".[test]"

Run unit tests with

pytest tests

and all tests with

pytest --flake8

The last command requires docker-compose to be installed. To run custom tests, go to the integration_tests/docker folder and run docker-compose up --build. For more instructions see integration_tests/docker/README.md.