Skip to content

Latest commit

 

History

History
51 lines (36 loc) · 2.29 KB

README.md

File metadata and controls

51 lines (36 loc) · 2.29 KB
Raw

OpenSidewalks e-mission deployment

This repo contains the deployment scripts and configs for the OpenSidewalks studies that use e-mission. It is based on the use of docker containers using the docker-compose version 3 orchestration framework.

Structure of this repo and deployments

Orchestration YAMLs

The docker-compose YAML files are in the top directory. By default, running docker-compose in the top directory will cascade two configs: docker-compose.yml and docker-compose.override.yml. The override config corresponds to a local dev config with forwarded ports for all services.

Service configs

Each software element is defined as a named service within the docker-compose YAMLs.

Configs for each service are are separated into development or production deployments, so the configuration for the caddy (reverse proxy) service in development (default) mode is in dev/configs/caddy.

Persisted data

Data is persisted using a local directory structure that is similar to configs. For example, in production mode the mongodb service stores its data in prod/data/mongodb.

Deploying in development mode

The default deployment mode is for development and testing purposes and can be instantiated by running docker-compose up in the top level directory. The reverse proxy will proxy the main e-mission service to http://localhost:2015, functionally equivalent to whatever domain the production deployment is pushed to such as emission.mydomain.com.

This mode will use the configs in dev/configs and persist data in dev/data.

Deploying in production mode

To deploy in production mode, you must explicitly cascade the base and production YML files. Starting the services would therefore be: docker-compose -f docker-compose.yml -f docker-compose.prod.yml up. Manipulating any of the services in production mode will require cascading these files, so keep this in mind if you use a subcommand like docker-compose logs, as it will need to be written as docker-compose -f docker-compose.yml -f docker-compose.prod.yml logs.

In addition, the host variable must be set as an environment variable. This is easiest done using a .env file in the top level directory. .env.example is provided as an example. Run cp .env.example .env and edit with plain text editor to configure.