Configuration via environment variables should be available
Current service configuration (according to doc-s) is done using config files.
Main configurations should be done via environment variables. This would provide ability to reconfigure service using the same artifact (image).
Example configurations that should be done via environment variables:
- all variables used for database connection (url, user, password)
- enabling caching via redis
- tenant name and any variables related to the ability to work with multitenancy
- log level
- any other configurations that are dependent on infrastructure configuration
These variables will be later set using ConfigMaps and Secrets, which will allow Ops team to swiftly react to any infrastructure changes, configurations for different environments in pipelines and re-usage of the same artifact for different environments.