The current behavior of workflow-manager is to use the AWS API to find the RDS instance that it should talk to. While very useful for production (the server need not be pre-configured with RDS specifics), it's more difficult to run an instance for development or integration testing.
This issue is for refactoring the server so that it can talk to:
An in-memory SQLite DB (similar to what we do for tests)
An in-cluster PostgreSQL DB (useful for integration tests or development)
This work should be done after #138 to prevent a very large merge conflict
The current behavior of workflow-manager is to use the AWS API to find the RDS instance that it should talk to. While very useful for production (the server need not be pre-configured with RDS specifics), it's more difficult to run an instance for development or integration testing.
This issue is for refactoring the server so that it can talk to:
This work should be done after #138 to prevent a very large merge conflict