As per the documentation, Spring Cloud Dataflow Service used RDBMS for storing stream/task definitions, application registration and job repositories. Instead of using RDBMS, is there a way to use Redis for storing this information.
RDBMS is the default repository implementation for the Data Flow server core. You can still override this default repositories (except task/job execution repositories) by having Redis based implementations from your custom Data Flow server configuration. While you can have 'redis' based repositories for stream/task definitions and application registration, you still need to have RDBMS for the task/batch job execution repositories. That's the reason Spring Cloud Data Flow by default goes with RDBMS based ones for all.