Switch approach for optional external database handling for Cloud Native Hybrid environments

When using external database servers, e.g. AWS RDS, with the Toolkit it will idempotently check if the required databases, users and extensions are set up and proceed to configure those if not in a graceful fashion.

This is currently done with the Ansible postgres collection via an Omnibus node as it's expected that the database server is internal only and therefore not accessible to Ansible directly by default.

For future Cloud Native scenarios where there may not be any Omnibus nodes we should switch to a better solution for all potential use cases. One such option is to deploy a container temporarily into the cluster supporting pool and run psql from there. Task is to explore.