Hey Rick 👋 I gave a tour of our setup at
tails.com at
Demo Day few weeks ago that you might find interesting. We treat meltano pretty much like any other app/service; branch based deployment through CI/CD onto kubernetes.
In prod we run one kubernetes pod each for the airflow UI, airflow scheduler and the meltano UI processes. Database is AWS RDS. We also use the KubernetesPodOperator in Airflow in stead of the default executor so that each of our invocations of
meltano elt ...
gets their own pod.