Unable to connect to RDS postgres DB, check-db container fails

[dylac]

Airflow version: 2.2.5
Chart version: 8.6.1

I am trying to point to an externalDatabase in my helm chart and this is an RDS multi-AZ Postgres deployment. Notably, I have another environment with the exact same airflow and helm chart version, literally the only difference is of course the RDS hostname. This environment is working great, although it initially failed until I uninstalled/re-did the install (same namespace same config):

helm upgrade --install --wait=false -f airflow-workingenv.yml airflow airflow-stable/airflow -n airflow-workingenv

I know adding --wait=false fixed a previous race condition I was having with this but I don't remember exactly what. I have tried different versions of including/excluding this option, deleting and recreating the whole namespace, resetting the DB itself. And of course I checked the DB itself is up and fine, I can psql connect to it from a pod in the same cluster.

I am not sure if I need to run something like airflow db init or airflow db reset in the namespace but I can't get an airflow pod running long enough to do so. They all fail on the check-db init container after getting our airflow image with simply:

Back-off restarting failed container

Is there a way to get more logs here, or a way to do a one-time skip of this check? I have also tried enabled: false in dbMigrations: or enabled but with runAsJob: true. I'm very confused because again this literally identical to my working env.