Add Crunchy Postgres Operator as Default

[ranchodeluxe]

What

It's typical for most k8 demos to use an in-memory database by default. I think our default should just install the Crunchy Postgres Operator. Then folks have high-available k8s alternative already set up to go

AC

• refactor db to just be a fixture loader for tests and examples
• make fixture loader and all services mount and pick up PGO created secrets
• how does using PGO dependencies affect publish eoapi as helm chart?
• write documentation
• pgo keeps us from having multiple deploys in the same namespace so rewrite ci/cd to use new namespaces per release and delete the namespace (ensuring everything in it is destroyed) on teardown
• the tipg fixture data is in the DB and the user has all the proper grants but we still cannot see the data (this seems to be b/c the /refresh logic of the background thread is not working as expected)
• file crunchydata issue on deps not being removed during helm uninstall eoapi
  • Not All Postgrescluster Resources Deleted on Helm Uninstall CrunchyData/postgres-operator#3928
• file crunchydata issue on not having an OCI repo for the postgrescluster chart
  • Install Postgres Cluster via Helm from OCI Registry CrunchyData/postgres-operator#3641
• write more docs around /eoapi/charts/[pgo | postgrescluser]

[ranchodeluxe]

Unfortunately crunchydata doesn't have a helm repo where they publish these so their docs recommend cloning their repo and installing from there. Nor are both charts OCI registries. So we have to copy pasta here

[ranchodeluxe]

This turned out to be a beast 😉 I've added some FINAL AC to finish this off but no reason not to start reviews

[ranchodeluxe]

merged