gcp-ts-k8s-ruby-on-rails-postgresql example app not working

The TODO app container appears to be invalid. I followed the README and successfully created the cluster and deployed the app, but the container is crash looping.

I connected to the cluster to debug by running

pulumi stack output kubeConfig > kubeconfig

and then running

KUBECONFIG=$PWD/kubeconfig kubectl logs -l app=rails-app

I saw the following logs:

Example:
    rails new ~/Code/Ruby/weblog

    This generates a skeletal Rails installation in ~/Code/Ruby/weblog.
* CREATING DATABASE (first time only):
* MIGRATING DATABASE (if needed):
* RUNNING...
/myapp/start: line 10: bin/rails: No such file or directory
/myapp/start: line 12: bin/rails: No such file or directory

It appears that the start script is not working. Changing it to

#!/bin/bash

# Spawn the Rails App.
echo "* SPAWNING RAILS:"
rails new . -f
rails s -p 3000 -b '0.0.0.0' &

# Wait for it to come up, and then ensure the database is created and migrated.
sleep 5
echo "* CREATING DATABASE (first time only):"
rails db:create
echo "* MIGRATING DATABASE (if needed):"
rails db:migrate

echo "* RUNNING..."
wait

got the server to start and stopped the container from crash looping, but when I navigate to the page, I'm seeing DB connection errors: ActiveRecord::ConnectionNotEstablished No connection pool with 'primary' found.

Start with the example's README and the container's start script, then reproduce the deployment using the shown pulumi stack output kubeConfig and kubectl logs commands. Trace both the missing bin/rails failure and the PostgreSQL connection error; done means the container stays running and the Rails app connects to its database.