Create Kubernetes deployment for testing Studio with and without GPU

[karlmutch]

As an integrator
I want to have available yaml files for deploying standalone studioML tests, and tests for AWS SQS/S3 deployments
In order that tests can be automated as a Kubernetes job on public and private infrastructure

The objective is to allow third parties to run the fully suite of GPU and non-GPU tests that leverage three configuration types on top of a Kubernetes deployment.

SQS with S3
Google PubSub with Google cloud storage running the S3 compatibility features
Standalone RabbitMQ and Minio in microk8s (non GPU), and any cloud providers generic Virtual Machines with stock Kubernetes

yaml should allow for templates, possibly using Go/Kubernetes Go notation github.com/karlmutch/duat/stencil so that images used during the test can be pulled from either local workstations (microk8s only), or from public image repositories such as Quay.io