Install TereminQ - Rancher 2.x HELM ThereminQ stacks on Ubuntu

Install vQbit Rancher 2.x HELM stack on Ubuntu

Minimum requirements:
- ES/Kibana + vQbit node in k3s: 4 cores and 4 GB RAM, 40 GB storage 
- ES/Kibana + vQbit node + Rancher 2 on Docker: 4 cores, 8GB RAM, 60 GB Storage

Advice:
- Install Rancher 2 on a separate controller node with at least 2 cores, 2 GB RAM and 20Gb Storage
- Deploy nodes with K3s/K3d (CPU) and/or Docker (GPU instances)
- Separate ES/Kibana stack from vQbit workloads 
- Set resource limits on the same node when sharing workloads or use separate nodes

# install Docker for CPU+GPU workloads
curl -fsSL https://get.docker.com | bash

and/or 

# install k3s/k3d cluster for CPU-only workloads
curl -sfL https://get.k3s.io | sh -
wget -q -O - https://raw.githubusercontent.com/k3d-io/k3d/main/install.sh | bash

# deploy rancher 2.x
docker run -d --restart=unless-stopped -p 8443:443 rancher/rancher

# start up your browser and point to https://<HOSTNAME>:8443 
# Where HOSTNAME is the IP or Name of the host where you just installed Rancher
# Create a password for the admin user 
# Add the HELM repo in Rancher reposities https://github.com/twobombs/helm.git

# two options; import or create cluster through Rancher 2 UI
- create a cluster (docker) when you want to use an NVidia/AMD/Intel card 
  ( requires docker and NV drivers on the host )
- import the cluster (k3s/k3d) when you want to use CPU cycles 
  ( k3s/k3d is very fast and very memory efficient )

- Create a project called vqbit and in that project create a namespace. eg: thereminq
- Deploy the Elasticsearch and Kibana APP in that namespace from the HELM repository
- Wait for both APPS to go Green, so that they are synced. 

# Be sure to add storage when using docker otherwise ES will stay red, missing storage.
- Add ThereminQ from the APP menu in the same namespace and wait for it to initialize.
- After a minute or 2 the CPU/GPUs load should increase to about the amount of CPUs/GPU avaliable in the system.

# When you have several GPUs or a large amount of CPU cores in the system you can add another workload via the (+) sign.
# Go to the Kibana UI by exposing port 5601 to the outside host in the Rancher UI or create an Ingress
- Select the filebeat stream for data output of ThereminQ
- You can also select the metricbeat for additional performance information

# When you want direct access to the ThereminQ container instance
- via the rancher shell UI option or on command line through the kubectl command
- when using docker on command line through the docker -i command

# when you need an XFCE desktop use cudacluster container instance ( included in HELM App stack )
- expose port 6080 to the host or create an Ingress and login with a webbrowser into NoVNC with password 00000000

Note: using cudacluster image won't increase the storage footprint as this is the base image of qracknet
Note: This gist is meant for Ubuntu 20.04+ for compatibility with the newest hardware and docker versions. 
      If you don't want a GPU you can deploy this with k3s/k3d and then should work with all version of ubuntu.
Note: default runs ( supremacy, tnn[-d], QFT entanglement & QFT cosmos ) will take days on CPU, hours on GPU.