Prerequisites

A cluster

Local

Minikube

https://minikube.sigs.k8s.io

# Create or start existing cluster
minikube start --cpus 5 --memory 10096

# Stop
minikube stop

# Delete
minikube delete

Docker Desktop

https://www.docker.com/products/docker-desktop

Cloud

GKE cluster

export cluster_name=mlstudio-cluster
export cluster_zone=us-central1-a

gcloud container clusters create $cluster_name \
    --machine-type=n1-standard-4 \
    --num-nodes 1 \
    --enable-autoscaling --min-nodes 0 --max-nodes 6 \
    --zone $cluster_zone
# TODO :: Save costs by using preemptible VMs. They offer the same machine types and options as regular compute instances and last for up to 24 hours 
#    --preemptible

# You can also create a gpu accelerated cluster or use your existing gpu accelretaed cluster
gcloud container clusters create $cluster_name \
#     --machine-type=n1-standard-4 \
#     --accelerator type=nvidia-tesla-t4,count=1 \
#     --num-nodes 1 \
#     --enable-autoscaling --min-nodes 0 --max-nodes 6 \
#     --zone $cluster_zone
# #    --preemptible

# # Install NVIDIA GPU device drivers
# kubectl apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/container-engine-accelerators/master/nvidia-driver-installer/cos/daemonset-preloaded.yaml

# Set kubectl locally
gcloud container clusters get-credentials $cluster_name --zone $cluster_zone

Kubectl

export KUBE_VERSION=v1.17.0 export KUBE_OS=darwin # windows # linux wget -q https://storage.googleapis.com/kubernetes-release/release/${KUBE_VERSION}/bin/${KUBE_OS}/amd64/kubectl -O /usr/local/bin/kubectl &&
chmod +x /usr/local/bin/kubectl

Helm 2.x

# TODO :: Install helm 2.x if not installed, https://helm.sh/docs/install/#installing-helm
export HELM_VERSION=v2.16.0 # v3.0.2
export HELM_OS=darwin # windows # linux
wget -q https://get.helm.sh/helm-${HELM_VERSION}-${HELM_OS}-amd64.tar.gz -O - | tar -xzO ${HELM_OS}-amd64/helm > /usr/local/bin/helm && \
chmod +x /usr/local/bin/helm

# Setup Tiller Service account
cat <<EOF | kubectl apply -f -
apiVersion: v1
kind: ServiceAccount
metadata:
  name: tiller
  namespace: kube-system
---
apiVersion: rbac.authorization.k8s.io/v1beta1
kind: ClusterRoleBinding
metadata:
  name: tiller
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: cluster-admin
subjects:
  - kind: ServiceAccount
    name: tiller
    namespace: kube-system
EOF

helm init --service-account=tiller
helm repo update

# Ensure that tiller is secure from access inside the cluster
kubectl patch deployment tiller-deploy --namespace=kube-system --type=json --patch='[{"op": "add", "path": "/spec/template/spec/containers/0/command", "value": ["/tiller", "--listen=localhost:44134"]}]'

# Verify helm and tiller were installed properly, by checking the client and server versions
helm version --short

# TODO :: Wait for tiller to get ready.

Istio

curl -L https://git.io/getLatestIstio | sh -

cd istio-*/

# Install the istio-init chart to bootstrap all the Istio’s CRDs:
helm install install/kubernetes/helm/istio-init --name istio-init --namespace istio-system

# Wait for all Istio CRDs to be created:
kubectl -n istio-system wait --for=condition=complete job --all

# Select a configuration profile and then install the istio chart corresponding to your chosen profile. The default profile is recommended for production deployments
helm install install/kubernetes/helm/istio --name istio --namespace istio-system

# Allow automatic Istio sidecar injector on all containers in default namespace
kubectl label namespace default istio-injection=enabled

# Verify auto injection worked
kubectl get namespace -L istio-injection

# Back
cd ../

# Remove istio directory
rm -R istio-*/

Install

helm repo add mlstudio https://ml-studio-app.github.io/helm-chart/
helm repo update

# If you are trying this on your local machine 
helm install mlstudio/mlstudio --set installLocally=true
# Otherwise
helm install mlstudio/mlstudio

Find the IP to access the app

kubectl get svc istio-ingressgateway -n istio-system

Log in as an admin user

  USER: admin
  PASSWORD: password

Cleanup

# Delete ML Studio release
helm del --purge mlstudio

# Delete the GKE cluster
gcloud container clusters delete $cluster_name --zone $cluster_zone