demo-script.sh

# script for minikube demo

# install xhyve driver - less resource overhead for the vm
# https://github.com/kubernetes/minikube/blob/master/DRIVERS.md#xhyve-driver
brew install docker-machine-driver-xhyve

# install minikube
# https://github.com/kubernetes/minikube/releases

# create a cluster. driver is important!

minikube start --vm-driver=xhyve

# install kubectl

brew install kubectl

# check that cluster is working
kubectl config set-context minikube


kubectl get componentstatus

kubectl cluster-info

# deploy an application and a service
# describe a namespace, deployment, and a service

kubectl create -f monitoring-namespace.yaml
kubectl create -f prometheus-config.yaml
kubectl create -f prometheus-deployment.yaml
kubectl create -f prometheus-service.yaml

# view via CLI:
kubectl get services --namespace=monitoring
kubectl get deployments --namespace=monitoring

# show the dashboard
minikube dashboard

# show node ports under services for prometheus. exposes

# show prometheus service

minikube service --namespace=monitoring prometheus

# click around
# show /targets

# show graph and query of container_memory_usage_bytes{kubernetes_namespace="monitoring"}

# deploy grafana
kubectl create -f grafana-deployment.yaml
kubectl create -f grafana-service.yaml

# show grafana
minikube service --namespace=monitoring grafana

# add datasource. make sure type is prometheus http://prometheus:8080
# describe kubernetes DNS

# create a graph:
#  container_memory_usage_bytes{kubernetes_namespace="monitoring"}
# {{kubernetes_pod_name}

# lets add node metrics
# deploy node exporter. explain daemonser
kubectl create -f node-exporter-daemonset.yml

# show new target in prometheus. explain it autodiscovering the pods

# now create a graph

# node_load1
# {{kubernetes_pod_node_name}}