Kubernetes Lab: Services & Service Discovery

Sample Code – https://github.com/devopsschool-demo-labs-projects/kubernetes-lab-demo/tree/main/Services-ServiceDiscovery

	ssh aen@c1-cp1
	cd ~/content/course/03/demos/

	#1 - Exposing and accessing applications with Services on our local cluster
	#ClusterIP

	#Imperative, create a deployment with one replica
	kubectl create deployment hello-world-clusterip \
	--image=gcr.io/google-samples/hello-app:1.0


	#When creating a service, you can define a type, if you don't define a type, the default is ClusterIP
	kubectl expose deployment hello-world-clusterip \
	--port=80 --target-port=8080 --type ClusterIP


	#Get a list of services, examine the Type, CLUSTER-IP and Port
	kubectl get service


	#Get the Service's ClusterIP and store that for reuse.
	SERVICEIP=$(kubectl get service hello-world-clusterip -o jsonpath='{ .spec.clusterIP }')
	echo $SERVICEIP


	#Access the service inside the cluster
	curl http://$SERVICEIP


	#Get a listing of the endpoints for a service, we see the one pod endpoint registered.
	kubectl get endpoints hello-world-clusterip
	kubectl get pods -o wide


	#Access the pod's application directly on the Target Port on the Pod, not the service's Port, useful for troubleshooting.
	#Right now there's only one Pod and its one Endpoint
	kubectl get endpoints hello-world-clusterip
	PODIP=$(kubectl get endpoints hello-world-clusterip -o jsonpath='{ .subsets[].addresses[].ip }')
	echo $PODIP
	curl http://$PODIP:8080


	#Scale the deployment, new endpoints are registered automatically
	kubectl scale deployment hello-world-clusterip --replicas=6
	kubectl get endpoints hello-world-clusterip


	#Access the service inside the cluster, this time our requests will be load balanced...whooo!
	curl http://$SERVICEIP


	#The Service's Endpoints match the labels, let's look at the service and it's selector and the pods labels.
	kubectl describe service hello-world-clusterip
	kubectl get pods --show-labels


	#Clean up these resources for the next demo
	kubectl delete deployments hello-world-clusterip
	kubectl delete service hello-world-clusterip




	#2 - Creating a NodePort Service
	#Imperative, create a deployment with one replica
	kubectl create deployment hello-world-nodeport \
	--image=gcr.io/google-samples/hello-app:1.0


	#When creating a service, you can define a type, if you don't define a type, the default is ClusterIP
	kubectl expose deployment hello-world-nodeport \
	--port=80 --target-port=8080 --type NodePort


	#Let's check out the services details, there's the Node Port after the : in the Ports column. It's also got a ClusterIP and Port
	#This NodePort service is available on that NodePort on each node in the cluster
	kubectl get service


	CLUSTERIP=$(kubectl get service hello-world-nodeport -o jsonpath='{ .spec.clusterIP }')
	PORT=$(kubectl get service hello-world-nodeport -o jsonpath='{ .spec.ports[].port }')
	NODEPORT=$(kubectl get service hello-world-nodeport -o jsonpath='{ .spec.ports[].nodePort }')

	#Let's access the services on the Node Port...we can do that on each node in the cluster and
	#from outside the cluster...regardless of where the pod actually is

	#We have only one pod online supporting our service
	kubectl get pods -o wide


	#And we can access the service by hitting the node port on ANY node in the cluster on the Node's Real IP or Name.
	#This will forward to the cluster IP and get load balanced to a Pod. Even if there is only one Pod.
	curl http://c1-cp1:$NODEPORT
	curl http://c1-node1:$NODEPORT
	curl http://c1-node2:$NODEPORT
	curl http://c1-node3:$NODEPORT


	#And a Node port service is also listening on a Cluster IP, in fact the Node Port traffic is routed to the ClusterIP
	echo $CLUSTERIP:$PORT
	curl http://$CLUSTERIP:$PORT


	#Let's delete that service
	kubectl delete service hello-world-nodeport
	kubectl delete deployment hello-world-nodeport




	#3 - Creating LoadBalancer Services in Azure or any cloud
	#Switch contexts into AKS, we created this cluster together in 'Kubernetes Installation and Configuration Fundamentals'
	#I've added a script to create a GKE and AKS cluster this course's downloads.
	kubectl config use-context 'CSCluster'


	#Let's create a deployment
	kubectl create deployment hello-world-loadbalancer \
	--image=gcr.io/google-samples/hello-app:1.0


	#When creating a service, you can define a type, if you don't define a type, the default is ClusterIP
	kubectl expose deployment hello-world-loadbalancer \
	--port=80 --target-port=8080 --type LoadBalancer


	#Can take a minute for the load balancer to provision and get an public IP, you'll see EXTERNAL-IP as <pending>
	kubectl get service


	LOADBALANCERIP=$(kubectl get service hello-world-loadbalancer -o jsonpath='{ .status.loadBalancer.ingress[].ip }')
	curl http://$LOADBALANCERIP:$PORT


	#The loadbalancer, which is 'outside' your cluster, sends traffic to the NodePort Service which sends it to the ClusterIP to get to your pods!
	#Your cloud load balancer will have health probes checking the health of the node port service on the real node IPs.
	#This isn't the health of our application, that still needs to be configured via readiness/liveness probes and maintained by your Deployment configuration
	kubectl get service hello-world-loadbalancer



	#Clean up the resources from this demo
	kubectl delete deployment hello-world-loadbalancer
	kubectl delete service hello-world-loadbalancer


	#Let's switch back to our local cluster
	kubectl config use-context kubernetes-admin@kubernetes



	#Declarative examples
	kubectl config use-context kubernetes-admin@kubernetes
	kubectl apply -f service-hello-world-clusterip.yaml
	kubectl get service


	#Creating a NodePort with a predefined port, first with a port outside of the NodePort range then a corrected one.
	kubectl apply -f service-hello-world-nodeport-incorrect.yaml
	kubectl apply -f service-hello-world-nodeport.yaml
	kubectl get service


	#Switch contexts to Azure to create a cloud load balancer
	kubectl config use-context 'CSCluster'
	kubectl apply -f service-hello-world-loadbalancer.yaml
	kubectl get service


	#Clean up these resources
	kubectl delete -f service-hello-world-loadbalancer.yaml
	kubectl config use-context kubernetes-admin@kubernetes
	kubectl delete -f service-hello-world-nodeport.yaml
	kubectl delete -f service-hello-world-clusterip.yaml

view raw 1-Services.sh hosted with

by GitHub

	ssh aen@c1-cp1
	cd ~/content/course/03/demos/


	#Service Discovery
	#Cluster DNS

	#Let's create a deployment in the default namespace
	kubectl create deployment hello-world-clusterip \
	--image=gcr.io/google-samples/hello-app:1.0


	#Let's create a deployment in the default namespace
	kubectl expose deployment hello-world-clusterip \
	--port=80 --target-port=8080 --type ClusterIP


	#We can use nslookup or dig to investigate the DNS record, it's CNAME @10.96.0.10 is the cluser IP of our DNS Server
	kubectl get service kube-dns --namespace kube-system


	#Each service gets a DNS record, we can use this in our applications to find services by name.
	#The A record is in the form <servicename>.<namespace>.svc.<clusterdomain>
	nslookup hello-world-clusterip.default.svc.cluster.local 10.96.0.10
	kubectl get service hello-world-clusterip


	#Create a namespace, deployment with one replica and a service
	kubectl create namespace ns1


	#Let's create a deployment with the same name as the first one, but in our new namespace
	kubectl create deployment hello-world-clusterip --namespace ns1 \
	--image=gcr.io/google-samples/hello-app:1.0


	kubectl expose deployment hello-world-clusterip --namespace ns1 \
	--port=80 --target-port=8080 --type ClusterIP


	#Let's check the DNS record for the service in the namespace, ns1. See how ns1 is in the DNS record?
	#<servicename>.<namespace>.svc.<clusterdomain>
	nslookup hello-world-clusterip.ns1.svc.cluster.local 10.96.0.10


	#Our service in the default namespace is still there, these are completely unique services.
	nslookup hello-world-clusterip.default.svc.cluster.local 10.96.0.10


	#Get the environment variables for the pod in our default namespace
	#More details about the lifecycle of variables in "Configuring and Managing Kubernetes Storage and Scheduling"
	#Only the kubernetes service is available? Why? I created the deployment THEN I created the service
	PODNAME=$(kubectl get pods -o jsonpath='{ .items[].metadata.name }')
	echo $PODNAME
	kubectl exec -it $PODNAME -- env \| sort


	#Environment variables are only created at pod start up, so let's delete the pod
	kubectl delete pod $PODNAME


	#And check the enviroment variables again...
	PODNAME=$(kubectl get pods -o jsonpath='{ .items[].metadata.name }')
	echo $PODNAME
	kubectl exec -it $PODNAME -- env \| sort


	#ExternalName
	kubectl apply -f service-externalname.yaml


	#The record is in the form <servicename>.<namespace>.<clusterdomain>. You may get an error that says ** server can't find hello-world.api.example.com: NXDOMAIN this is ok.
	nslookup hello-world-api.default.svc.cluster.local 10.96.0.10




	#Let's clean up our resources in this demo
	kubectl delete service hello-world-api
	kubectl delete service hello-world-clusterip
	kubectl delete service hello-world-clusterip --namespace ns1
	kubectl delete deployment hello-world-clusterip
	kubectl delete deployment hello-world-clusterip --namespace ns1
	kubectl delete namespace ns1

view raw 2-ServiceDiscovery.sh hosted with

by GitHub

Rajesh Kumar

I’m a DevOps/SRE/DevSecOps/Cloud Expert passionate about sharing knowledge and experiences. I am working at Cotocus. I blog tech insights at DevOps School, travel stories at Holiday Landmark, stock market tips at Stocks Mantra, health and fitness guidance at My Medic Plus, product reviews at I reviewed , and SEO strategies at Wizbrand.

Please find my social handles as below;

Rajesh Kumar Personal Website
Rajesh Kumar at YOUTUBE
Rajesh Kumar at INSTAGRAM
Rajesh Kumar at X
Rajesh Kumar at FACEBOOK
Rajesh Kumar at LINKEDIN
Rajesh Kumar at PINTEREST
Rajesh Kumar at QUORA
Rajesh Kumar at WIZBRAND