Greetings Application

This is a simple demo application demonstrating a Java application's journey from development to production.

It comprises of a Spring Boot application that exposes a single API endpoint that returns a greeting as a string. Cliche. Right? That needs to find its way to Kubernetes. Not so cliche for a Greetings app. Huh?

Packaging to jar

$ ./gradlew build

Build and skip tests...

$ ./gradlew build -x test

Test application....

$ curl localhost:8080

Packaging to Docker image using gradle plugin jib

$ ./gradlew jibDockerBuild

List docker images....

$ docker images

Running a container from the docker image....

$ docker run --name greetings-app -p 8080:8080 -d greetings:1.0.0-RELEASE

or with an environment variable

$ docker run --name greetings-app -p 8080:8080 --env GREETINGS_SALUTATION=Niaje! -d greetings:1.0.0-RELEASE

Test application....

$ curl localhost:8080

To developer Kubernetes instance (minikube)

Minikube hack to build image to kubernetes

$ eval $(minikube docker-env)
$ ./gradlew jibDockerBuild

Deploy app to kubernetes on developer machine

$ kubectl apply -f manifest/vanilla/greetings-application-v1.yaml

Get deployments

$ kubectl get deployments -n ns-workshop

NAME READY UP-TO-DATE AVAILABLE AGE

greetings 1/1 1 1 2m26s

Get pods

$ kubectl get pods -n ns-workshop

NAME READY STATUS RESTARTS AGE

greetings-6b7c4bbf77-jnbvx 1/1 Running 0 2m36s

Get services

$ kubectl get services -n ns-workshop

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE

greetings-app NodePort 10.104.188.88 8080:30245/TCP 19s

Test deployed app

$ minikube service -n ns-workshop greetings-app

Switch to production kubernetes context

$ kubectl config get-contexts

CURRENT NAME CLUSTER AUTHINFO NAMESPACE

•     gke_my-lab-...   gke_my-lab-...   gke_my-lab-...
  
      minikube.istio   minikube.istio   minikube.istio   
  

$ kubectl config use-context <context-name>

Helm

What is helm?

Test Deploy

$ helm install ./manifest/helm-charts/greetings --debug --dry-run

Deploy

$ helm install ./manifest/helm-charts/greetings

NAME: yellow-ferret LAST DEPLOYED: Wed Jul 24 23:04:14 2019 NAMESPACE: default STATUS: DEPLOYED

RESOURCES: ==> v1/Service NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE yellow-ferret-greetings ClusterIP 10.36.12.175 8080/TCP 1s

==> v1/Deployment NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE yellow-ferret-greetings 1 1 1 0 1s

==> v1/Pod(related) NAME READY STATUS RESTARTS AGE yellow-ferret-greetings-547b4d5bd9-ns828 0/1 ContainerCreating 0 1s

Delete helm install

$ helm delete yellow-ferret

Istio

What is istio?

Why istio?

Out of the box features:

• Request routing
• Fault injection
• Traffic shifting
• Querying metrics
• Visualizing metrics
• Collecting logs
• Rate limiting
• Ingress gateways
• Accessing external services
• Visualizing your mesh

Querying metrics with Prometheus

Open Prometheus dashboard

$ kubectl -n istio-system port-forward $(kubectl -n istio-system get pod -l app=prometheus -o jsonpath='{.items[0].metadata.name}') 9090:9090 &

Go to http://localhost:9090

Querying metrics from Prometheus

Execute a Prometheus query.

In the “Expression” input box at the top of the web page, enter the text: istio_requests_total. Then, click the Execute button

Total count of all requests to greetings-app service

$ istio_requests_total{destination_service="greetings-app.ns-workshop.svc.cluster.local"}

Total count of all requests to version 4 of greetings-app service

$ istio_requests_total{destination_service="greetings-app.ns-workshop.svc.cluster.local", destination_version="v3"}

Visualize metrics with Grafana

Open Grafana dashboard

$ kubectl -n istio-system port-forward $(kubectl -n istio-system get pod -l app=grafana -o jsonpath='{.items[0].metadata.name}') 3000:3000 &

Go to http://localhost:3000/dashboard/db/istio-mesh-dashboard

Go to http://localhost:3000/dashboard/db/istio-service-dashboard

Go to http://localhost:3000/dashboard/db/istio-workload-dashboard

Linkerd

What is linkerd?

Why linkerd?

Out of the box features:

• HTTP, HTTP/2, and gRPC Proxying
• TCP Proxying and Protocol Detection
• Retries and Timeouts
• Automatic mTLS
• Automatic Proxy Injection
• Service Profiles
• Ingress
• Telemetry and Monitoring
• Load Balancing
• Dashboard and Grafana

Open Linkerd dashboard

$ linkerd dashboard

Injecting proxy to nairobijvm namepsace if auto-injection is disabled

This command retrieves all of the deployments running in the nairobijvm namespace, runs the set of Kubernetes resources through inject, and finally reapplies it to the cluster.

$ kubectl get -n ns-workshop deploy -o yaml \
  | linkerd inject - \
  | kubectl apply -f -

Test deployed app

$ minikube service -n ns-workshop greetings-app
$ while true; do curl http://192.168.39.9:31081/; done

Cleanup

To clean up the resources you created in your cluster:

Delete greetings service

$ kubectl delete service greetings-app -n ns-workshop

Delete greetings deployment

$ kubectl delete deployment greetings -n ns-workshop

Packaging to Docker image and Push to a registry using gradle plugin jib

$ ./gradlew jib

Pull and use registry:v2 to quickly setup a local docker image registry.