Application Delivery

The best application is a running application. For our we worked on creating some amazing REST API so far but its no use to anyone if we don't have it accessible to others.

Kubernetes

We can run the docker-compose in a virtual machine and expose it onto the internet that way. But these days for many software teams Kubernetes is the defacto place to deploy containers. Kubernetes is an open-source tool for managing container workloads. It has some awesome properties like self healing and auto-scaling. Its also the center of many cloud native technologies that makes the application delivery experience even better.

For this example we will deploy to a GKE Standard cluster. To enable our cloud-sql-proxy to connect to our database we will have to set up some IAM policies as shown here

git clone origin/containers-02

If you are not already in the folder

cd rest-api-container

kubectl apply -f kubernetes/

After a while we should see that our application is deployed

kubectl get deploy

NAME      READY   UP-TO-DATE   AVAILABLE   AGE
restapi   1/1     1            1           64m

We can get the external IP from our svc

kubectl get svc

NAME         TYPE           CLUSTER-IP    EXTERNAL-IP    PORT(S)        AGE
kubernetes   ClusterIP      10.52.0.1     <none>         443/TCP        5d8h
restapi      LoadBalancer   10.52.7.179   <SOME_IP>	     80:30981/TCP   69m

We can access our application at the IP

curl -s http://<SOME_IP>/api/v1/users/1 | jq

We should get

{
  "id": 1,
  "name": "Travis Miller",
  "email": "michellebrooks@williams.net",
  "company": "Russell-Rowe",
  "interests": [
    "Jenkins",
    "Jupyter Notebook",
    "Swift",
    "Scrum",
    "Spark",
    "Kubeflow",
    "GraphQL",
    "CSS",
    "Java",
    "Go",
    "Openshift",
    "C++",
    "Data Encryption"
  ]
}