Birthday-server

Google Container Engine based app

The application

• It's a web application to wish happy birthday or how many days left to your birthday
• Takes inputs from the URL query form and return the result or error message in JSON format
• Please seed the test data in tests directory.
• in the Makefile there's a rudimentary example of spawning all of the service and connecting them, plus sending a test request

Environment

• the Go compiler is installed and the go command is on the path

• make and anything else needed by the Makefile is installed and is on the path

• docker is on the path and the user running make is in the docker group

• a gcloud login was already performed so docker push just works

• kubectl is on the path

• a Kubernetes cluster is set up and the the default kubectl context points to that cluster. We can create the cluster using below commands

  gcloud config set project my-project-id
  gcloud config set compute/zone us-central1-f
  gcloud container clusters create example --zone us-central1-a \
   --num-nodes 2 --enable-autoscaling --min-nodes 1 --max-nodes 3 \
   --node-locations us-central1-a,us-central1-b,us-central1-f
  gcloud container clusters get-credentials --cluster example
  

Building the Docker Image

• Do a git init in the code and set respective user and email address

git config --global user.email "[EMAIL_ADDRESS]"
git config --global user.name "[USERNAME]"

• Make initial commits to the source code repository
• Create a repository to host your code

 gcloud source repos create sample-app
 git config credential.helper gcloud.sh

• Add your newly created repository as remote

export PROJECT=$(gcloud info --format='value(config.project)')
git remote add origin https://source.developers.google.com/p/$PROJECT/r/example

• Push your code to the new repository's master branch

git push origin master

• Create a Tag for the code and push

git tag v1.0.0
git push --tags

Configuring the build triggers :

• In GCP console click Build Triggers in the container registry seciotn
• Select Cloud Source Repository and click Continue
• Select example repository and select cloudbuild.yaml location: gcp/cloudbuild.yaml.
• Select Trigger type as Tag and in the Tag regex provide v.*

Build and Test

• Build images using make docker-images DOCKER_REPO=gcr.io/birthday-server
• Create and test cluster using make kubernetes-test
• Clean up after testing using make kubernetes-clean

Deployment

• Do the necessary code changes , tag them and push to the repo. Once pushed we can check the auto build in the build history
• We have three ways to do rolling update.

  • Using set image

    $ kubectl set image deployment <deployment> <container>=<image> --record
    # example
    git:mastkubectl set image deployment web-controller  web=example:v1.0.0 --record
    deployment "web-controller" image updated
    
    

• Modify the kubernetes manifest and update the docker image version.

  # format
  $ kubectl replace -f <yaml> --record
  # example
  $ kubectl replace -f gcp/birthday-server.yml --record
  
  

• Using Edit

  # format
  $ kubectl edit deployment <deployment> --record
  git:master *⚡ kubectl edit deployment web-controller --record
  
  

• Check Status/Pause/Resume Deployments

  kubectl rollout status deployment web-controller # Check Status
  kubectl rollout pause deployment web-controller # Pause Rolling Update
  kubectl rollout resume deployment web-controller # Resume Rolling Update
  

• To create and list the firewall rule that allows traffic into the cluster I use a gcloud compute command

gcloud compute firewall-rules create --allow=tcp:80 --target-tags=gke-example-default-pool-12d321682 example #creating the firewall rule
gcloud compute forwarding-rules list

Autoscaling

• When you use kubectl autoscale, you specify a maximum and minimum number of replicas for your application, as well as a CPU utilization target.
• For example, to set the maximum number of replicas to 4 and the minimum to 3, with a CPU utilization target of 50% utilization, run the following command:

git:master *⚡ kubectl  get deployments  # getting deployments list
NAME             DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
web-controller   2         2        2            2           1h

git:master *%=⚡ kubectl autoscale deployment web-controller --max 4 --min 3 --cpu-percent 50
deployment "web-controller" autoscaled

Rollback

to previous revision

 $ kubectl rollout undo deployment web-controller
 git:master *⚡ kubectl rollout undo deployment web-controller
 deployment "web-controller" rolled back

to specific revision

  $ kubectl rollout undo deployment web-controller --to-revision=<revision>
  # exmaple
  $ kubectl rollout undo deployment web-controller --to-revision=v1.0.1