weather-api-task

task description:

create a deployment pipeline, packing node.js application in Docker image, and deployed using HELM in k8s cluster as mentioned in https://github.com/culture-trip/weather-api

Clone the repo and run:

source init.sh

By running source init.sh file the following actions are taken:

1. check if virtualenv exists, if not - create it
2. install all dependencies required
3. trigger the main_run.py script

main_run.py script

1. import dependencies, check running dir and create Docker client
2. if weather-api repo does not exist, clone it
3. copy Docker files content into the weather-api repo: Dockerfile - copy app files to docker, run npm install, set entrypoint npm start docker-compose - set image name, version, and port
4. interactively create a .env file with API_KEY, PORT and version.
5. login to docker hub registry and push the image

K8s deployment

for testing I created a free tier k8s cluster with 2 nodes 1 master using this guide: https://aws.plainenglish.io/how-to-setup-a-kubernetes-cluster-with-aws-free-tier-and-a-free-domain-514d010ae456

created a hosted-zone, security group, bucket for the cluster state and evantually spin it up using this command:

kops create cluster --name=kube.dynst.ml --state=s3://yishai-bucket --zones=eu-west-1a --node-count=2 --node-size=t3.medium --master-size=t3.medium --dns-zone=kube.dynst.ml --yes

5. validate k8s context

6. helm install using chart in weather-api-task repo pass --set selected version and port pass the api_key in encrypted secret.yaml Api key and port are saved as environment variables in container

   equivilant to the following commands

   $ helm install --set service.port=8080 --set image.tag=v3  --create-namespace  -n weather -f values.yaml weather .
   $ kubectl -n weather port-forward weather-chart-56db6d6b74-bn4lx 3000:3000

   to uninstall:

   $ helm uninstall -n weather weather 

   Destroy free cluster:

   $ kops delete cluster --name=kube.dynst.ml --state=s3://yishai-bucket --yes

For additional Monitoring: open another tab and run:

$ kubectl apply -f "https://cloud.weave.works/k8s/scope.yaml?k8s-version=$(kubectl version | base64 | tr -d '\n')"

$ kubectl port-forward -n weave "$(kubectl get -n weave pod --selector=weave-scope-component=app -o jsonpath='{.items..metadata.name}')" 4040

this will give you an overview of the containers used resources, uptime, networking and footprint etcetera

For scaling we can increase the replica count and implement load balancing to achieve High availability using Nginx Controllers.