README.md

Deployment with kubernetes (locally)

Deployment to kubernetes cluster

• The 2 config files deployment.yaml and service.yaml can be used for deploying to kubernetes cluster

Instructions to setup kubernetes cluster (locally) and deploying the docker image with the FastAPI ML application to the kubernetes cluster

• Install kubectl and kind
• Setup a kind cluster with the following command

kind create cluster

• To check the cluster info, run the following commands

kubectl cluster-info --context kind-kind

• To check services, pods, deployments; run the following commands

kubectl get service
kubectl get pod
kubectl get deployment

• Load the docker image for the FastAPI ML application to the kubernetes cluster

kind load docker-image fastapi_water_potability

• Setup the deployment, the config for deployment is available in deployment.yaml

kubectl apply -f kubernetes_deployment/deployment.yaml

• Run port forwarding and check the deployment

kubectl get pod
kubectl port-forward <pod_name> 5000:5000
python3 test_post_request.py

• Setup the service, the config for service is available in service.yaml

kubectl apply -f kubernetes_deployment/service.yaml

• Run port forwarding and check the service

kubectl port-forward service/fastapi-water-potability 5000:80
python3 test_post_request.py

• NOTE: In the port forwarding command, the first port is port on the host, the second port is the port of the deployment pod or the service which can be found in the respective yaml files