Deploy Python container on Kubernetes cluster on GCP

 Source

https://www.youtube.com/watch?v=GKuk-TBmNcI

• Purpose

The goal of this blog is to show how a docker container can be deployed on a Kubernetes cluster. The motivation of deploying containers is as follows

Once you build a container, you may want to scale it up and down according to the changing demand. TO do that, Kubernetes has a Loadbalancer which distributes http requests over a set of Nodes.

Kubernetes architecture

A Kubernetes cluster is formed of nodes. A node can be either worker or master node

Within each Node there are multiple pods. A container is deployed on a pod and the load balancer distributes the requests across the nodes and pods

Build a container

• On cloud shell create a folder and place the following files

• Thats how the Dockerfile looks like

• Build the container using gcloud command

• Create a Kubernetes cluster using this command- Change the number of nodes as you wish

• You can change the number of nodes by modifying the deployment.yaml file as follows

• Then apply this as follows

• To modify the ports of the loadbalancer edit the services.yaml file as follows

• To get the IP to call type the following command

• Just place this IP in the browser and the container will run