k8s

Here, we will go through the steps of configuring our k8s environment, playing around with the necessary commands and such, then distributing some manifests to our cluster.

Steps

• install local dependencies
• cloud setup
• enable k8s engine
• create service account
• git and configure
• test cluster connectivity

install local dependencies

You need kubectl (cube-control) to manage and operate your k8s cluster from a local machine.

• Mac

brew install kubernetes-cli

You will need helm too

• Mac

brew install kubernetes-helm

You should have gcloud on your system as well. This way you can manipulate gcp from the local cli.

• Mac

brew cask install google-cloud-sdk

• Linux/Ubuntu
• Windows

enable k8s engine

• Navigate to K8s Engine in GCP Console
• Enable the K8s API

create service account

• Navigate to the IAM section
• Navigate to Service Accounts
• Create a new service account
• Give it admin credentials
• Generate a .json key_file
• Store the .json key_file locally somewhere you can access later

git and configure

• clone the repo
• if you don't have terraform, please download it or build it from source
• change into the k8s dir

cd k8s

• read the comments and make the necessary changes
• initialize

terraform init

• plan

terraform plan

• apply

terraform apply

• if you need references:
  • k8s cluster on gcp
  • google provider

test cluster connectivity

Assuming all went well and we all have k8s clusters on GCP, we should be able to navigate to our cluster from the k8s engine tab from the GCP console. Click on the connect button and copy/paste the gcloud command into your local terminal. Now run

kubectl get nodes

... I bet you see your cluster. If you don't, just let me know.

If all is well then we can move on to the k8sWithIstio directory and dig into the tool.