yuweisung / cks_kubeadm

Certified Kubernetes Security Specialist Stuffs

---

My terraform, ansible, and kubeadm scripts for CKS exam (K8S v1.20.0)

Prerequsites (Mac):

• terraform (brew install terraform)
• ansible (brew install ansible)
• git (brew install git)
• kubectl (brew install kubectl)
• GCP Project with a Google cloud managed dns (publilc_zone)

  wget https://dl.google.com/dl/cloudsdk/channels/rapid/downloads/google-cloud-sdk-324.0.0-darwin-x86_64.tar.gz
  tar zxvf google-cloud-sdk-324.0.0-darwin-x86_64.tar.gz
  cd google-cloud-sdk
  ./install.sh
  gcloud init
  gcloud auth login
  gcloud iam service-accounts create <service-account-name> --project <gcp-project-id>
  gcloud iam service-accounts keys create ~/.ssh/<service-account-name>.json --iam-account <service-account-name>@dataengineeringdemos.iam.gserviceaccount.
com
  gcloud projects add-iam-policy-binding <gcp-project-id> --member='serviceAccount:<service-account-name>@dataengineeringdemos.iam.gserviceaccount.com' --role=roles/editor
  

Installation:

1. Clone this repo

git clone https://github.com/ssung-yugabyte/cks-kubeadm.git
cd cks-kubeadm 

2. Prepare variables.tf

mv variables.tf.example variables.tf

3. Review and modify the variables.tf

cat > variables.tf <<EOF
variable "gcp_profile" {
  description = "GCP Configuration"
  type = map
  default = {
    project = "XXXXXXX"                  <== Your GCP Project
    region = "us-central1"               <== Your Preferred GCP Region
    zone = "us-central1-c"               <== Your Preferred GCP Zone
    credentials = "~/.ssh/XXXXXXX.json"  <== Your GCP Service Account Credential
  }
  sensitive = true
}

variable "gce_vm" {
  description = "GCE Instance Configuration"
  type = map
  default = {
    instance_type = "n2-standard-2"
    os_project = "ubuntu-os-cloud"
    os_family = "ubuntu-2004-lts"
    boot_disk_size = 200
    ssh_user = "XXXX"                    <== Your local user_name
    ssh_pub = "~/.ssh/id_rsa.pub"        <== Your ssh public key
  }
}

variable "master_count" {
  description = "K8s Master instances"
  type = number
  default = 1
}

variable "worker_count" {
  description = "K8s worker instances"
  type          = number
  default       = 3
}

variable "k8s_version" {
  type		= string
  default	= "1.20.0"
}

variable "gcp_private_dns_zone" {
  description = "Google Managed DNS zone - private zone name"
  type = map
  default = {
    zone_name = "XXXXX-private"            <== Your Google Cloud Managed Zone (private)
    dns_name = "cks.yugabyte.lab."           <== Your Google Cloud Managed Zone DNS Name (private)
  } 
}

variable "gcp_public_dns_zone" {
  description = "Google Managed DNS zone - public (preconfig required). If no public zone, external api uses public IP"
  type = map
  default = {
    enabled = false   <== OPTIOINAL: google domain resource requires 24hrs to be affected. You need to pre config a google domain in advance, then create a public DNS zone.
    zone_name = "ysung-public-ats-zone"
  }

}

variable "vpc_subnet_cidr" {
  description  = "VPC custom subnet CIDR"
  type          = string
  default       = "192.168.20.0/24"
}

variable "k8s_pod_cidr" {
  description = "K8s pod subnet CIDR"
  type          = string
  default       = "10.244.0.0/16"
}

variable "k8s_service_cidr" {
  description = "K8s service CIDR"
  type		= string
  default	= "10.96.0.0/12"
}
EOF

4. Init Terraform plugins

terraform init

5. Reivew Terraform plan

terraform plan

6. Apply Terraform plan

terraform apply --auto-approve

7. Check the k8s cluster

kubectl get nodes

8. Kubectl away...

cd kubectl/deployments

Reset:

1. Destroy the terraform plan K8S will use gce to create legacy gce disks (pvc) /firewall rules/health-check/load-balancer/target-pool. When destory, those will be ignored as they are defined outside of terraform. In order to clean up those google cloud resources, you will need gcloud cmd.

terraform destroy --auto-approve

ToDo:

• [] Instance groups
• [] Kubeadm upgrade
• Control plane HA
  • GCP LoadBalancer: Layer4 TCP Load Balancer
  • GCP SSL health check (haproxy + keepalived)
  • GCP target pool
  • kubeadm join --control-plane