Deploy a Kubernetes cluster for free, using k3s and Oracle Cloud always free resources.

Lorenzo Garuti · February 22, 2022

k3s-logo

Deploy a Kubernetes cluster for free, using K3s and Oracle always free resources.

Table of Contents

Note choose a region with enough ARM capacity

Important notes

  • This is tutorial shows only how to use terraform with the Oracle Cloud infrastructure and use only the always free resources. This examples are not for a production environment.
  • At the end of your trial period (30 days). All the paid resources deployed will be stopped/terminated
  • At the end of your trial period (30 days), if you have a running compute instance it will be stopped/hibernated

Requirements

To use this tutorial you will need:

  • an Oracle Cloud account. You can register here

Once you get the account, follow the Before you begin and 1. Prepare step in this document.

you need also:

  • Terraform - Terraform is an open-source infrastructure as code software tool that provides a consistent CLI workflow to manage hundreds of cloud services. Terraform codifies cloud APIs into declarative configuration files.
  • kubectl - The Kubernetes command-line tool (optional)
  • oci cli - Oracle command line interface (optional)

Example RSA key generation

To use terraform with the Oracle Cloud infrastructure you need to generate an RSA key. Generate the rsa key with:

openssl genrsa -out ~/.oci/<your_name>-oracle-cloud.pem 4096
chmod 600 ~/.oci/<your_name>-oracle-cloud.pem
openssl rsa -pubout -in ~/.oci/<your_name>-oracle-cloud.pem -out ~/.oci/<your_name>-oracle-cloud_public.pem

replace ** with your name or a string you prefer.

NOTE ~/.oci/-oracle-cloud_public.pem this string will be used on the *terraform.tfvars* used by the Oracle provider plugin, so please take note of this string.

Project setup

You can clone this repository and work in the example/ directory. You have to edit the main.tf file and you have to create the terraform.tfvars file. For more detail see Oracle provider setup and Pre flight checklist.

Or if you prefer you can create an new empty directory in your workspace and create this three files:

The main.tf file will look like:

variable "compartment_ocid" {

}

variable "tenancy_ocid" {

}

variable "user_ocid" {

}

variable "fingerprint" {

}

variable "private_key_path" {

}

variable "region" {
  default = "<change_me>"
}

module "k3s_cluster" {
  region              = var.region
  availability_domain = "<change_me>"
  compartment_ocid    = var.compartment_ocid
  my_public_ip_cidr   = "<change_me>"
  cluster_name        = "<change_me>"
  environment         = "staging"
  k3s_token           = "<change_me>"
  source              = "github.com/garutilorenzo/k3s-oci-cluster"
}

output "k3s_servers_ips" {
  value = module.k3s_cluster.k3s_servers_ips
}

output "k3s_workers_ips" {
  value = module.k3s_cluster.k3s_workers_ips
}

output "public_lb_ip" {
  value = module.k3s_cluster.public_lb_ip
}

For all the possible variables see Pre flight checklist

The provider.tf will look like:

provider "oci" {
  tenancy_ocid     = var.tenancy_ocid
  user_ocid        = var.user_ocid
  private_key_path = var.private_key_path
  fingerprint      = var.fingerprint
  region           = var.region
}

Now we can init terraform with:

terraform init

terraform init
Initializing modules...
Downloading git::https://github.com/garutilorenzo/k3s-oci-cluster.git for k3s_cluster...
- k3s_cluster in .terraform/modules/k3s_cluster

Initializing the backend...

Initializing provider plugins...
- Reusing previous version of hashicorp/oci from the dependency lock file
- Reusing previous version of hashicorp/template from the dependency lock file
- Using previously-installed hashicorp/template v2.2.0
- Using previously-installed hashicorp/oci v4.64.0

Terraform has been successfully initialized!

You may now begin working with Terraform. Try running "terraform plan" to see
any changes that are required for your infrastructure. All Terraform commands
should now work.

If you ever set or change modules or backend configuration for Terraform,
rerun this command to reinitialize your working directory. If you forget, other
commands will detect it and remind you to do so if necessary.

Generate sel signed SSL certificate for the public LB (L7)

NOTE If you already own a valid certificate skip this step and set the correct values for the variables: PATH_TO_PUBLIC_LB_CERT and PATH_TO_PUBLIC_LB_KEY

We need to generate the certificates (sel signed) for our public load balancer (Layer 7). To do this we need openssl, open a terminal and follow this step:

Generate the key:

openssl genrsa 2048 > privatekey.pem
Generating RSA private key, 2048 bit long modulus (2 primes)
.......+++++
...............+++++
e is 65537 (0x010001)

Generate the a new certificate request:

openssl req -new -key privatekey.pem -out csr.pem
You are about to be asked to enter information that will be incorporated
into your certificate request.
What you are about to enter is what is called a Distinguished Name or a DN.
There are quite a few fields but you can leave some blank
For some fields there will be a default value,
If you enter '.', the field will be left blank.
-----
Country Name (2 letter code) [AU]:IT
State or Province Name (full name) [Some-State]:Italy
Locality Name (eg, city) []:Brescia
Organization Name (eg, company) [Internet Widgits Pty Ltd]:GL Ltd
Organizational Unit Name (eg, section) []:IT
Common Name (e.g. server FQDN or YOUR name) []:testlb.domainexample.com
Email Address []:email@you.com

Please enter the following 'extra' attributes
to be sent with your certificate request
A challenge password []:
An optional company name []:

Generate the public CRT:

openssl x509 -req -days 365 -in csr.pem -signkey privatekey.pem -out public.crt
Signature ok
subject=C = IT, ST = Italy, L = Brescia, O = GL Ltd, OU = IT, CN = testlb.domainexample.com, emailAddress = email@you.com
Getting Private key

This is the final result:

ls

csr.pem  privatekey.pem  public.crt

Now set the variables:

  • PATH_TO_PUBLIC_LB_CERT: ~/full_path/public.crt
  • PATH_TO_PUBLIC_LB_KEY: ~/full_path/privatekey.pem

Oracle provider setup

This is an example of the terraform.tfvars file:

fingerprint      = "<rsa_key_fingerprint>"
private_key_path = "~/.oci/<your_name>-oracle-cloud_public.pem"
user_ocid        = "<user_ocid>"
tenancy_ocid     = "<tenency_ocid>"
compartment_ocid = "<compartment_ocid>"

To find your tenency_ocid in the Ocacle Cloud console go to: Governance and Administration > Tenency details, then copy the OCID.

To find you user_ocid in the Ocacle Cloud console go to User setting (click on the icon in the top right corner, then click on User settings), click your username and then copy the OCID

The compartment_ocid is the same as tenency_ocid.

The fingerprint is the fingerprint of your RSA key, you can find this vale under User setting > API Keys

Pre flight checklist

Once you have created the terraform.tfvars file edit the main.tf file (always in the example/ directory) and set the following variables:

Var Required Desc
region yes set the correct OCI region based on your needs
availability_domain yes Set the correct availability domain. See how to find the availability domain
compartment_ocid yes Set the correct compartment ocid. See how to find the compartment ocid
cluster_name yes the name of your K3s cluster. Default: k3s-cluster
k3s_token yes The token of your K3s cluster. How to generate a random token
my_public_ip_cidr yes your public ip in cidr format (Example: 195.102.xxx.xxx/32)
environment yes Current work environment (Example: staging/dev/prod). This value is used for tag all the deployed resources
PATH_TO_PUBLIC_LB_CERT yes Path to the public LB certificate. See how to generate the certificate
PATH_TO_PUBLIC_LB_KEY yes Path to the public LB key. See how to generate the key
compute_shape no Compute shape to use. Default VM.Standard.A1.Flex. NOTE Is mandatory to use this compute shape for provision 4 always free VMs
os_image_id no Image id to use. Default image: Canonical-Ubuntu-20.04-aarch64-2022.01.18-0. See how to list all available OS images
oci_core_vcn_dns_label no VCN DNS label. Default: defaultvcn
oci_core_subnet_dns_label10 no First subnet DNS label. Default: defaultsubnet10
oci_core_subnet_dns_label11 no Second subnet DNS label. Default: defaultsubnet11
oci_core_vcn_cidr no VCN CIDR. Default: oci_core_vcn_cidr
oci_core_subnet_cidr10 no First subnet CIDR. Default: 10.0.0.0/24
oci_core_subnet_cidr11 no Second subnet CIDR. Default: 10.0.1.0/24
oci_identity_dynamic_group_name no Dynamic group name. This dynamic group will contains all the instances of this specific compartment. Default: Compute_Dynamic_Group
oci_identity_policy_name no Policy name. This policy will allow dynamic group ‘oci_identity_dynamic_group_name’ to read OCI api without auth. Default: Compute_To_Oci_Api_Policy
k3s_load_balancer_name no Internal LB name. Default: k3s internal load balancer
public_load_balancer_name no Public LB name. Default: K3s public LB
kube_api_port no Kube api default port Default: 6443
public_lb_shape no LB shape for the public LB. Default: flexible. NOTE is mandatory to use this kind of shape to provision two always free LB (public and private)
http_lb_port no http port used by the public LB. Default: 80
https_lb_port no http port used by the public LB. Default: 443
k3s_server_pool_size no Number of k3s servers deployed. Default 2
k3s_worker_pool_size no Number of k3s workers deployed. Default 2
install_nginx_ingress no Boolean value, install kubernetes nginx ingress controller instead of Traefik. Default: true. For more information see Nginx ingress controller
install_longhorn no Boolean value, install longhorn “Cloud native distributed block storage for Kubernetes”. Default: true
longhorn_release no Longhorn release. Default: v1.2.3
unique_tag_key no Unique tag name used for tagging all the deployed resources. Default: k3s-provisioner
unique_tag_value no Unique value used with unique_tag_key. Default: https://github.com/garutilorenzo/k3s-oci-cluster
PATH_TO_PUBLIC_KEY no Path to your public ssh key (Default: “~/.ssh/id_rsa.pub)
PATH_TO_PRIVATE_KEY no Path to your private ssh key (Default: “~/.ssh/id_rsa)

Generate random token

Generate random k3s token with:

cat /dev/urandom | tr -dc 'a-zA-Z0-9' | fold -w 55 | head -n 1

How to find the availability doamin name

To find the list of the availability domains run this command on che Cloud Shell:

oci iam availability-domain list
{
  "data": [
    {
      "compartment-id": "<compartment_ocid>",
      "id": "ocid1.availabilitydomain.oc1..xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
      "name": "iAdc:EU-ZURICH-1-AD-1"
    }
  ]
}

How to list all the OS images

To filter the OS images by shape and OS run this command on che Cloud Shell:

oci compute image list --compartment-id <compartment_ocid> --operating-system "Canonical Ubuntu" --shape "VM.Standard.A1.Flex"
{
  "data": [
    {
      "agent-features": null,
      "base-image-id": null,
      "billable-size-in-gbs": 2,
      "compartment-id": null,
      "create-image-allowed": true,
      "defined-tags": {},
      "display-name": "Canonical-Ubuntu-20.04-aarch64-2022.01.18-0",
      "freeform-tags": {},
      "id": "ocid1.image.oc1.eu-zurich-1.aaaaaaaag2uyozo7266bmg26j5ixvi42jhaujso2pddpsigtib6vfnqy5f6q",
      "launch-mode": "NATIVE",
      "launch-options": {
        "boot-volume-type": "PARAVIRTUALIZED",
        "firmware": "UEFI_64",
        "is-consistent-volume-naming-enabled": true,
        "is-pv-encryption-in-transit-enabled": true,
        "network-type": "PARAVIRTUALIZED",
        "remote-data-volume-type": "PARAVIRTUALIZED"
      },
      "lifecycle-state": "AVAILABLE",
      "listing-type": null,
      "operating-system": "Canonical Ubuntu",
      "operating-system-version": "20.04",
      "size-in-mbs": 47694,
      "time-created": "2022-01-27T22:53:34.270000+00:00"
    },

Note: this setup was only tested with Ubuntu 20.04

Notes about OCI always free resources

In order to get the maximum resources available within the oracle always free tier, the max amount of the k3s servers and k3s workers must be 2. So the max value for k3s_server_pool_size and k3s_worker_pool_size is 2.

In this setup we use two LB, one internal LB and one public LB (Layer 7). In order to use two LB using the always free resources, one lb must be a network load balancer an the other must be a load balancer. The public LB must use the flexible shape (public_lb_shape variable).

Notes about K3s

In this environment the High Availability of the K3s cluster is provided using the Embedded DB. More details here

The default installation of K3s install Traefik as ingress the controller. In this environment Traefik is replaced by Nginx ingress controller. To install Traefik as the ingress controller set the variable install_nginx_ingress to false. For more details on Nginx ingress controller see the Nginx ingress controller section.

Infrastructure overview

The final infrastructure will be made by:

  • two instance pool:
    • one instance pool for the server nodes named “k3s-servers”
    • one instance pool for the worker nodes named “k3s-workers”
  • one internal load balancer that will route traffic to K3s servers
  • one external load balancer that will route traffic to K3s workers

The other resources created by terraform are:

  • two instance configurations (one for the servers and one for the workers) used by the instance pools
  • one vcn
  • two public subnets
  • two security list
  • one dynamic group
  • one identity policy

k3s-infra

Cluster resource deployed

This setup will automatically install longhorn. Longhorn is a Cloud native distributed block storage for Kubernetes. To disable the longhorn deployment set install_longhorn variable to false

Nginx ingress controller

In this environment Nginx ingress controller is used instead of the standard Traefik ingress controller.

The installation is the bare metal installation, the ingress controller then is exposed via a LoadBalancer Service.

---
apiVersion: v1
kind: Service
metadata:
  name: ingress-nginx-controller-loadbalancer
  namespace: ingress-nginx
spec:
  selector:
    app.kubernetes.io/component: controller
    app.kubernetes.io/instance: ingress-nginx
    app.kubernetes.io/name: ingress-nginx
  ports:
    - name: http
      port: 80
      protocol: TCP
      targetPort: 80
    - name: https
      port: 443
      protocol: TCP
      targetPort: 80
  type: LoadBalancer

To properly configure all the Forwarded HTTP Headers (L7 Headers) this parameters are added to che ConfigMap:

---
apiVersion: v1
data:
  allow-snippet-annotations: "true"
  use-forwarded-headers: "true"
  compute-full-forwarded-for: "true"
  enable-real-ip: "true"
  forwarded-for-header: "X-Forwarded-For"
  proxy-real-ip-cidr: "0.0.0.0/0"
kind: ConfigMap
metadata:
  labels:
    app.kubernetes.io/component: controller
    app.kubernetes.io/instance: ingress-nginx
    app.kubernetes.io/managed-by: Helm
    app.kubernetes.io/name: ingress-nginx
    app.kubernetes.io/part-of: ingress-nginx
    app.kubernetes.io/version: 1.1.1
    helm.sh/chart: ingress-nginx-4.0.16
  name: ingress-nginx-controller
  namespace: ingress-nginx

Deploy

We are now ready to deploy our infrastructure. First we ask terraform to plan the execution with:

terraform plan

...
...
      + id                             = (known after apply)
      + ip_addresses                   = (known after apply)
      + is_preserve_source_destination = false
      + is_private                     = true
      + lifecycle_details              = (known after apply)
      + nlb_ip_version                 = (known after apply)
      + state                          = (known after apply)
      + subnet_id                      = (known after apply)
      + system_tags                    = (known after apply)
      + time_created                   = (known after apply)
      + time_updated                   = (known after apply)

      + reserved_ips {
          + id = (known after apply)
        }
    }

Plan: 27 to add, 0 to change, 0 to destroy.

Changes to Outputs:
  + k3s_servers_ips = [
      + (known after apply),
      + (known after apply),
    ]
  + k3s_workers_ips = [
      + (known after apply),
      + (known after apply),
    ]
  + public_lb_ip    = (known after apply)

──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────

Note: You didn't use the -out option to save this plan, so Terraform can't guarantee to take exactly these actions if you run "terraform apply" now.

now we can deploy our resources with:

terraform apply

...
...
      + is_preserve_source_destination = false
      + is_private                     = true
      + lifecycle_details              = (known after apply)
      + nlb_ip_version                 = (known after apply)
      + state                          = (known after apply)
      + subnet_id                      = (known after apply)
      + system_tags                    = (known after apply)
      + time_created                   = (known after apply)
      + time_updated                   = (known after apply)

      + reserved_ips {
          + id = (known after apply)
        }
    }

Plan: 27 to add, 0 to change, 0 to destroy.

Changes to Outputs:
  + k3s_servers_ips = [
      + (known after apply),
      + (known after apply),
    ]
  + k3s_workers_ips = [
      + (known after apply),
      + (known after apply),
    ]
  + public_lb_ip    = (known after apply)

  Do you want to perform these actions?
  Terraform will perform the actions described above.
  Only 'yes' will be accepted to approve.
  Enter a value: yes

...
...

module.k3s_cluster.oci_network_load_balancer_backend.k3s_kube_api_backend[0]: Still creating... [50s elapsed]
module.k3s_cluster.oci_network_load_balancer_backend.k3s_kube_api_backend[0]: Still creating... [1m0s elapsed]
module.k3s_cluster.oci_network_load_balancer_backend.k3s_kube_api_backend[0]: Creation complete after 1m1s [...]

Apply complete! Resources: 27 added, 0 changed, 0 destroyed.

Outputs:

k3s_servers_ips = [
  "X.X.X.X",
  "X.X.X.X",
]
k3s_workers_ips = [
  "X.X.X.X",
  "X.X.X.X",
]
public_lb_ip = tolist([
  "X.X.X.X",
])

Now on one master node you can check the status of the cluster with:

ssh X.X.X.X -lubuntu

ubuntu@inst-iwlqz-k3s-servers:~$ sudo su -
root@inst-iwlqz-k3s-servers:~# kubectl get nodes

NAME                     STATUS   ROLES                       AGE     VERSION
inst-axdzf-k3s-workers   Ready    <none>                      4m34s   v1.22.6+k3s1
inst-hmgnl-k3s-servers   Ready    control-plane,etcd,master   4m14s   v1.22.6+k3s1
inst-iwlqz-k3s-servers   Ready    control-plane,etcd,master   6m4s    v1.22.6+k3s1
inst-lkvem-k3s-workers   Ready    <none>                      5m35s   v1.22.6+k3s1

Public LB check

We can now test the public load balancer, nginx ingress controller and the security list ingress rules. On your local PC run:

curl -v http://<PUBLIC_LB_IP>

*   Trying PUBLIC_LB_IP:80...
* TCP_NODELAY set
* Connected to PUBLIC_LB_IP (PUBLIC_LB_IP) port 80 (#0)
> GET / HTTP/1.1
> Host: PUBLIC_LB_IP
> User-Agent: curl/7.68.0
> Accept: */*
> 
* Mark bundle as not supporting multiuse
< HTTP/1.1 404 Not Found
< Date: Fri, 25 Feb 2022 14:03:09 GMT
< Content-Type: text/html
< Content-Length: 146
< Connection: keep-alive
< 
<html>
<head><title>404 Not Found</title></head>
<body>
<center><h1>404 Not Found</h1></center>
<hr><center>nginx</center>
</body>
</html>
* Connection #0 to host PUBLIC_LB_IP left intact

404 is a correct response since the cluster is empty. We can test also the https listener/backends:

curl -k -v https://<PUBLIC_LB_IP>

* Trying PUBLIC_LB_IP:443...
* TCP_NODELAY set
* Connected to PUBLIC_LB_IP (PUBLIC_LB_IP) port 443 (#0)
* ALPN, offering h2
* ALPN, offering http/1.1
* successfully set certificate verify locations:
*   CAfile: /etc/ssl/certs/ca-certificates.crt
  CApath: /etc/ssl/certs
* TLSv1.3 (OUT), TLS handshake, Client hello (1):
* TLSv1.3 (IN), TLS handshake, Server hello (2):
* TLSv1.2 (IN), TLS handshake, Certificate (11):
* TLSv1.2 (IN), TLS handshake, Server key exchange (12):
* TLSv1.2 (IN), TLS handshake, Server finished (14):
* TLSv1.2 (OUT), TLS handshake, Client key exchange (16):
* TLSv1.2 (OUT), TLS change cipher, Change cipher spec (1):
* TLSv1.2 (OUT), TLS handshake, Finished (20):
* TLSv1.2 (IN), TLS handshake, Finished (20):
* SSL connection using TLSv1.2 / ECDHE-RSA-AES256-GCM-SHA384
* ALPN, server accepted to use http/1.1
* Server certificate:
*  subject: C=IT; ST=Italy; L=Brescia; O=GL Ltd; OU=IT; CN=testlb.domainexample.com; emailAddress=email@you.com
*  start date: Feb 25 10:28:29 2022 GMT
*  expire date: Feb 25 10:28:29 2023 GMT
*  issuer: C=IT; ST=Italy; L=Brescia; O=GL Ltd; OU=IT; CN=testlb.domainexample.com; emailAddress=email@you.com
*  SSL certificate verify result: self signed certificate (18), continuing anyway.
> GET / HTTP/1.1
> Host: PUBLIC_LB_IP
> User-Agent: curl/7.68.0
> Accept: */*
> 
* Mark bundle as not supporting multiuse
< HTTP/1.1 404 Not Found
< Date: Fri, 25 Feb 2022 13:48:19 GMT
< Content-Type: text/html
< Content-Length: 146
< Connection: keep-alive
< 
<html>
<head><title>404 Not Found</title></head>
<body>
<center><h1>404 Not Found</h1></center>
<hr><center>nginx</center>
</body>
</html>
* Connection #0 to host PUBLIC_LB_IP left intact

Longhorn check

To check if longhorn was successfully installed run on one master nodes:

kubectl get ns
NAME              STATUS   AGE
default           Active   9m40s
kube-node-lease   Active   9m39s
kube-public       Active   9m39s
kube-system       Active   9m40s
longhorn-system   Active   8m52s   <- longhorn namespace 


root@inst-hmgnl-k3s-servers:~# kubectl get pods -n longhorn-system
NAME                                        READY   STATUS    RESTARTS        AGE
csi-attacher-5f46994f7-8w9sg                1/1     Running   0               7m52s
csi-attacher-5f46994f7-qz7d4                1/1     Running   0               7m52s
csi-attacher-5f46994f7-rjqlx                1/1     Running   0               7m52s
csi-provisioner-6ccbfbf86f-fw7q4            1/1     Running   0               7m52s
csi-provisioner-6ccbfbf86f-gwmrg            1/1     Running   0               7m52s
csi-provisioner-6ccbfbf86f-nsf84            1/1     Running   0               7m52s
csi-resizer-6dd8bd4c97-7l67f                1/1     Running   0               7m51s
csi-resizer-6dd8bd4c97-g66wj                1/1     Running   0               7m51s
csi-resizer-6dd8bd4c97-nksmd                1/1     Running   0               7m51s
csi-snapshotter-86f65d8bc-2gcwt             1/1     Running   0               7m50s
csi-snapshotter-86f65d8bc-kczrw             1/1     Running   0               7m50s
csi-snapshotter-86f65d8bc-sjmnv             1/1     Running   0               7m50s
engine-image-ei-fa2dfbf0-6rpz2              1/1     Running   0               8m30s
engine-image-ei-fa2dfbf0-7l5k8              1/1     Running   0               8m30s
engine-image-ei-fa2dfbf0-7nph9              1/1     Running   0               8m30s
engine-image-ei-fa2dfbf0-ndkck              1/1     Running   0               8m30s
instance-manager-e-31a0b3f5                 1/1     Running   0               8m26s
instance-manager-e-37aa4663                 1/1     Running   0               8m27s
instance-manager-e-9cc7cc9d                 1/1     Running   0               8m20s
instance-manager-e-f39d9f2c                 1/1     Running   0               8m29s
instance-manager-r-1364d994                 1/1     Running   0               8m26s
instance-manager-r-c1670269                 1/1     Running   0               8m20s
instance-manager-r-c20ebeb3                 1/1     Running   0               8m28s
instance-manager-r-c54bf9a5                 1/1     Running   0               8m27s
longhorn-csi-plugin-2qj94                   2/2     Running   0               7m50s
longhorn-csi-plugin-4t8jm                   2/2     Running   0               7m50s
longhorn-csi-plugin-ws82l                   2/2     Running   0               7m50s
longhorn-csi-plugin-zmc9q                   2/2     Running   0               7m50s
longhorn-driver-deployer-784546d78d-s6cd2   1/1     Running   0               8m58s
longhorn-manager-l8sd8                      1/1     Running   0               9m1s
longhorn-manager-r2q5c                      1/1     Running   1 (8m30s ago)   9m1s
longhorn-manager-s6wql                      1/1     Running   0               9m1s
longhorn-manager-zrrf2                      1/1     Running   0               9m
longhorn-ui-9fdb94f9-6shsr                  1/1     Running   0               8m59s

Deploy a sample stack

Finally to test all the components of the cluster we can deploy a sample stack. The stack is composed by the following components:

  • MariaDB
  • Nginx
  • Wordpress

Each component is made by: one deployment and one service. Wordpress and nginx share the same persistent volume (ReadWriteMany with longhorn storage class). The nginx configuration is stored in four ConfigMaps and the nginx service is exposed by the nginx ingress controller.

Deploy the resources with:

kubectl apply -f https://raw.githubusercontent.com/garutilorenzo/k3s-oci-cluster/master/deployments/mariadb/all-resources.yml
kubectl apply -f https://raw.githubusercontent.com/garutilorenzo/k3s-oci-cluster/master/deployments/nginx/all-resources.yml
kubectl apply -f https://raw.githubusercontent.com/garutilorenzo/k3s-oci-cluster/master/deployments/wordpress/all-resources.yml

and check the status:

kubectl get deployments
NAME        READY   UP-TO-DATE   AVAILABLE   AGE
mariadb       1/1     1            1           92m
nginx         1/1     1            1           79m
wordpress     1/1     1            1           91m

kubectl get svc
NAME            TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)    AGE
kubernetes        ClusterIP   10.43.0.1       <none>        443/TCP    5h8m
mariadb-svc       ClusterIP   10.43.184.188   <none>        3306/TCP   92m
nginx-svc         ClusterIP   10.43.9.202     <none>        80/TCP     80m
wordpress-svc     ClusterIP   10.43.242.26    <none>        9000/TCP   91m

Now you are ready to setup WP, open the LB public ip and follow the wizard. NOTE nginx and the Kubernetes Ingress rule are configured without virthual host/server name.

k3s-wp-install

To clean the deployed resources:

kubectl delete -f https://raw.githubusercontent.com/garutilorenzo/k3s-oci-cluster/master/deployments/mariadb/all-resources.yml
kubectl delete -f https://raw.githubusercontent.com/garutilorenzo/k3s-oci-cluster/master/deployments/nginx/all-resources.yml
kubectl delete -f https://raw.githubusercontent.com/garutilorenzo/k3s-oci-cluster/master/deployments/wordpress/all-resources.yml

Clean up

terraform destroy

Twitter, Facebook