How to set up a Kubernetes cluster in an automated fashion on our LAB environment? This article will walk you through kubernetes installation on Ubuntu on my windows 10 workstation. Getting hands-on on kubernetes is not a big task nowadays. There are hundreds of kubernetes variants available in the market including managed kubernetes environments like EKS, AKS, GKE, etc …
Minikube is a lightweight Kubernetes implementation that creates a VM on your local machine and deploys a simple cluster containing only one node. This is the perfect platform for developers and beginners to get hands-on with kubernetes. In my environment, I have set up a mainstream kubernetes environment.
In My LAB, I have set up the Kubernetes environment with 1 Master and 2 worker nodes using Kubespray.
Infrastructure:
In Windows 10, I have installed VMware Player which is free for non-commercial purposes. VMware workstation offers more features compared to the player and VirtualBox is a very popular open-source hypervisor. You can use any one of the virtualization technology to set up a home LAB.
I have created 3 Ubuntu virtual machines on a VMware player. The plan is to create 1 master and 2 worker nodes. All three VMs will have internet connectivity to download the images from the internet for various deployments.
Setting up k8s Cluster:
There are multiple automation methods available in the market to get a production-grade Kubernetes cluster up and running. kOps is one of the popular methods to set up a k8s cluster but it’s tightly integrated with unique features. For people with familiarity with Ansible, existing Ansible deployments, or the desire to run a Kubernetes cluster across multiple platforms, Kubespray is a good choice. I have used kubespary to bootstrap the cluster.
I do not have a dedicated node for an ansible host to set up kubespary. I have used the master node to perform the cluster setup. Password-less authentication has been configured between both worker nodes.
root@uamaster1:~# ssh uaworker1 uname -n uaworker1 root@uamaster1:~# ssh uaworker2 uname -n uaworker2 root@uamaster1:~#
1. Clone the kubespary git repo.
# git clone https://github.com/kubernetes-sigs/kubespray.git
2. Create the inventory file to reflect the master and worker node details.
root@uamaster1:~/# cd kubespray/inventory/sample root@uamaster1:~/kubespray/inventory/sample# cat uak8s.ini |grep -v "#" [all] uamaster1 ansible_host=192.168.23.132 ip=192.168.23.132 uaworker1 ansible_host=192.168.23.133 ip=192.168.23.133 uaworker2 ansible_host=192.168.23.134 ip=192.168.23.134 [kube_control_plane] uamaster1 [etcd] uamaster1 [kube_node] uaworker1 uaworker2 [calico_rr] [k8s_cluster:children] kube_control_plane kube_node calico_rr root@uamaster1:~/kubespray/inventory/sample#
3. Start the cluster installation using the following command.
root@uamaster1:~/kubespray/inventory/sample# ansible-playbook -i uak8s.ini ../../cluster.yml -b -v
TASK [network_plugin/calico : Check if inventory match current cluster configuration] ****************************
ok: [uamaster1] => {
"changed": false,
"msg": "All assertions passed"
}
TASK [network_plugin/calico : Check kdd calico_datastore if calico_apiserver_enabled] *****************************
skipping: [uamaster1] => {"changed": false, "skip_reason": "Conditional result was False"}
TASK [network_plugin/calico : Check kdd calico_datastore if typha_enabled] ****************************************
skipping: [uamaster1] => {"changed": false, "skip_reason": "Conditional result was False"}
PLAY RECAP *******************************************************************************************************
localhost : ok=4 changed=0 unreachable=0 failed=0 skipped=0 rescued=0 ignored=0
uamaster1 : ok=743 changed=113 unreachable=0 failed=0 skipped=1298 rescued=0 ignored=5
uaworker1 : ok=503 changed=67 unreachable=0 failed=0 skipped=770 rescued=0 ignored=1
uaworker2 : ok=503 changed=67 unreachable=0 failed=0 skipped=769 rescued=0 ignored=1
4. Once the playbook is completed successfully, you can able to list the nodes using the “kubectl” command.
root@uamaster1:~/kubespray/inventory/sample# kubectl get nodes NAME STATUS ROLES AGE VERSION uamaster1 Ready control-plane,master 16m v1.23.5 uaworker1 Ready <none> 14m v1.23.5 uaworker2 Ready <none> 14m v1.23.5 root@uamaster1:~/kubespray/inventory/sample#
5. Let’s list all the kubernetes objects.
root@uamaster1:~/kubespray/inventory/sample# kubectl get all -A NAMESPACE NAME READY STATUS RESTARTS AGE kube-system pod/calico-kube-controllers-75fcdd655b-rb24b 1/1 Running 0 13m kube-system pod/calico-node-9mktm 1/1 Running 0 14m kube-system pod/calico-node-kjc2b 1/1 Running 0 14m kube-system pod/calico-node-vdf2b 1/1 Running 0 14m kube-system pod/coredns-76b4fb4578-khfkn 1/1 Running 0 12m kube-system pod/coredns-76b4fb4578-zjmhz 1/1 Running 0 12m kube-system pod/dns-autoscaler-7979fb6659-j6md2 1/1 Running 0 12m kube-system pod/kube-apiserver-uamaster1 1/1 Running 1 16m kube-system pod/kube-controller-manager-uamaster1 1/1 Running 2 (11m ago) 16m kube-system pod/kube-proxy-826q6 1/1 Running 0 14m kube-system pod/kube-proxy-rfknc 1/1 Running 0 15m kube-system pod/kube-proxy-zzn5x 1/1 Running 0 15m kube-system pod/kube-scheduler-uamaster1 1/1 Running 2 (11m ago) 16m kube-system pod/nginx-proxy-uaworker1 1/1 Running 0 15m kube-system pod/nginx-proxy-uaworker2 1/1 Running 0 15m kube-system pod/nodelocaldns-l5zkd 1/1 Running 0 12m kube-system pod/nodelocaldns-qpd6r 1/1 Running 0 12m kube-system pod/nodelocaldns-wn6qm 1/1 Running 0 12m NAMESPACE NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE default service/kubernetes ClusterIP 10.233.0.1 <none> 443/TCP 16m kube-system service/coredns ClusterIP 10.233.0.3 <none> 53/UDP,53/TCP,9153/TCP 12m NAMESPACE NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE kube-system daemonset.apps/calico-node 3 3 3 3 3 kubernetes.io/os=linux 14m kube-system daemonset.apps/kube-proxy 3 3 3 3 3 kubernetes.io/os=linux 16m kube-system daemonset.apps/nodelocaldns 3 3 3 3 3 kubernetes.io/os=linux 12m NAMESPACE NAME READY UP-TO-DATE AVAILABLE AGE kube-system deployment.apps/calico-kube-controllers 1/1 1 1 13m kube-system deployment.apps/coredns 2/2 2 2 12m kube-system deployment.apps/dns-autoscaler 1/1 1 1 12m NAMESPACE NAME DESIRED CURRENT READY AGE kube-system replicaset.apps/calico-kube-controllers-75fcdd655b 1 1 1 13m kube-system replicaset.apps/coredns-76b4fb4578 2 2 2 12m kube-system replicaset.apps/dns-autoscaler-7979fb6659 1 1 1 12m root@uamaster1:~/kubespray/inventory/sample#
Deploy a new application:
Let’s deploy a new application to test the cluster environment.
1. Create a new deployment file like the one below.
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
namespace: nginx-test
labels:
app: nginx
spec:
replicas: 3
selector:
matchLabels:
app: nginx
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:1.14.2
ports:
- containerPort: 80
2. Create a new namespace
lingesh@uamaster1:~$ kubectl create namespace nginx-test namespace/nginx-test created lingesh@uamaster1:~$
3. Create the deployment.
lingesh@uamaster1:~$ kubectl create -f test_ngnix.yaml deployment.apps/nginx-deployment created lingesh@uamaster1:~$
4. Check the deployed resources on the newly created namespace.
lingesh@uamaster1:~$ kubectl get all -n nginx-test NAME READY STATUS RESTARTS AGE pod/nginx-deployment-9456bbbf9-k92b4 0/1 ContainerCreating 0 7s pod/nginx-deployment-9456bbbf9-vnm9d 0/1 ContainerCreating 0 7s pod/nginx-deployment-9456bbbf9-z7qcn 0/1 ContainerCreating 0 7s NAME READY UP-TO-DATE AVAILABLE AGE deployment.apps/nginx-deployment 0/3 3 0 7s NAME DESIRED CURRENT READY AGE replicaset.apps/nginx-deployment-9456bbbf9 3 3 0 7s lingesh@uamaster1:~$
Here we can see that container is creating. If you check after a few seconds container should be up and running like below.
lingesh@uamaster1:~$ kubectl get all -n nginx-test NAME READY STATUS RESTARTS AGE pod/nginx-deployment-9456bbbf9-k92b4 1/1 Running 0 26s pod/nginx-deployment-9456bbbf9-vnm9d 1/1 Running 0 26s pod/nginx-deployment-9456bbbf9-z7qcn 1/1 Running 0 26s NAME READY UP-TO-DATE AVAILABLE AGE deployment.apps/nginx-deployment 3/3 3 3 26s NAME DESIRED CURRENT READY AGE replicaset.apps/nginx-deployment-9456bbbf9 3 3 3 26s lingesh@uamaster1:~$
We have successfully set up the Kubernetes environment using kubespary. In the upcoming article, we will see how to expose the service outside of the cluster using node port and ingress controllers. There will be a few more components like an ingress controller, certificate manager, and Load balancer that need to be deployed in an order to simulate the real kubernetes production environment.