Kubernetes 1.3 HA Walkthrough – NGINX

Table of Contents

You can find all of the config files on the GitHub page.

Overview

Let’s test out the environment by installing NGINX.

Create the NGINX deployment

We are going to create an NGINX deployment where we will have an NGINX replica on each of our Kubernetes worker nodes. A replica will be Kubernetes pod:

kubectl run nginx –image=nginx –port=80 –replicas=3

If successful, you will see: deployment “nginx” created

Let’s get a listing of each of the pods:

kubectl get pods -o wide

NAME                   READY STATUS  RESTARTS AGE IP          NODE
nginx-2032906785-44ndh 1/1   Running 2        16d 172.16.43.3 kube-worker2
nginx-2032906785-mvpft 1/1   Running 2        16d 172.16.19.2 kube-worker1
nginx-2032906785-tyfmu 1/1   Running 2        16d 172.16.23.2 kube-worker0

Here we can see that there is a pod on each of the worker nodes. Notice the IP address. Each of the pods (not docker container) will get an IP address from a network that was given out by etcd. Each worker node will be given a unique network range by etcd.

Create the NGNIX service

Now we need to create a Kubernetes Service to expose our NGINX deployment so we can reach it. I’m going to make the service of type NodePort, which will make the service accessible on any pod using the same port.

kubectl expose deployment nginx –port=80 –type=NodePort

View the service’s details

kubectl describe services nginx

Name:             nginx
Namespace:        default
Labels:           run=nginx
Selector:         run=nginx
Type:             NodePort
IP:               172.16.242.11
Port:              80/TCP
NodePort:          31153/TCP
Endpoints:        172.16.19.2:80,172.16.23.2:80,172.16.43.3:80
Session Affinity: None

Notice the NodePort and Endpoints. The Endpoints are our Kubernetes pods. If we launch a web browser and go to one of our worker nodes and use port 31153, we should see NGINX:

Lastly, let’s check out one of the pod’s details:

kubectl describe pod nginx-2032906785-44ndh

Name: nginx-2032906785-44ndh
Namespace: default
Node: kube-worker2/192.168.3.184
Start Time: Thu, 22 Sep 2016 20:20:45 -0600
Labels: pod-template-hash=2032906785
 run=nginx
Status: Running
IP: 172.16.43.3
Controllers: ReplicaSet/nginx-2032906785
Containers:
 nginx:
 Container ID: docker://19ffedded8de834da2e072f012c5081655b7149172d2c00d31944c7fe2499766
 Image: nginx
 Image ID: docker://sha256:ba6bed934df2e644fdd34e9d324c80f3c615544ee9a93e4ce3cfddfcf84bdbc2
 Port: 80/TCP
 State: Running
 Started: Sat, 08 Oct 2016 14:59:03 -0600
 Last State: Terminated
 Reason: Completed
 Exit Code: 0
 Started: Sun, 02 Oct 2016 14:38:18 -0600
 Finished: Mon, 03 Oct 2016 08:27:44 -0600
 Ready: True
 Restart Count: 2
 Environment Variables: <none>
Conditions:
 Type Status
 Initialized True
 Ready True
 PodScheduled True
Volumes:
 default-token-jeqk5:
 Type: Secret (a volume populated by a Secret)
 SecretName: default-token-jeqk5
QoS Tier: BestEffort
No events.

I’m not going to go into all of the details here, but you can read the Kubernetes Documentation.