see this article
A deployment and service needs to be implemented to create :
a pod with an application in this case called marshyon/test-node-listener - see this repo for the code behind this
a service of type ClusterIP , the default if not specified and by its nature only available internally within the K8s cluster
The latter is important to note as if this were to be of type
LoadBalancer, the service would be exposed by the underlying cloud fabric and this would be a 1:1 relationship in that for everyserviceof typeLoadBalancerthere would be a separate load balancer and external IP address created. With large or complicated applications this could quickly become unwieldy. By deploying an ingress controller, a single load balancer is created that can distribute requests to different back end services by host name and or urls.
Here is a manifest file for a deployment and service - post.yaml :
apiVersion: v1
kind: Service
metadata:
name: post1
spec:
ports:
- port: 80
targetPort: 4000
selector:
app: post1
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: post1
spec:
selector:
matchLabels:
app: post1
replicas: 2
template:
metadata:
labels:
app: post1
spec:
containers:
- name: post1
image: marshyon/test-node-listener
args:
- "-text=post1"
ports:
- containerPort: 4000
it is applied with :
kubectl apply -f .\post1.yaml
the deployment can be seen to be running with :
kubectl.exe get deployments
NAME READY UP-TO-DATE AVAILABLE AGE
post1 2/2 2 2 15m
running pods :
kubectl get pods
NAME READY STATUS RESTARTS AGE
post1-68fb864b7d-4clbh 1/1 Running 0 6m42s
post1-68fb864b7d-msz2x 1/1 Running 0 6m42s
Kubernetes has standard deployments for popular Cloud providers, at time of writing this includes Amazon, Azure, Google, Digital Ocean and Scaleway. For specific deployments see https://kubernetes.github.io/ingress-nginx/deploy
The following is for Digital Ocean :
kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/controller-v0.41.2/deploy/static/provider/do/deploy.yaml
Taking a look at the url there are a number of manifests each separated by --- character breaks and consisting of 10 kinds :
curl -s https://raw.githubusercontent.com/kubernetes/ingress-nginx/controller-v0.41.2/deploy/static/provider/do/deploy.yaml | egrep "^kind: " | cut -f2 -d' ' | grep -v Namespace | sort | uniq
ClusterRole
ClusterRoleBinding
ConfigMap
Deployment
Job
Role
RoleBinding
Service
ServiceAccount
ValidatingWebhookConfiguration
it is worth reviewing these as there are 17 in total :
curl -s https://raw.githubusercontent.com/kubernetes/ingress-nginx/controller-v0.41.2/deploy/static/provider/do/deploy.yaml | egrep "^kind: " | cut -f2 -d' ' | grep -v Namespace | sort | wc -l
17
and this can offer insight into how complex services are deployed in K8s and can also serve to inform how such ingress deployments may be crafted to suit applications for different applications and clients.
To this point all of the infrastructure has been implemented but there is no exposure of the running service to the world outside of K8s.
The ingress configuration file : ingress.yaml
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
metadata:
name: post-ingress
spec:
rules:
- host: post1.notapplicable.info
http:
paths:
- backend:
serviceName: post1
servicePort: 80
is applied with :
kubectl apply -f .\ingress.yaml
its status is obtained from :
kubectl get ingress
Warning: extensions/v1beta1 Ingress is deprecated in v1.14+, unavailable in v1.22+; use networking.k8s.io/v1 Ingress
NAME CLASS HOSTS ADDRESS PORTS AGE
echo-ingress <none> post1.notapplicable.info 206.189.245.136 80 79m
Here the ip address 206.189.245.136 is created by the underlying Digital Ocean infrastructure.
A final step, entirely outside of K8s is required to create a DNS A record of post1.notapplicable.info ( replacing post1.notapplicable.info with the domain name that you own ) and resolving to the IP address in your ingress configuration.