Skip to content

loadbalancer-examples.md

Latest commit

 

History

History
124 lines (97 loc) · 3.48 KB

loadbalancer-examples.md

File metadata and controls

124 lines (97 loc) · 3.48 KB
Raw

Scaleway LoadBalancer Service example

This example will show you how to use the cloud-controller-manager to create a service of type: LoadBalancer.

Requirements

First, you need a cluster running a cloud-controller-manager, this could be a Scaleway Kapsule or your own installed kubernetes.

To create a load balancer you first have to have an running application. In the example below we are going to create a webserver. This web server will then be reached from the outside by creating a LoadBalancer.

# examples/demo-nginx-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment
spec:
  replicas: 2
  selector:
    matchLabels:
      app: demo-nginx
  template:
    metadata:
      labels:
        app: demo-nginx
    spec:
      containers:
      - name: nginx
        image: nginx
        ports:
        - containerPort: 80

Execute it

kubectl create -f examples/demo-nginx-deployment.yaml

Deploy a TCP Loadbalancer

The example below will expose the deployment and via a loadbalancer. Note that the service type is set to LoadBalancer.

# examples/demo-nginx-svc.yaml
kind: Service
apiVersion: v1
metadata:
  name: demo-nginx
spec:
  selector:
    app: demo-nginx
  type: LoadBalancer
  ports:
  - name: http
    port: 80
    targetPort: 80

Execute it

kubectl create -f examples/demo-nginx-svc.yaml

Deploying a loadbalancer takes few seconds, wait until EXTERNAL-IP address appear. This will be the load balancer IP which you can use to connect to your service.

Note: you can append --watch to the command to automaticaly refresh the result.

$ kubectl get svc
NAME            CLUSTER-IP     EXTERNAL-IP      PORT(S)        AGE
demo-nginx      10.96.97.137   203.0.113.10     80:30132/TCP   3m

You can now access your service via the provisioned load balancer

curl -i http://203.0.113.10

Deploy a loadbalancer with specific IP

Instead of creating a loadbalancer service with an ephemeral IP (which changes each time your delete/create the service), you can use a reserved loadbalancer address.

Warning: LoadBalancer ips are different from Instance ips. You cannot use an instance ip on a load-balancer and vice versa.

To reserve a loadbalancer IP using the Scaleway CLI:

scw lb ip create zone=$SCW_DEFAULT_ZONE project-id=$SCW_DEFAULT_PROJECT_ID -o json | jq -r .id

Now specify the ID of this IP address in the service.beta.kubernetes.io/scw-loadbalancer-ip-ids annotation.

kind: Service
apiVersion: v1
metadata:
  name: demo-nginx
  annotations:
    service.beta.kubernetes.io/scw-loadbalancer-ip-ids: IP_ID
spec:
  selector:
    app: demo-nginx
  type: LoadBalancer
  ports:
  - name: http
    port: 80
    targetPort: 80

Convert an ephemeral IP into reserved IP

It's possible to keep an ephemeral address (converting into a reserved address) by patching the associated service.

export CURRENT_EXTERNAL_IP=$(kubectl get svc $SERVICE_NAME -o json | jq -r .status.loadBalancer.ingress[0].ip)
export CURRENT_EXTERNAL_IP_ID=$(scw lb ip list zone=$SCW_DEFAULT_ZONE ip-address=$CURRENT_EXTERNAL_IP -o json | jq -r '.[0].id')
kubectl patch svc $SERVICE_NAME --type merge --patch "{\"metadata\":{\"annotations\": {\"service.beta.kubernetes.io/scw-loadbalancer-ip-ids\": \"$CURRENT_EXTERNAL_IP_ID\"}}}"

This way, the ephemeral ip would not be deleted when the service will be deleted and could be reused in another service.