Install on premises

This guide shows you how to install HAProxy Kubernetes Ingress Controller into an on-premises Kubernetes cluster. You can install either with Helm or kubectl.

Pre-installation checklist Jump to heading #

• A running Kubernetes cluster
• The helm command-line tool
• The kubectl command-line tool

Choose one of the following installation methods:

Install with Helm Jump to heading #

1. Add the HAProxy Technologies Helm repository:

   bash
   helm repo add haproxytech https://haproxytech.github.io/helm-charts

   bash
   helm repo add haproxytech https://haproxytech.github.io/helm-charts

2. Update your list of Helm charts:

   bash
   helm repo update

   bash
   helm repo update

3. Install the latest version of the ingress controller:

   bash
   helm install haproxy-kubernetes-ingress haproxytech/kubernetes-ingress \
     --create-namespace \
     --namespace haproxy-controller

   bash
   helm install haproxy-kubernetes-ingress haproxytech/kubernetes-ingress \
     --create-namespace \
     --namespace haproxy-controller

   About ingress class

   By default, the Helm chart adds --ingress.class=haproxy to the ingress controller. That means that it will use Ingress resources only if they specify ingressClassName: haproxy. You can disable this by setting --set controller.ingressClass=null when calling helm install.

   This change was introduced in version 1.6.

Install a Specific Version Jump to heading #

1. You can use helm search repo with the --versions flag to list all of the available versions of the Helm chart:

   bash
   helm search repo haproxy --versions

   bash
   helm search repo haproxy --versions

   output
   bash
   NAME                            CHART VERSION  APP VERSION  DESCRIPTION
   haproxytech/kubernetes-ingress  1.16.3         1.6.7        A Helm chart for HAProxy Kubernetes Ingress Con...
   haproxytech/kubernetes-ingress  1.16.2         1.6.5        A Helm chart for HAProxy Kubernetes Ingress Con...
   haproxytech/kubernetes-ingress  1.16.1         1.6.5        A Helm chart for HAProxy Kubernetes Ingress Con...
   haproxytech/kubernetes-ingress  1.16.0         1.6.3        A Helm chart for HAProxy Kubernetes Ingress Con...
   haproxytech/kubernetes-ingress  1.15.4         1.6.3        A Helm chart for HAProxy Kubernetes Ingress Con...

   output
   bash
   NAME                            CHART VERSION  APP VERSION  DESCRIPTION
   haproxytech/kubernetes-ingress  1.16.3         1.6.7        A Helm chart for HAProxy Kubernetes Ingress Con...
   haproxytech/kubernetes-ingress  1.16.2         1.6.5        A Helm chart for HAProxy Kubernetes Ingress Con...
   haproxytech/kubernetes-ingress  1.16.1         1.6.5        A Helm chart for HAProxy Kubernetes Ingress Con...
   haproxytech/kubernetes-ingress  1.16.0         1.6.3        A Helm chart for HAProxy Kubernetes Ingress Con...
   haproxytech/kubernetes-ingress  1.15.4         1.6.3        A Helm chart for HAProxy Kubernetes Ingress Con...

2. Then, use a specific chart version when you call helm install:

   bash
   helm install haproxy-kubernetes-ingress haproxytech/kubernetes-ingress \
     --create-namespace \
     --namespace haproxy-controller \
     --version 1.16.2

   bash
   helm install haproxy-kubernetes-ingress haproxytech/kubernetes-ingress \
     --create-namespace \
     --namespace haproxy-controller \
     --version 1.16.2

Install with preset NodePort values Jump to heading #

By default, the ingress controller creates a Kubernetes service that assigns random NodePort ports. You can see this by calling kubectl get service kubernetes-service. In the example below, the ingress controller listens on port 30706 for HTTP traffic and 30675 for HTTPS traffic. Port 31441 publishes the HAProxy Stats page.

bash
kubectl get service haproxy-kubernetes-ingress --namespace haproxy-controller

bash
kubectl get service haproxy-kubernetes-ingress --namespace haproxy-controller

output
bash
NAME                         TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                                     AGE
haproxy-kubernetes-ingress   NodePort    10.97.154.7      <none>        80:30706/TCP,443:30675/TCP,1024:31441/TCP   32s

output
bash
NAME                         TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                                     AGE
haproxy-kubernetes-ingress   NodePort    10.97.154.7      <none>        80:30706/TCP,443:30675/TCP,1024:31441/TCP   32s

You can choose specific NodePort ports instead by setting the controller.service.nodePort.http, controller.service.nodePort.https, and controller.service.nodePort.stat Helm chart fields.

bash
helm install haproxy-kubernetes-ingress haproxytech/kubernetes-ingress \
  --create-namespace \
  --namespace haproxy-controller \
  --set controller.service.nodePorts.http=30000 \
  --set controller.service.nodePorts.https=30001 \
  --set controller.service.nodePorts.stat=30002

bash
helm install haproxy-kubernetes-ingress haproxytech/kubernetes-ingress \
  --create-namespace \
  --namespace haproxy-controller \
  --set controller.service.nodePorts.http=30000 \
  --set controller.service.nodePorts.https=30001 \
  --set controller.service.nodePorts.stat=30002

Run the ingress controller as a DaemonSet Jump to heading #

By default, the ingress controller is installed using a Deployment resource. To use a DaemonSet instead, set the controller.kind field to DaemonSet and controller.daemonset.useHostPort to true.

bash
helm install haproxy-kubernetes-ingress haproxytech/kubernetes-ingress \
  --create-namespace \
  --namespace haproxy-controller \
  --set controller.kind=DaemonSet \
  --set controller.daemonset.useHostPort=true

bash
helm install haproxy-kubernetes-ingress haproxytech/kubernetes-ingress \
  --create-namespace \
  --namespace haproxy-controller \
  --set controller.kind=DaemonSet \
  --set controller.daemonset.useHostPort=true

Install with kubectl Jump to heading #

If you prefer to install without Helm, then you can install the Deployment YAML file directly.

1. Download the latest source code from the HAProxy Kubernetes Ingress Controller GitHub page.

2. Use the kubectl apply command to deploy the controller:

   bash
   kubectl apply -f deploy/haproxy-ingress.yaml

   bash
   kubectl apply -f deploy/haproxy-ingress.yaml

   You can also download the YAML file and modify it for your needs before you apply it (such as changing the namespace in which the controller runs and/or its exposed ports).

Check your installation Jump to heading #

Verify that the controller is installed into your Kubernetes cluster by using the command kubectl get pods:

bash
kubectl get pods --namespace haproxy-controller

bash
kubectl get pods --namespace haproxy-controller

output
bash
NAME                                       READY   STATUS    RESTARTS   AGE
haproxy-kubernetes-ingress-7dd4cc4b-x5fkv  1/1     Running   0          1m

output
bash
NAME                                       READY   STATUS    RESTARTS   AGE
haproxy-kubernetes-ingress-7dd4cc4b-x5fkv  1/1     Running   0          1m

Call kubectl get svc to see which ports were mapped. In this example, port 32063 forwards to port 80 on the ingress controller.

bash
kubectl get svc --namespace haproxy-controller

bash
kubectl get svc --namespace haproxy-controller

output
bash
NAME                         TYPE       CLUSTER-IP       EXTERNAL-IP   PORT(S)                                     AGE
haproxy-kubernetes-ingress   NodePort   10.104.173.167   <none>        80:30264/TCP,443:31575/TCP,1024:31785/TCP   155m

output
bash
NAME                         TYPE       CLUSTER-IP       EXTERNAL-IP   PORT(S)                                     AGE
haproxy-kubernetes-ingress   NodePort   10.104.173.167   <none>        80:30264/TCP,443:31575/TCP,1024:31785/TCP   155m