pihole

Installs pihole in kubernetes

Source Code

• https://github.com/MoJo2600/pihole-kubernetes/tree/master/charts/pihole
• https://pi-hole.net/
• https://github.com/pi-hole
• https://github.com/pi-hole/docker-pi-hole

Installation

Jeff Geerling on YouTube made a video about the installation of this chart:

Add Helm repository

helm repo add mojo2600 https://mojo2600.github.io/pihole-kubernetes/
helm repo update

Configure the chart

The following items can be set via --set flag during installation or configured by editing the values.yaml directly.

Configure the way how to expose pihole service:

• Ingress: The ingress controller must be installed in the Kubernetes cluster.
• ClusterIP: Exposes the service on a cluster-internal IP. Choosing this value makes the service only reachable from within the cluster.
• LoadBalancer: Exposes the service externally using a cloud provider’s load balancer.

My settings in values.yaml

dnsmasq:
  customDnsEntries:
    - address=/nas/192.168.178.10
 
  customCnameEntries:
    - cname=foo.nas,nas

persistentVolumeClaim:
  enabled: true

serviceWeb:
  loadBalancerIP: 192.168.178.252
  annotations:
    metallb.universe.tf/allow-shared-ip: pihole-svc
  type: LoadBalancer

serviceDns:
  loadBalancerIP: 192.168.178.252
  annotations:
    metallb.universe.tf/allow-shared-ip: pihole-svc
  type: LoadBalancer

Configuring Upstream DNS Resolvers

By default, pihole-kubernetes will configure pod DNS automatically to use Google's 8.8.8.8 nameserver for upstream DNS resolution. You can configure this, or opt-out of pod DNS configuration completely.

Changing The Upstream DNS Resolver

For example, to use Cloudflare's resolver:

podDnsConfig:
  enabled: true
  policy: "None"
  nameservers:
  - 127.0.0.1
  - 1.1.1.1

Disabling Pod DNS Configuration

If you have other DNS policy at play (for example, when running a service mesh), you may not want to have pihole-kubernetes control this behavior. In that case, you can disable DNS configuration on pihole pods:

podDnsConfig:
  enabled: false

Upgrading

To 2.0.0

This version splits the DHCP service into its own resource and puts the configuration to serviceDhcp.

If you have not changed any configuration for serviceDns, you don’t need to do anything.

If you have changed your serviceDns configuration, copy your serviceDns section into a new serviceDhcp section.

To 1.8.22

To enhance compatibility for Traefik, we split the TCP and UDP service into Web and DNS. This means, if you have a dedicated configuration for the service, you have to update your values.yaml and add a new configuration for this new service.

Before (In my case, with metallb):

serviceTCP:
  loadBalancerIP: 192.168.178.252
  annotations:
    metallb.universe.tf/allow-shared-ip: pihole-svc

serviceUDP:
  loadBalancerIP: 192.168.178.252
  annotations:
    metallb.universe.tf/allow-shared-ip: pihole-svc

After:

serviceWeb:
  loadBalancerIP: 192.168.178.252
  annotations:
    metallb.universe.tf/allow-shared-ip: pihole-svc

serviceDns:
  loadBalancerIP: 192.168.178.252
  annotations:
    metallb.universe.tf/allow-shared-ip: pihole-svc

Version 1.8.22 has switched from the deprecated ingress api extensions/v1beta1 to the go forward version networking.k8s.io/v1. This means that your cluster must be running 1.19.x as this api is not available on older versions. If necessary to run on an older Kubernetes Version, it can be done by modifying the ingress.yaml and changing the api definition back. The backend definition would also change from:

            backend:
              service:
                name: \{\{ $serviceName \}\}
                port:
                  name: http

to:

            backend:
              serviceName: \{\{ $serviceName \}\}
              servicePort: http

Uninstallation

To uninstall/delete the my-release deployment (NOTE: --purge is default behaviour in Helm 3+ and will error):

helm delete --purge my-release

Configuration

The following table lists the configurable parameters of the pihole chart and the default values.

Values

Maintainers

Remarks

MetalLB 0.8.1+

pihole seems to work without issue in MetalLB 0.8.1+

MetalLB 0.7.3

MetalLB 0.7.3 has a bug, where the service is not announced anymore, when the pod changes (e.g. update of a deployment). My workaround is to restart the metallb-speaker-* pods.

Credits

Pi-hole®

Contributing

Feel free to contribute by making a pull request.

Please read Contribution Guide for more information on how you can contribute to this Chart.

Contributors ✨

Thanks goes to these wonderful people:

This project follows the all-contributors specification. Contributions of any kind welcome!

Autogenerated from chart metadata using helm-docs v1.10.0