README.md

Add the following lines to /boot/firmware/nobtcmd.txt

cgroup_enable=memory cgroup_memory=1

and /boot/firmware/cmdline.txt

net.ifnames=0 dwc_otg.lpm_enable=0 console=serial0,115200 console=tty1 root=LABEL=writable rootfstype=ext4 elevator=deadline rootwait fixrtc cgroup_enable=cpuset cgroup_enable=memory cgroup_memory=1

to enable cgroups for microk8s on a raspberry pi.

Enable the following plugins

microk8s enable dashboard
microk8s enable dns
microk8s enable prometheus
microk8s enable helm3
microk8s enable ingress
microk8s enable storage

TCP-/UDP Ingress

In order to get other nodes to connect to your beacon node, you need make sure, that the ports 13000/TCP and 12000/UDP are open to the outside. In this example, the nginx ingress already enabled is the weapon of choice. For both ports, the traffic flow will look roughly like this:

internet -> host machine -> nginx-ingress-controller -> beacon-node-service -> beacon-node-pod

Following the official documentation, we have to parameterize the ingress controller with config maps containing the routing decisions for the opened ports. Therefore we start by creating the according config maps in the ingress namespace:

$ kubectl apply -n ingress -f ./ingress-tcp-services-configmap.yaml
$ kubectl apply -n ingress -f ./ingress-udp-services-configmap.yaml

Further, the default implementation of the microk8s ingress controller daemonset is not parameterized with the created configmaps. In the example below, the highlighted lines have to be added:

$ kubectl describe -n ingress daemonsets.apps nginx-ingress-microk8s-controller
Name:           nginx-ingress-microk8s-controller
Selector:       name=nginx-ingress-microk8s
Node-Selector:  <none>
Labels:         microk8s-application=nginx-ingress-microk8s
Annotations:    deprecated.daemonset.template.generation: 3
                kubectl.kubernetes.io/last-applied-configuration:
                  {"apiVersion":"apps/v1","kind":"DaemonSet","metadata":{"annotations":{},"labels":{"microk8s-application":"nginx-ingress-microk8s"},"name":...
Desired Number of Nodes Scheduled: 1
Current Number of Nodes Scheduled: 1
Number of Nodes Scheduled with Up-to-date Pods: 1
Number of Nodes Scheduled with Available Pods: 1
Number of Nodes Misscheduled: 0
Pods Status:  1 Running / 0 Waiting / 0 Succeeded / 0 Failed
Pod Template:
  Labels:           name=nginx-ingress-microk8s
  Service Account:  nginx-ingress-microk8s-serviceaccount
  Containers:
   nginx-ingress-microk8s:
    Image:       quay.io/kubernetes-ingress-controller/nginx-ingress-controller-arm64:0.25.1
    Ports:       80/TCP, 443/TCP
    Host Ports:  80/TCP, 443/TCP
    Args:
      /nginx-ingress-controller
      --configmap=$(POD_NAMESPACE)/nginx-load-balancer-microk8s-conf
      --publish-status-address=127.0.0.1
      --tcp-services-configmap=$(POD_NAMESPACE)/tcp-services
      --udp-services-configmap=$(POD_NAMESPACE)/udp-services
    Liveness:  http-get http://:10254/healthz delay=30s timeout=5s period=10s #success=1 #failure=3
    Environment:
      POD_NAME:        (v1:metadata.name)
      POD_NAMESPACE:   (v1:metadata.namespace)
    Mounts:           <none>
  Volumes:            <none>
Events:
  Type    Reason            Age   From                  Message
  ----    ------            ----  ----                  -------
  Normal  SuccessfulCreate  29m   daemonset-controller  Created pod: nginx-ingress-microk8s-controller-8t9p9

We therefore have to edit the daemonset, i.e. with the following command:

$ kubectl edit -n ingress daemonsets.apps nginx-ingress-microk8s-controller

Afterwards we have to restart the ingress controller by deleting the controller pod:

kubectl delete -n ingress po nginx-ingress-microk8s-controller-<pod-id>

if everything worked out, you can see in the logs that the according configmaps were loaded. Check with:

$ kubectl logs -n ingress nginx-ingress-microk8s-<pod-id> | grep "'CREATE' ConfigMap ingress"
I0819 20:56:49.807074       7 event.go:258] Event(v1.ObjectReference{Kind:"ConfigMap", Namespace:"ingress", Name:"tcp-services", UID:"5b280307-8969-4b00-916d-4f26fd4c6a3c", APIVersion:"v1", ResourceVersion:"2636901", FieldPath:""}): type: 'Normal' reason: 'CREATE' ConfigMap ingress/tcp-services
I0819 20:56:49.810126       7 event.go:258] Event(v1.ObjectReference{Kind:"ConfigMap", Namespace:"ingress", Name:"nginx-load-balancer-microk8s-conf", UID:"51384a02-9989-447d-819f-9cbc6d6a8b93", APIVersion:"v1", ResourceVersion:"2569059", FieldPath:""}): type: 'Normal' reason: 'CREATE' ConfigMap ingress/nginx-load-balancer-microk8s-conf
I0819 20:56:49.832790       7 event.go:258] Event(v1.ObjectReference{Kind:"ConfigMap", Namespace:"ingress", Name:"udp-services", UID:"33a70fe1-1467-4676-ad7a-53755d8d0909", APIVersion:"v1", ResourceVersion:"2639337", FieldPath:""}): type: 'Normal' reason: 'CREATE' ConfigMap ingress/udp-services

And verify with netcat that the ports are in fact open, and are being routed to the beacon node:

$ nc -z -v <host-machine> 13000                                                                                                                                      [±master ●●]
Connection to <host-machine> 13000 port [tcp/*] succeeded!
$ nc -z -v -u <host-machine> 12000                                                                                                                                   [±master ●●]
Connection to <host-machine> 12000 port [udp/*] succeeded!