Skip to content

tbernacchi/home-lab

BranchesTags

Repository files navigation

home-lab

This is my personal Kubernetes setup for my home-lab running on my Raspberry Pi4 cluster.

Core Components

  • K3s - Lightweight Kubernetes distribution perfect for IoT & Edge computing
  • Cilium - eBPF-based networking, observability & security

Ingress Controller

  • Traefik - Cloud native ingress controller for handling incoming traffic and routing requests

Monitoring

GitOps

Database

Hand's on

For this setup I've disable wlan0 interface and use only eth0 for performance reasons. I also disable brcmfmac_wcc, brcmfmac, brcmutil and cfg80211 modules to avoid the wlan0 interface to be used.

modprobe -r brcmfmac_wcc
modprobe -r brcmfmac
modprobe -r brcmutil
modprobe -r cfg80211

echo "blacklist brcmfmac_wcc" > /etc/modprobe.d/blacklist-brcmfmac.conf
echo "blacklist brcmfmac" >> /etc/modprobe.d/blacklist-brcmfmac.conf
shutdown -r now

k3s

export K3S_KUBECONFIG_MODE="644"
export INSTALL_K3S_EXEC=" --flannel-backend=none --disable-network-policy --disable servicelb --disable traefik" 
curl -sfL https://get.k3s.io | sh -

After everything it's up and running I've changed k3s to use etcd as a data store.

I've added --cluster-init on /etc/systemd/system/k3s.service

ExecStart=/usr/local/bin/k3s \
    server \
        '--cluster-init' \
	'--flannel-backend=none' \
	'--disable-network-policy' \
	'--disable' \
	'servicelb' \
	'--disable' \
	'traefik' \
systemctl daemon-reload
systemctl restart k3s

cluster-init
https://k3s.io/

Join nodes to the cluster

To install additional agent nodes and add them to the cluster, run the installation script with the K3S_URL and K3S_TOKEN environment variables. Here is an example showing how to join an agent:

curl -sfL https://get.k3s.io | K3S_URL=https://myserver:6443 K3S_TOKEN=mynodetoken sh -

Setting the K3S_URL parameter causes the installer to configure K3s as an agent, instead of a server. The K3s agent will register with the K3s server listening at the supplied URL. The value to use for K3S_TOKEN is stored at /var/lib/rancher/k3s/server/node-token on your server node.

https://docs.k3s.io/quick-start

Cilium

helm repo add cilium https://helm.cilium.io/
helm repo update
helm upgrade --install cilium cilium/cilium --version v1.15.6 \
  --set operator.replicas=1 \
  --set ipam.operator.clusterPoolIPv4PodCIDRList=10.42.0.0/16 \
  --set ipv4NativeRoutingCIDR=10.42.0.0/16 \
  --set ipv4.enabled=true \
  --set loadBalancer.mode=dsr \
  --set routingMode=native \
  --set autoDirectNodeRoutes=true \
  --set l2announcements.enabled=true \
  --set kubeProxyReplacement=true \
  --set k8sClientRateLimit.qps=50 \
  --set k8sClientRateLimit.burst=100 \
  --set k8sServiceHost=192.168.1.106 \
  --set k8sServicePort=6443 \
  --set l2announcements.leaseDuration=3s \
  --set l2announcements.leaseRenewDeadline=1s \
  --set l2announcements.leaseRetryPeriod=200ms \
  --set ingressController.Enabled=true \
  --set enable-bgp-control-plane.enabled=true
  • Be attention to your k8sServicePort, which it's the interface advertised from your k3s.
kubectl edit cm -n kube-system cilium-config
bpf-lb-sock-hostns-only: "true"
enable-host-legacy-routing: "true"
device: eth0
enable-bpf-masquerade: "true"
kubectl -n kube-system rollout restart ds/cilium
kubectl create -f cilium/CiliumL2AnnouncementPolicy-IPPool.yaml

https://docs.cilium.io/en/stable/installation/k8s-install-helm/

Certs

Run certificate.sh on certs folder.

./certificate.sh
Certificate request self-signature ok
subject=C = BR, ST = SP, L = Sao Paulo, O = MyKubernetes, CN = traefik.mykubernetes.com
secret/traefik-dashboard-cert created

Add your ca.crt to the system keychain. If you are using macOS:

First time installation

sudo security add-trusted-cert -d -r trustRoot -k /Library/Keychains/System.keychain ./ca.crt

Updating certificate (if already exists)

If you need to update the certificate and there are already old certificates in the keychain:

  1. Find existing certificates:
security find-certificate -a -c "MyKubernetes CA" -Z /Library/Keychains/System.keychain | grep "SHA-1 hash" | awk '{print $3}'
  1. Remove old certificates by SHA-1 hash:
sudo security delete-certificate -Z <SHA-1_HASH> /Library/Keychains/System.keychain

Repeat for each old certificate hash found in step 1.

  1. Add the new certificate:
sudo security add-trusted-cert -d -r trustRoot -k /Library/Keychains/System.keychain ./ca.crt
  1. Verify the certificate was added:
security find-certificate -c "MyKubernetes CA" /Library/Keychains/System.keychain -Z | grep "SHA-1 hash"

The SHA-1 hash should match your new certificate:

openssl x509 -in ca.crt -noout -fingerprint -sha1 | cut -d= -f2 | tr ':' ' ' | tr -d ' '
  1. Clear browser HSTS cache (Chrome):
  • Open chrome://net-internals/#hsts
  • In "Delete domain security policies", enter: traefik.mykubernetes.com
  • Click "Delete"
  • Or clear all browser cache and restart the browser

Monitoring

helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm repo update
kubectl create namespace monitoring
helm install my-kube-prometheus-stack prometheus-community/kube-prometheus-stack --version 68.3.0 -n monitoring -f monitoring/prometheus-values.yaml

Setting my monitoring solution to reach at /prometheus, /grafana and /alertmanager.

See monitoring/README.md for more details.

# Add the secret to the monitoring namespace
kubectl get secret traefik-dashboard-cert -n traefik -o yaml | sed 's/namespace: traefik/namespace: monitoring/' | kubectl apply -f -

# Create the ingressroute for the monitoring namespace
kubectl create -f monitoring/ingressroute.yaml

Traefik

kubectl create namespace traefik
helm repo add traefik https://traefik.github.io/charts
helm repo update
helm install traefik traefik/traefik --namespace traefik --values traefik/values.yaml
kubectl create -f traefik/dashboard.yaml

https://doc.traefik.io/traefik/getting-started/install-traefik/#use-the-helm-chart
https://github.com/traefik/traefik-helm-chart/blob/master/traefik/values.yaml

argo-workflow

kubectl create namespace argo
kubectl apply -n argo -f https://github.com/argoproj/argo-workflows/releases/download/v3.5.8/install.yaml

In this setup I've changed the Base_HREF to /argo/ to be able to reach the workflows UI at /argo/.

kubectl edit deploy/argo-server -n argo
- args: 
  - server
  - --auth-mode=server
  env:
  - name: BASE_HREF
    value: /argo/

Argo Workflows need a service account in the respective namespace where the workloads it's going to run order to work properly. This service account needs some permissions to manage workflows, interact with pods and etctera. You can find more info here.

kubectl get secret traefik-dashboard-cert -n traefik -o yaml | sed 's/namespace: traefik/namespace: argo/' | kubectl apply -f -
kubectl create -f argo/rbac.yaml

argo-cd

To install argo-cd I've followed the https://argo-cd.readthedocs.io/en/stable/getting_started/.

kubectl create namespace argocd
kubectl apply -n argocd -f https://raw.githubusercontent.com/argoproj/argo-cd/stable/manifests/install.yaml

For this installation I've managed the UI through Traefik to access on traefik.mykubernetes.com/argocd.

kubectl patch deployment argocd-server -n argocd --type='json' -p='[{"op":"replace","path":"/spec/template/spec/containers/0/args","value":["/usr/local/bin/argocd-server","--insecure","--basehref=/argocd","--rootpath=/argocd"]}]'
kubectl get deploy argocd-server -n argocd -o jsonpath='{.spec.template.spec.containers[0].args}' | jq
kubectl create -f argo-stack/argo-cd/ingressroute.yaml

Reference: https://argo-cd.readthedocs.io/en/latest/operator-manual/ingress/

UI password:

kubectl get secret/argocd-initial-admin-secret -n argocd -o jsonpath='{.data.password}' | base64 --decode

-> argo-cd has an issue using basehref with /argocd;

Bug fix:

kubectl apply -n argocd -f https://raw.githubusercontent.com/argoproj/argo-cd/v2.14.10/manifests/install.yaml

-> Don't forget to patch the deploment after the upgrade.

argo-rollouts

To enable the argo-rollouts on the UI I've use this extension: https://github.com/argoproj-labs/rollout-extension

kubectl create namespace argo-rollouts
helm repo add argo-rollouts https://argoproj.github.io/argo-helm -n argo-rollouts
helm repo update
helm install argo-rollouts argo/argo-rollouts -n argo-rollouts --set dashboard.enabled=true
kubectl patch deployment argocd-server -n argocd --patch "$(cat argo-stack/argo-rollouts/patch-argocd-server.yaml)"

argo-image-updater

kubectl apply -n argocd -f https://raw.githubusercontent.com/argoproj-labs/argocd-image-updater/master/manifests/install.yaml

We've to create a secret from Docker Hub registry to store the credentials for the image updater.

kubectl create secret docker-registry regcred \
  --namespace argocd \
  --docker-server=https://index.docker.io/v1/ \
  --docker-username=ambrosiaaaaa\
  --docker-password=token-xyz\
  --docker-email=myemail@gmail.com

Also we need a credentials for the GitHub registry.

kubectl create secret generic git-creds \
  -n argocd \
  --from-literal=username=myusername \
  --from-literal=password=token-xyz

Now we just need to "annotate" the application with the image updater.

kubectl get application -n argocd
NAME     SYNC STATUS   HEALTH STATUS
foobar   Synced        Healthy

kubectl annotate application foobar \
  argocd-image-updater.argoproj.io/credentials="docker.io=secret:dockerhub-secret" \
  argocd-image-updater.argoproj.io/image-list="ambrosiaaaaa/foobar-api" \
  argocd-image-updater.argoproj.io/update-strategy="semver" \
  argocd-image-updater.argoproj.io/write-back-method="git:secret:argocd/git-creds" \
  -n argocd

Argo CD Image Updater
Update Strategies
Git Write-back Method
Exemples

About

This is a repo to store all my settings from my home-lab.

Resources

Readme
Activity

Stars

1 star

Watchers

1 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages