Serious Lab Only (No Bulto Allowed)

Aquí no se viene a inventar… esto es serio, pana mio

---

⚙️ Overview

This lab is designed for real testing, real configs, and real consequences. No shortcuts, no guesswork, no bulto allowed.

---

Rules of Engagement

• Always snapshot before making changes
• Test before pushing anything to “prod-like”
• Document what you break (and hopefully fix)
• If it works… verify again 😄

---

What’s Not Allowed

• Random configs without validation
• “Works on my machine” excuses
• Skipping backups
• Bulto. Ninguno.

---

✅ Goal

Build, break, fix, and learn the right way.

---

Status: Active Mode: Diablo 😈

---

Stack

Component	Role
Ubuntu 26.04 LTS	Guest OS, provisioned via cloud-init
k3s	Lightweight Kubernetes (single-node)
Prometheus + Alertmanager	Metrics collection and alerting
Grafana	Dashboards and visualization
Loki + Promtail	Log aggregation
NGINX Ingress	HTTP routing by hostname
MetalLB	LoadBalancer IP assignment (bare-metal)

---

Host requirements

• OS: CachyOS or any Arch-based Linux with KVM support (You can use whatever distro you want… but let’s be real, use Arch, mi pana :V 😄)
• CPU: 4+ cores, VT-x or AMD-V enabled in BIOS
• RAM: 16 GB (8 GB assigned to the VM, 8 GB for the host)
• Disk: 100 GB free on SSD/NVMe at /var/lib/libvirt/images/

---

Directory structure

---

Deployment

1. Install host dependencies

sudo pacman -S qemu-full libvirt virt-install edk2-ovmf bridge-utils
sudo systemctl enable --now libvirtd
sudo usermod -aG kvm,libvirt $USER && newgrp libvirt

2. Create a network bridge

The VM needs a bridged interface to get a real IP on your LAN. Replace enp3s0 with your actual NIC name (ip link show to find it).

nmcli con add type bridge ifname br0 con-name br0
nmcli con add type bridge-slave ifname enp3s0 master br0
nmcli con modify br0 bridge.stp no
nmcli con up br0

3. Configure cloud-init

Edit vm/cloud-init.yaml and replace the two placeholders

4. Validate and deploy

bash scripts/validate.sh      # fix any FAIL items first
bash vm/deploy.sh             # (image download + provisioning)

5. Access the cluster

VM_IP=$(virsh domifaddr k8s-lab | awk '/ipv4/{print $4}' | cut -d/ -f1)
mkdir -p ~/.kube
scp labuser@${VM_IP}:/home/labuser/.kube/config ~/.kube/k8s-lab.yaml
sed -i "s/127.0.0.1/${VM_IP}/" ~/.kube/k8s-lab.yaml
export KUBECONFIG=~/.kube/k8s-lab.yaml
kubectl get nodes

---

Installing the monitoring stack

# SSH into the VM
ssh labuser@<vm-ip>

# Install NGINX Ingress Controller
helm install ingress-nginx ingress-nginx/ingress-nginx \
  --namespace ingress-nginx --create-namespace \
  --values kubernetes/ingress/nginx-values.yaml

# Install MetalLB — edit the IP range first (see kubernetes/networking/metallb-pool.yaml)
helm install metallb metallb/metallb \
  --namespace metallb-system --create-namespace
kubectl -n metallb-system wait pod \
  --for=condition=Ready -l app.kubernetes.io/name=metallb --timeout=90s
kubectl apply -f kubernetes/networking/metallb-pool.yaml

# Get the IP MetalLB assigned, then configure DNS on your host
kubectl -n ingress-nginx get svc ingress-nginx-controller
bash scripts/setup-dns.sh 192.168.1.200   # replace with actual IP

# Install Prometheus, Grafana, and Alertmanager
kubectl create namespace monitoring
helm install kube-prometheus-stack prometheus-community/kube-prometheus-stack \
  --namespace monitoring \
  --values kubernetes/monitoring/stack-values.yaml

# Expose Grafana and Prometheus in the browser
kubectl apply -f kubernetes/monitoring/grafana-ingress.yaml
kubectl apply -f kubernetes/monitoring/prometheus-ingress.yaml

Open http://grafana.lab — login: admin / bolecajuil123 Open http://prometheus.lab

---

Snapshots

Always snapshot before making changes you might want to undo… future you will either thank you, or file a ticket against past you 😄

bash scripts/snapshot.sh create  "before-monitoring"
bash scripts/snapshot.sh list
bash scripts/snapshot.sh revert  "before-monitoring-20260503-1430"

---