MetalLB Bare Metal Load Balancer

💡 Quick Answer: Install MetalLB and configure an IPAddressPool with your available IPs. Choose L2 mode for simple setups (ARP-based, single node handles traffic) or BGP mode for production (announces routes to your network router, true load distribution).

The Problem

On bare-metal Kubernetes, type: LoadBalancer Services stay in Pending forever — there’s no cloud provider to create a load balancer. MetalLB fills this gap, providing LoadBalancer IP allocation for bare-metal, on-premises, and homelab clusters.

The Solution

Install MetalLB

helm repo add metallb https://metallb.github.io/metallb
helm install metallb metallb/metallb --namespace metallb-system --create-namespace

L2 Mode (Simple)

apiVersion: metallb.io/v1beta1
kind: IPAddressPool
metadata:
  name: production-pool
  namespace: metallb-system
spec:
  addresses:
    - 192.168.1.200-192.168.1.220
---
apiVersion: metallb.io/v1beta1
kind: L2Advertisement
metadata:
  name: production
  namespace: metallb-system
spec:
  ipAddressPools:
    - production-pool

BGP Mode (Production)

apiVersion: metallb.io/v1beta2
kind: BGPPeer
metadata:
  name: router
  namespace: metallb-system
spec:
  myASN: 64500
  peerASN: 64501
  peerAddress: 192.168.1.1
---
apiVersion: metallb.io/v1beta1
kind: BGPAdvertisement
metadata:
  name: production
  namespace: metallb-system
spec:
  ipAddressPools:
    - production-pool
  peers:
    - router

Use LoadBalancer Service

apiVersion: v1
kind: Service
metadata:
  name: web-app
  annotations:
    metallb.universe.tf/address-pool: production-pool
spec:
  type: LoadBalancer
  selector:
    app: web-app
  ports:
    - port: 80
      targetPort: 8080

graph TD
    subgraph L2 Mode
        CLIENT1[Client] -->|ARP| LEADER[Leader Node<br/>handles all traffic]
        LEADER --> POD1[Pod 1]
        LEADER --> POD2[Pod 2]
    end
    subgraph BGP Mode
        CLIENT2[Client] -->|BGP routes| ROUTER[Network Router]
        ROUTER --> N1[Node 1 → Pod A]
        ROUTER --> N2[Node 2 → Pod B]
        ROUTER --> N3[Node 3 → Pod C]
    end

Common Issues

Service still Pending after MetalLB install: Check IP pool has available addresses: kubectl get ipaddresspool -n metallb-system. Ensure the IP range is on the same subnet as your nodes (L2 mode).

L2 mode: all traffic goes to one node: Expected behavior. L2 mode uses ARP — one node handles all traffic. For true load distribution, use BGP mode.

Best Practices

• L2 for homelab/dev — simple, no router config needed
• BGP for production — true load distribution across nodes
• Separate pools for different service tiers
• Don’t overlap pools with DHCP range — causes IP conflicts
• Cilium can replace MetalLB — if already using Cilium, use its LB IPAM

Key Takeaways

• MetalLB provides LoadBalancer Services on bare-metal Kubernetes
• L2 mode is simple (ARP) but sends all traffic through one node
• BGP mode distributes traffic across nodes via router integration
• IPAddressPool defines available IPs; Advertisement controls how they’re announced
• Essential for on-premises, homelab, and edge Kubernetes deployments