Amazon EKS — Project 4 — High Availability Cluster

1 minute read

Project 4 — High Availability EKS Cluster

Build a production-grade cluster that survives AZ failures, node crashes, and traffic spikes.

HA Architecture

Multi-AZ HA Cluster
┌──────────────────────────────────────────────────────┐
│  Region: us-east-1                                   │
│                                                      │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐ │
│  │   AZ-1a     │  │   AZ-1b     │  │   AZ-1c     │ │
│  │             │  │             │  │             │ │
│  │ ┌─────────┐ │  │ ┌─────────┐ │  │ ┌─────────┐ │ │
│  │ │ Node 1  │ │  │ │ Node 3  │ │  │ │ Node 5  │ │ │
│  │ │ Node 2  │ │  │ │ Node 4  │ │  │ │ Node 6  │ │ │
│  │ └─────────┘ │  │ └─────────┘ │  │ └─────────┘ │ │
│  │             │  │             │  │             │ │
│  │ ┌─────────┐ │  │ ┌─────────┐ │  │ ┌─────────┐ │ │
│  │ │ EBS Vol │ │  │ │ EBS Vol │ │  │ │ EBS Vol │ │ │
│  │ └─────────┘ │  │ └─────────┘ │  │ └─────────┘ │ │
│  └─────────────┘  └─────────────┘  └─────────────┘ │
│                                                      │
│  ┌──────────────────────────────────────────────┐   │
│  │ EFS (shared storage — all AZs)               │   │
│  └──────────────────────────────────────────────┘   │
│                                                      │
│  ┌──────────────────────────────────────────────┐   │
│  │ RDS Multi-AZ (automatic failover)            │   │
│  └──────────────────────────────────────────────┘   │
│                                                      │
└──────────────────────────────────────────────────────┘

Step 1: Cluster with Multi-AZ

# cluster-ha.yaml
apiVersion: eksctl.io/v1alpha5
kind: ClusterConfig
metadata:
  name: ha-cluster
  region: us-east-1
  version: "1.29"
availabilityZones:
  - us-east-1a
  - us-east-1b
  - us-east-1c
managedNodeGroups:
  - name: critical
    instanceType: m5.xlarge
    desiredCapacity: 6
    minSize: 3
    maxSize: 12
    availabilityZones:
      - us-east-1a
      - us-east-1b
      - us-east-1c

Step 2: Pod Anti-Affinity (Spread Across AZs)

spec:
  affinity:
    podAntiAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:
        - weight: 100
          podAffinityTerm:
            labelSelector:
              matchExpressions:
                - key: app
                  operator: In
                  values: ["web"]
            topologyKey: topology.kubernetes.io/zone

Step 3: PodDisruptionBudgets

apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: web-pdb
spec:
  minAvailable: "50%"
  selector:
    matchLabels:
      app: web

Step 4: Health Checks

livenessProbe:
  httpGet:
    path: /health
    port: 8080
  initialDelaySeconds: 30
  periodSeconds: 10
  failureThreshold: 3
readinessProbe:
  httpGet:
    path: /ready
    port: 8080
  initialDelaySeconds: 5
  periodSeconds: 5

What You Practiced

  • Multi-AZ node distribution
  • Pod anti-affinity for zone spreading
  • PodDisruptionBudgets for safe maintenance
  • Health checks for automatic recovery
  • Auto-scaling for demand handling

What’s Next?

Save money: Project 5 — Cost Optimization.

You may also enjoy