AWS Spot Instances: A Practical Guide

Spot Instances: 60-90% Savings

AWS Spot Instances let you use spare EC2 capacity at up to 90% discount compared to On-Demand pricing. The catch? AWS can reclaim them with a 2-minute warning. But with the right architecture, Spot is safe — and incredibly cost-effective.

What Workloads Are Spot-Safe?

Best Practices for Production Spot

1. Diversify instance types — Request capacity across 5-10 instance types and 3 AZs. This dramatically reduces interruption risk.
2. Use Spot Fleet or EC2 Auto Scaling mixed instances — Let AWS choose the cheapest available combination.
3. Handle interruptions gracefully — Monitor the instance metadata endpoint for the 2-minute warning and drain connections.
4. Mix with On-Demand — Use a 70/30 Spot/On-Demand ratio in Auto Scaling Groups for guaranteed baseline capacity.

Spot in Auto Scaling Groups

{
  "MixedInstancesPolicy": {
    "InstancesDistribution": {
      "OnDemandBaseCapacity": 2,
      "OnDemandPercentageAboveBaseCapacity": 30,
      "SpotAllocationStrategy": "capacity-optimized"
    },
    "LaunchTemplate": {
      "Overrides": [
        {"InstanceType": "m5.large"},
        {"InstanceType": "m5a.large"},
        {"InstanceType": "m6i.large"},
        {"InstanceType": "m6a.large"},
        {"InstanceType": "m5.xlarge", "WeightedCapacity": "2"}
      ]
    }
  }
}

Handling Interruptions

#!/bin/bash
# Check for Spot interruption notice
while true; do
  HTTP_CODE=$(curl -s -o /dev/null -w "%{http_code}" \
    http://169.254.169.254/latest/meta-data/spot/instance-action)
  if [ "$HTTP_CODE" -eq 200 ]; then
    echo "Spot interruption received! Draining..."
    # Deregister from ALB, drain connections, save state
    exit 0
  fi
  sleep 5
done

Eazy SaaS Tip: Start with CI/CD and dev environments. Once comfortable, expand Spot to stateless web tiers behind Auto Scaling Groups. Most clients achieve 60-70% average savings across their fleet.