What is AWS Route 53 and how do you implement DNS failover?

Amazon Route 53 is a scalable and highly available DNS web service that routes end users to internet applications and supports domain registration.

Key Features

DNS Resolution

Route 53 translates domain names (example.com) into IP addresses. It supports all standard DNS record types: A, AAAA, CNAME, MX, TXT, NS, SOA, and Route 53-specific alias records.

Routing Policies

• Simple: Route traffic to a single resource
• Weighted: Split traffic by percentage between resources (A/B testing, gradual rollouts)
• Latency: Route to the region with lowest network latency
• Geolocation: Route based on user’s geographic location
• Geoproximity: Route based on geographic location with configurable bias
• Failover: Active-passive failover routing
• Multivalue Answer: Responds with up to 8 healthy records

Implementing DNS Failover

Active-Passive Failover Setup

1. Create Health Checks

• Configure health checks for your primary endpoint (HTTP/HTTPS/TCP)
• Set evaluation period, failure threshold, and interval

1. Create Primary Record

   Type: A
   Routing Policy: Failover
   Failover Type: Primary
   Health Check: my-primary-health-check
   TTL: 60

3. Create Secondary Record

   Type: A
   Routing Policy: Failover
   Failover Type: Secondary
   Value: [backup IP or S3 static site]
   TTL: 60

4. Failover Behavior

• If primary health check fails, Route 53 routes to secondary
• When primary recovers, traffic automatically returns

Active-Active Failover

Use Weighted routing with health checks:

• Both endpoints active with equal weight (50/50)
• Route 53 automatically removes unhealthy endpoints
• Traffic redistributes to healthy endpoints

Multi-Region Failover Pattern

Route 53 (Latency routing)
├── us-east-1 ALB (Primary)
│   └── Auto Scaling Group
└── eu-west-1 ALB (Failover)
    └── Auto Scaling Group

Health Check Types

• Endpoint health checks: HTTP/HTTPS/TCP checks on IP or domain
• Calculated health checks: Combine results of multiple health checks
• CloudWatch alarm health checks: Based on CloudWatch alarm state