Route 53 – Health Checks

Overview

• Allow Route 53 to automatically fail over to backup endpoints.

• Only directly possible for public endpoints — for private resources, CloudWatch integration is required.

1. Endpoint Monitoring — checks an application or service via HTTP/HTTPS/TCP.

2. Calculated Health Checks — combine results of multiple health checks using logical AND/OR/NOT.

3. CloudWatch Alarm-based — monitors the state of a CloudWatch alarm (for custom/private checks).

How Endpoint Health Checks Work

• About 15 global health checkers probe the endpoint.

• Thresholds: default is 3 consecutive failed/passed checks.

• Interval: 30s (default) or 10s (higher cost).

• Healthy if >18% of checkers report it healthy.

• Pass conditions: HTTP 2xx or 3xx, or matching body content in first 5120 bytes.

Calculated Health Checks

• Combine up to 256 child health checks.

• Can define a threshold (e.g., require 3 of 5 checks to be healthy).

• Useful for avoiding full failover during maintenance or partial outages.

Health Checks in Private Hosted Zones

• Route 53 health checkers cannot reach private endpoints directly.

• Workaround:
• Create a CloudWatch metric (from Lambda, EC2 agent, etc.).
• Attach a CloudWatch alarm.
• Use a Route 53 health check that monitors that alarm.