Amazon Route 53
A highly available Domain Name System (DNS) service that provides integrated domain registration, DNS routing, and health checking
Overview
Amazon Route 53 is a highly available and scalable cloud DNS service. It provides integrated domain name registration, DNS record management, traffic routing, and endpoint health checking. As one of the few AWS services with a 100% availability SLA, it delivers low-latency name resolution through a globally distributed network of DNS servers. It offers a variety of routing policies including simple, weighted, latency-based, failover, geoproximity, and multivalue answer routing, enabling flexible traffic distribution and disaster recovery configurations. Alias records are an AWS-specific feature that lets you map AWS resources such as CloudFront, ELB, and S3 directly to the Zone Apex (naked domain).
Alias Records and Zone Apex Handling
Route 53's alias records are a proprietary feature that lets you map AWS resources directly to the Zone Apex (naked domain, e.g. example.com), where DNS specifications prohibit setting a CNAME. You can specify CloudFront distributions, ALBs, and S3 static website hosting endpoints as alias targets, and DNS query charges are waived for alias records. Azure DNS lacks this mechanism, so for Zone Apex records you need to combine Azure Traffic Manager or Azure Front Door as a workaround. Unlike CNAMEs, alias records do not include intermediate CNAME chains in the DNS response, reducing name resolution latency. Behind Route 53's 100% availability SLA is a globally distributed DNS server network combined with anycast routing that automatically distributes queries.
Routing Policy Combination Strategies
Route 53 offers seven routing policies that can be combined for sophisticated traffic management. Weighted routing is effective for canary releases, directing just 10% of traffic to a new version. Latency-based routing directs users to the nearest region's resources, optimizing response times for global applications. Failover routing works with health checks to automatically switch to a secondary resource when the primary is unhealthy. Geolocation routing controls traffic based on the user's country, addressing data sovereignty and content localization requirements. These policies can be nested - for example, you can use latency-based routing to select the nearest region, then apply weighted routing within that region for canary deployments. Specialized books (Amazon) cover practical routing policy combination patterns in detail.
Health Checks and Failure Detection
Route 53 health checks periodically monitor endpoint availability from checkers around the world, and when an anomaly is detected, traffic is automatically rerouted based on the configured routing policy. Three protocol types are supported - HTTP, HTTPS, and TCP - and HTTP/HTTPS checks can include response body string matching. The check interval defaults to 30 seconds, with a fast option that reduces it to 10 seconds. Calculated health checks that link CloudWatch alarm states to health check results enable routing decisions based on application-specific metrics such as CPU utilization or error rates. In a failover configuration, health checks are set on both primary and secondary resources, routing to the secondary only when the primary becomes unhealthy. Health check results are published as CloudWatch metrics, so combining them with SNS notifications lets you build an immediate failure detection system.