Route53
Route53 is AWS's highly available, globally distributed DNS service. It also provides domain registration, health checks, and traffic routing policies for directing user traffic across regions and endpoints.
Route53 is a traffic cop at a city intersection — they can direct cars by shortest path (latency), by neighborhood (geolocation), or weighted between two roads (weighted). They also check if a road is open (health checks) before directing anyone onto it.
Routing policies: Simple (single resource), Weighted (A/B testing, gradual migration), Latency-based (route to lowest-latency region), Failover (active-passive with health checks), Geolocation (by country/continent), Geoproximity (by distance with bias), Multi-value (up to 8 healthy records), IP-based. Health checks: monitor endpoints every 10 or 30 seconds, trigger failover automatically. Alias records point to AWS resources (ALB, CloudFront, S3 website) with no TTL cost — unlike CNAMEs, Alias records work at the zone apex (naked domain).
Route53 Resolver handles DNS resolution within VPCs and between VPCs and on-prem via Resolver Endpoints. Route53 Resolver DNS Firewall blocks DNS queries to known malicious domains. Private Hosted Zones associate DNS records with VPCs — internal-only DNS. DNS failover: when a health check fails, Route53 automatically routes traffic to the standby endpoint within seconds (TTL-dependent). Route53 Traffic Flow provides a visual policy editor for complex multi-policy routing. DNSSEC provides cryptographic signing of DNS records to prevent DNS spoofing. Latency routing uses AWS's internal network latency data, not geographic distance — a closer region isn't always the lowest latency pick.
Route53 handles DNS with smart routing policies. For multi-region active-active, I use Latency routing with health checks so traffic automatically shifts if one region degrades. For gradual deployments, Weighted routing lets me send 10% to new versions. Alias records for AWS resources avoid CNAME costs and work at the zone apex. Failover routing for active-passive DR automatically promotes the standby when health checks fail.
Using latency routing and expecting it to route by geographic proximity. Route53 uses actual measured network latency to AWS regions, not straight-line geographic distance. A geographically closer region can have higher measured latency.