On this page
On this page
- What Is Route 53
- Record Types
- Alias Records vs CNAME
- Hosted Zones
- Public Hosted Zone
- Private Hosted Zone
- Split-Horizon DNS
- TTL (Time to Live)
- Routing Policies
- Simple
- Weighted
- Latency-Based
- Failover
- Geolocation
- Geoproximity
- Multivalue Answer
- IP-Based
- Health Checks
- Endpoint Health Checks
- CloudWatch Alarm Health Checks
- Calculated Health Checks
- Route 53 Resolver (Hybrid DNS)
- Inbound Endpoint — On-Premises Resolves an AWS Private Hostname
- Outbound Endpoint — EC2 Resolves an On-Premises Hostname
- Resolver Rules
- Key Exam Scenarios
- CloudWatch Integration
- Automatically Published (AWS/Route53 namespace)
- Requires Setup
Route 53
Route 53 is AWS's DNS (Domain Name System) service. It translates human-readable
domain names like api.mycompany.com into IP addresses that computers use to
connect to your servers. It also does much more than basic DNS — it can route
traffic intelligently based on geography, latency, health, and weights.
DNS in Plain Terms
When you type a URL into a browser, your computer needs to find the IP address of the server hosting that site. It asks a DNS resolver, which consults a hierarchy of DNS servers until it finds the answer. Route 53 is the authoritative DNS server for your domains — the final authority that answers "what is the IP address for api.mycompany.com?"
Routing Policies
Route 53's power comes from its routing policies, which determine which answer to return when a DNS query arrives:
Simple — return one IP address. The default, no-frills option.
Weighted — split traffic between multiple endpoints by percentage. Send 90% of traffic to your stable deployment and 10% to a new version for canary testing.
Latency-based — Route 53 measures latency from the requester to each of your endpoints in different regions and routes to the one with the lowest latency. Users in Tokyo get routed to your Tokyo endpoint; users in Paris to your European one.
Failover — designate a primary and secondary endpoint. Route 53 monitors the primary with health checks and automatically fails over to the secondary if the primary becomes unhealthy.
Geolocation — route based on the country or continent the query comes from. Send European users to servers in eu-west-1, Asian users to ap-southeast-1.
Geoproximity — like geolocation, but lets you shift traffic toward or away from specific regions using a "bias" value.
Simple = "just tell me the IP" · Weighted = "what percentage?" · Latency = "what's fastest?" · Failover = "is the primary alive?" · Geolocation = "where is the user FROM?" (their country) · Geoproximity = "what's NEAR the user?" (distance + bias dial) · Multivalue = "give me several healthy options" · IP-based = "which network is the user on?"
Health Checks
Route 53 can monitor your endpoints (HTTP, HTTPS, TCP) and mark them healthy or unhealthy. This feeds into failover and other routing policies. You can also create calculated health checks that aggregate the health of multiple underlying checks — useful for "consider this endpoint healthy only if at least 2 of 3 instances are healthy."
Private Hosted Zones
Route 53 can serve DNS internally within a VPC using private hosted zones.
Your services can address each other as orders.internal instead of by private
IP, making the architecture more readable and easier to change.