Have you ever wondered why an OTT subscription registered in India won’t work in the USA or why a soccer game live in Europe isn’t available in Singapore?

The answer often lies in sophisticated DNS routing policies designed to deliver content efficiently, securely, and sometimes legally.

Let’s explore six key DNS routing policies that power global-scale applications and their real-world applications, benefits, and limitations.

Download a high-resolution PDF of this diagram here.

1. Simple Routing

DNS maps a domain (e.g., app.techops.com) to a resource like an EC2 instance, load balancer, or Kubernetes service, resolving to a single IP.

It uses A/AAAA records for static IPs or CNAME records for dynamic resources like ALBs.
If the resource fails, traffic halts unless manually rerouted.

2. Weighted Routing

Multiple resources are assigned weights, and DNS resolves requests proportionally to these weights. Example: 70% to v1.techops.com, 30% to v2.techops.com

Each DNS entry gets a percentage weight.
If all resources have weight=0, traffic is evenly distributed

3. Failover Routing

Configures a primary and secondary resource for disaster recovery (DR). Traffic routes to the primary unless health checks detect a failure, in which case it redirects to the secondary.

Monitors HTTP(S), TCP, or even custom endpoint responses for health checks.
DNS TTL impacts failover speed; typical values range from 30–60 seconds.

4. Latency Based Routing

Routes users to the resource with the lowest network latency. Latency is dynamically calculated using regional edge locations and network topology.

DNS resolver queries latency maps updated by the DNS provider (e.g., Route 53 or Akamai).
Relies on GeoIP databases and BGP routing data.

5. Geolocation Routing

Routes users based on their IP location to pre-defined backend regions.

Uses GeoIP databases to map user IPs to locations.
Pre-configured locations specify which region serves traffic.
Supports default routing for unmatched locations.

6. Geoproximity Routing

Routes users to the nearest resource while adjusting traffic distribution with a bias factor.

Bias ranges from -99 to 99:
- Positive bias sends more traffic to a resource.
- Negative bias sends less traffic to a resource.
Geo-proximity calculations use lat/lon coordinates of resources and user IPs.

Pro Tip: When implementing these policies, always monitor real-world traffic patterns and tweak TTL, weights, and biases dynamically for optimal results.

Looking to promote your company, product, service, or event to 30,000+ TechOps Professionals? Let's work together.