Cloud concept · Updated

Latency-Based Routing - Latency-based routing directs clients to an eligible endpoint expected to provide the lowest network delay.

In short

Latency-based routing selects among multiple application endpoints using measured or estimated network latency rather than geographic distance or fixed traffic percentages. It helps distributed applications reduce connection and response times, but its effectiveness depends on measurement quality, endpoint health, DNS behavior, and the application's data architecture.

Cloud Concepts Hub →

6 key aspects of Latency-Based Routing

Effective latency-based routing combines a suitable latency model, healthy regional endpoints, predictable DNS behavior, and an application architecture that supports regional execution.

01
Selection by expected network delay

Endpoints are ranked using measured or estimated latency rather than physical distance alone. Verify what the routing platform measures and whether that model represents the actual client-to-endpoint path.

02
DNS granularity and caching

DNS-based decisions occur during name resolution and may be shared by many clients behind a recursive resolver. Cached answers and persistent connections prevent routing changes from taking effect immediately.

03
Health-aware endpoint eligibility

Health checks determine whether an endpoint should participate, while latency determines which eligible endpoint is preferred. Review all-unhealthy behavior and use checks that represent application readiness.

04
Distinct traffic-management objectives

Latency routing optimizes expected delay; geographic routing enforces location rules, and weighted routing controls response proportions. Combine policies only when the platform supports the required hierarchy and its behavior is understood.

05
Hybrid DNS path awareness

Enterprise forwarding, shared resolvers, and private connectivity can change how a client's location is represented. Test endpoint selection from on-premises networks and each participating cloud account rather than relying only on public DNS tests.

06
Application and data readiness

Every selected region must have suitable capacity, application state, and access to required data. Evaluate end-to-end response times because lower network latency can be offset by cross-region database calls or replication constraints.

Cloud services for Latency-Based Routing

Amazon Route 53 and Azure Traffic Manager apply latency-oriented selection at the DNS layer, returning a destination that clients subsequently contact directly. They are typically alternatives for authoritative traffic management rather than services that must be chained. AWS Global Accelerator offers a different approach for supported AWS endpoints: a stable anycast frontend with network-level regional steering. In any design, regional applications and load balancers must expose useful health signals and sufficient capacity. Choose the traffic-management layer according to protocol support, DNS caching tolerance, network requirements, and operational complexity.

Amazon Route 53Amazon Route 53

Latency-based DNS records select among configured AWS Regions using Route 53's latency measurements and the query's available location information. Health checks or supported alias target-health evaluation can influence endpoint eligibility; DNS caching still limits how quickly clients observe changes.

AWS Global AcceleratorAWS Global Accelerator

Global Accelerator provides static anycast IP addresses and directs connections toward healthy regional endpoints using factors including client location, endpoint health, and configured traffic controls. It is a complementary performance-oriented approach rather than a Route 53 latency DNS policy, and it carries traffic over the AWS global network.

Latency-Based Routing exam questions

Real practice exam questions on Latency-Based Routing, with the correct answer and its explanation. Answer them with every option in the exam simulator.

SAA-C03

A company provides a stateless engineering application through Application Load Balancers (ALBs) in Europe and Asia Pacific. Amazon Route 53 has correctly configured latency alias records for the application's hostname. Both deployments are healthy, have comparable processing times, and can serve any employee.

Employees in London and Singapore resolve public DNS names through a single corporate recursive DNS resolver in London. The resolver does not send EDNS Client Subnet information. Singapore employees consistently receive the European ALB's address, even though network tests from Singapore show a faster path to the Asia Pacific deployment. Clearing client DNS caches does not change this behavior.

The company wants DNS routing decisions to reflect network conditions near each office while continuing to use AWS network-latency estimates to select a Region. Which change should a solutions architect recommend?

Show answerCORRECT ANSWER

Configure each office to use a recursive DNS resolver located in that office's geographic area for public DNS queries.

Without EDNS Client Subnet information, Route 53 bases latency routing on the recursive resolver's location. Using geographically local resolvers gives Route 53 a location more representative of each office, allowing its regional latency estimates to guide selection appropriately. It does not guarantee the fastest endpoint for every individual request.

SAP-C03

A company provides interactive image enhancement through a single public hostname. It has identical application stacks in three AWS Regions. Each stack includes an internet-facing Application Load Balancer (ALB), application servers, and an Amazon SageMaker AI real-time inference endpoint. Each stack has sufficient capacity to serve the entire workload, and requests require no shared session state or remote datasets.

Currently, the hostname resolves to a central proxy fleet in one Region. The proxies forward requests to the regional stacks and relay the responses. Large request and response payloads generate substantial inter-Region data transfer charges.

Testing shows that clients in the same country can experience different network performance to the regional stacks. The best-performing Region for a client network can also change over time, and the geographically closest Region is not always fastest. Model processing times and operating costs are equivalent across the stacks.

The company wants clients to connect directly to a responsive, healthy regional stack while eliminating routine inter-Region payload forwarding. It must retain the existing hostname and minimize routing administration. DNS caching and DNS-based failover convergence are acceptable.

Which solution best meets these requirements?

Show answerCORRECT ANSWER

Create Route 53 latency alias records for the hostname, associating each regional ALB with its AWS Region. Configure target-health evaluation and have each stack invoke its local SageMaker AI endpoint and return results directly to clients.

Route 53 uses its Region-associated latency measurements to select an eligible endpoint for the querying client network or resolver. Target-health evaluation helps exclude unhealthy ALBs. Clients then connect directly to the selected stack, which performs inference locally, eliminating the central proxy's routine inter-Region payload transfers. DNS selection is not a fresh measurement of every request or of model processing time.

ANS-C01

A company operates an application behind internet-facing Application Load Balancers (ALBs) in us-east-1, eu-west-1, and ap-southeast-1. Each deployment is in a separate AWS account and can serve any user. A central networking account owns the Route 53 public hosted zone for the application.

Employees access the application through local internet connections at offices worldwide. However, all offices forward public DNS queries to recursive resolvers at the company's US headquarters. These resolvers do not send EDNS Client Subnet information. Testing shows that offices usually receive the same regional endpoint, even when AWS network-latency estimates favor another Region for an office's network.

The company wants DNS answers to reflect AWS's estimated network delay for each office's network, avoid unhealthy ALBs, and require no manually maintained country-to-Region assignments. Which configuration best meets these requirements?

Show answerCORRECT ANSWER

Create latency alias records with the same application name, one for each regional ALB, specifying each ALB's Region and enabling Evaluate Target Health. Have offices resolve the public application name through branch-local recursive resolvers that use their local internet connections.

Latency records let Route 53 choose among eligible Regions using AWS latency measurements. Evaluate Target Health allows Route 53 to exclude unhealthy ALB targets when healthy alternatives exist. Without EDNS Client Subnet, the recursive resolver's source address represents the querying network, so branch-local resolution avoids making every office appear to query from headquarters. The central public hosted zone can contain aliases to ALBs in the other accounts.

Practice Latency-Based Routing in a full exam

Questions on Latency-Based Routing in each practice exam, with every answer option and an explanation for each:

Start a free practice exam

What is Latency-Based Routing

3 min read

Latency-based routing is a traffic-management policy that chooses an endpoint according to expected network delay from the client or a representative network location. It is commonly used for applications deployed in multiple cloud regions, where users can access equivalent instances of the same service. The objective is to improve responsiveness, not necessarily to choose the geographically nearest deployment or distribute traffic evenly.

In a DNS implementation, an authoritative DNS service selects an answer when a recursive resolver queries an application hostname. The decision may use historical network measurements, the resolver's location, or client-network information supplied through EDNS Client Subnet when supported. The client then connects directly to the returned endpoint; the DNS service does not proxy application traffic. This differs from connection-level steering through a global network frontend, which can select a destination without relying on a new DNS answer.

For example, an application might run in eastern North America and western Europe. European users would often receive the European endpoint, but Internet routing and peering can make a more distant region faster for some networks. The routing service generally uses its own measurement model rather than timing a fresh request from every individual client. In Amazon Route 53, latency records identify AWS Regions, and selection uses latency measurements associated with those Regions. This is not a direct measurement of each application's processing time.

Health and latency are separate concerns. A low-latency endpoint might have a failed application, an overloaded database, or insufficient capacity. Configure health checks or supported target-health evaluation so that routing can avoid unhealthy destinations, and understand the provider's behavior when every destination is unhealthy. Application readiness checks should reflect whether an endpoint can actually serve useful requests, not merely whether a TCP port accepts connections.

DNS caching limits how quickly routing changes reach clients. Recursive resolvers and applications can reuse an answer until its cache lifetime expires, and existing connections normally continue using their original destination. Shorter time-to-live values can improve responsiveness to changes but increase DNS query volume and do not guarantee immediate failover. Centralized enterprise resolvers can also make many geographically dispersed users appear to originate from one location, reducing selection accuracy.

Hybrid and multi-account architectures require particular attention to DNS visibility and query paths. Establish which public or private namespace owns the application name, how accounts receive access to private zones, and how on-premises resolvers forward queries. Conditional forwarding and resolver endpoints provide name-resolution connectivity; they do not themselves implement latency-based endpoint selection. Public Internet latency estimates may poorly represent a private path over a VPN or dedicated circuit. Validate the chosen endpoint and actual application latency from each relevant network.

Distinguish latency routing from other exam-tested policies. Geolocation routing selects destinations according to geographic rules, such as country-specific service requirements. Weighted routing assigns configured proportions of DNS responses, useful for gradual releases, but does not inherently select the fastest endpoint. Failover routing prefers a primary destination until health conditions require a secondary. Where supported, hierarchical DNS policies can combine regional latency selection with weighted deployment choices, but each policy layer needs explicit health and fallback behavior.

Finally, faster network access does not guarantee faster transactions. A nearby application that synchronously accesses a distant database can perform worse than a more distant, self-contained deployment. Plan data replication, session handling, consistency, regional capacity, and residency constraints before enabling cross-region selection. Test with both DNS queries and end-to-end application measurements rather than assuming the selected region is always optimal.

Learn Latency-Based Routing in lessons

3 short lessons on Latency-Based Routing in 2 certification courses.

Open the lessons