AWS Global Accelerator and NLB for Hybrid Workloads

SAA-C03 By islever Editorial Last updated

Short answer

AD Use AWS Global Accelerator with Network Load Balancers to route TCP and UDP traffic to regional endpoints.

Question

A company wants to improve the availability and performance of its hybrid application. The application consists of a stateful TCP-based workload hosted on Amazon EC2 instances in different AWS Regions and a stateless UDP-based workload hosted on premises. Which combination of actions should a solutions architect take to improve availability and performance? (Choose two.)
  1. Create an accelerator using AWS Global Accelerator. Add the load balancers as endpoints. Correct Answer
  2. Create an Amazon CloudFront distribution with an origin that uses Amazon Route 53 latency-based routing to route requests to the load balancers.
  3. Configure two Application Load Balancers in each Region. The first will route to the EC2 endpoints, and the second will route to the on-premises endpoints.
  4. Configure a Network Load Balancer in each Region to address the EC2 endpoints. Configure a Network Load Balancer in each Region that routes to the on-premises endpoints. Correct Answer
  5. Configure a Network Load Balancer in each Region to address the EC2 endpoints. Configure an Application Load Balancer in each Region that routes to the on-premises endpoints.

Key Takeaways

The core trap is confusing ALBs with NLBs for non-HTTP traffic; NLBs are required for TCP/UDP, while Global Accelerator provides the global routing layer.

This question tests the selection of AWS networking services to optimize availability and performance for mixed stateful TCP and stateless UDP hybrid workloads across regions and on-premises locations.

Common mistake: Many learners incorrectly choose B (CloudFront) because they associate it with performance, but CloudFront is limited to HTTP/HTTPS and cannot handle raw TCP or UDP traffic.

Expert Analysis

Why the Answer Is Correct

Option A is correct because AWS Global Accelerator uses the AWS global backbone to route traffic to healthy endpoints in different regions, significantly improving availability and reducing latency compared to public internet routing. Option D is correct because Network Load Balancers (NLBs) operate at Layer 4 and support both TCP and UDP protocols, making them suitable for the stateful TCP EC2 workload and the stateless UDP on-premises workload respectively. Application Load Balancers (ALBs) only support Layer 7 (HTTP/HTTPS), so they cannot handle the UDP traffic.

Why the Other Options Are Wrong

Option B is incorrect because Amazon CloudFront is a Content Delivery Network that only supports HTTP and HTTPS traffic; it cannot route raw TCP or UDP packets. Options C and E are incorrect because they suggest using Application Load Balancers (ALBs). ALBs are Layer 7 load balancers designed for web traffic and do not support UDP or raw TCP load balancing, which would fail for the specified workloads.

Community Comment Notes

Community consensus strongly supports AD. As user Ricksaurus noted, 'TCP >> NLB non-http >> accelerator,' highlighting the protocol requirements. Another user, muhammadahmer36, reinforced this by stating 'TCP use NLB FOR non-http use accelerator.' User swati1508 also confirmed 'AD -TCP use NLB FOR non-http use accelerator.' One learner questioned the multi-region deployment of NLBs for on-premises endpoints, but Global Accelerator requires endpoints in multiple regions to provide its high-availability benefits.

Exam Strategy

Always match the protocol (TCP/UDP vs HTTP) to the load balancer type (NLB vs ALB). When global low-latency routing is needed for non-web protocols, think Global Accelerator first.

Frequently Asked Questions

Why can't I use an Application Load Balancer for the UDP traffic?

ALBs operate at Layer 7 and only support HTTP/HTTPS. They cannot process raw UDP or TCP streams required by the application.

Does Global Accelerator work with on-premises endpoints?

Yes, Global Accelerator can route traffic to on-premises servers via IP addresses as endpoints, provided they are reachable from the AWS edge locations.

More SAA-C03 FAQ →

Community Votes

AD
100%

100% of anonymous learners picked answer AD. Votes are pick records left by other test-takers — they are not the verified answer.

Discussion 4 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 4

  1. Ricksaurus 👍 6
    TCP >> NLB non-http >> accelerator.
  2. swati1508 👍 3
    AD -TCP use NLB FOR non-http use accelerator.
  3. muhammadahmer36 👍 3
    TCP >> NLB non-http >> accelerator
  4. GOTJ 👍 1
    But, regarding to option "D", what's the reason behind the multi-regional deploy of the NLB for on-premises UDP endpoints? Shouldn't it be deployed on the nearest region only?

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Log in to comment, report an error, or add a note about this question.

Submitted for moderation before publishing. Keep it helpful and respectful.

Practice All SAA-C03 Questions

Access 120 questions with complete answers and detailed explanations.

View Full SAA-C03 Practice Test →

← Back to SAA-C03 Study Guide