AWS SAA-C03 HIGH AVAILABILITY GUIDE

ALB vs NLB vs Gateway Load Balancer

Choose by what must be understood and distributed: HTTP requests, network connections, or traffic that must pass through security appliances.

Updated 18 August 2026 · Original ITCertPath learning resource

Match the load balancer to the traffic decision

Application Load Balancer

Layer 7 HTTP/HTTPS routing. Choose it when rules must inspect application content such as host, path, header, method, query string or source IP.

Network Load Balancer

Layer 4 connection and flow distribution for TCP, UDP, TLS and supported QUIC protocols. It suits high-throughput network traffic and static-IP requirements.

Gateway Load Balancer

Layer 3 transparent gateway for scaling virtual appliances such as firewalls and intrusion prevention or deep-packet inspection systems.

DecisionALBNLBGWLB
OSI focusLayer 7Layer 4Layer 3
Primary trafficHTTP and HTTPSTCP, UDP, TLS and supported QUICIP packets to appliances
Special strengthContent-based routing and WAF integrationStatic per-AZ IP and very high connection scaleTransparent appliance fleet insertion
Typical targetEC2, IP, containers or LambdaInstance, IP or ALB targetVirtual appliance instances

Architecture scenarios that reveal the answer

One domain, several microservices

Requests for /orders must reach an orders target group while /catalog reaches a catalog target group. Choose an ALB because listener rules evaluate HTTP request content and health checks operate per target group.

Low-latency TCP service with allow-listed IPs

A partner requires fixed destination IP addresses and the application uses TCP rather than HTTP. Choose an NLB, enable the required Availability Zones, and optionally associate an Elastic IP per enabled subnet for an internet-facing design.

Central inspection for multiple application VPCs

Traffic must pass transparently through a scalable fleet of third-party firewalls. Choose a GWLB with Gateway Load Balancer endpoints and route tables that insert the appliance service into the traffic path.

Load balancing and Auto Scaling solve different problems

Route 53 or clientResolves the application endpoint
Load balancerDistributes only to healthy targets
Target groupDefines targets and health checks
Auto Scaling groupMaintains and adjusts EC2 capacity

Elastic Load Balancing scales its own capacity and routes around unhealthy registered targets. EC2 Auto Scaling launches, replaces or terminates instances according to desired capacity and scaling policies. Attach the target group to the Auto Scaling group so instances register and deregister automatically.

Health-check trap: an instance can be running while the application is unhealthy. Use a health-check endpoint that tests the dependency level needed for safe traffic, configure grace periods for startup, and inspect the target-group reason code before replacing infrastructure.

Real-world troubleshooting sequence

  1. Listener: confirm the protocol, port, certificate and default action.
  2. Rule: for ALB, verify priority and whether the request actually matches host or path conditions.
  3. Target group: verify target type, port, protocol and registered targets.
  4. Health check: inspect path, success codes, timeout, thresholds and application startup time.
  5. Network path: verify load-balancer and target security groups, NACLs and return traffic.
  6. Capacity: compare healthy-host count, request or connection metrics, response time and Auto Scaling activity.

Fast SAA-C03 decision rule

HTTP content decides the destination: ALB. Network flow, non-HTTP protocol or static IP matters: NLB. Traffic must traverse a scalable virtual-appliance fleet: GWLB.

Official AWS references

Frequently asked questions

Which AWS load balancer supports path-based routing?

Application Load Balancer supports Layer 7 rules based on request content, including URL paths and host headers, making it suitable for HTTP and HTTPS applications and microservices.

Which load balancer should I choose when clients require static IP addresses?

A Network Load Balancer provides a static IP address for each enabled Availability Zone and can optionally use an Elastic IP per subnet for an internet-facing load balancer.

What is Gateway Load Balancer used for?

Gateway Load Balancer deploys and scales fleets of virtual network appliances such as firewalls, intrusion prevention systems and deep-packet inspection systems while preserving transparent traffic flow.

Does a load balancer replace EC2 Auto Scaling?

No. Elastic Load Balancing distributes traffic and checks target health. EC2 Auto Scaling maintains and adjusts instance capacity. Used together, new instances register with target groups and terminated instances deregister.