Course contentsShow
← Cloud and Infrastructure
Lesson 259 of 2,9895. Load Balancing and Traffic ManagementFree lesson

Cross-Zone Load Balancing for ALBs

Distribute traffic evenly across availability zones to improve fault tolerance and avoid zone-specific overloading.

Cross-Zone Load Balancing for ALBs

What you'll learn: How Application Load Balancers distribute traffic evenly across availability zones to prevent zone-specific overloads and improve fault tolerance.

The Problem: Uneven Zone Distribution

Imagine a shopping mall with three entrances, but one entrance has only 2 stores while the other two have 10 stores each. If customers are sent proportionally to each entrance, the 2-store entrance gets overwhelmed while the others run smoothly. This is what happens without cross-zone load balancing.

By default, load balancers receive traffic and distribute it to targets (instances, containers, IPs) within their own availability zone. If you have unequal numbers of healthy targets across zones, some zones handle disproportionately more requests per target.

What Cross-Zone Load Balancing Does

With cross-zone load balancing enabled (which is the default for ALBs), your load balancer distributes traffic evenly across all healthy targets in all enabled zones, regardless of which zone received the incoming request.

Example scenario:

  • Zone A: 1 healthy target
  • Zone B: 3 healthy targets
  • Zone C: 2 healthy targets

Without cross-zone: Each zone gets ~33% of traffic. Zone A's single target handles 33%, while Zone B's three targets share 33% (~11% each).

With cross-zone: All 6 targets receive ~16.67% of traffic each, regardless of zone.

Why This Matters

Cross-zone load balancing prevents zone-specific "hot spots" and ensures even workload distribution. This improves fault tolerance—if a zone becomes degraded, the load is already balanced across all zones. For ALBs, this feature is enabled by default and highly recommended to keep enabled.

Key Takeaway: Cross-zone load balancing for ALBs distributes traffic evenly across all healthy targets in all availability zones, preventing individual zones from becoming bottlenecks and improving overall fault tolerance.