Question about capacity reservation for ALB and NLB

You're right to be confused :) . ALBs have "redundancy within a Region" is correct, but it's crucial to understand what this means in the context of capacity reservations.

1. ALBs are designed to distribute traffic across multiple Availability Zones (AZs) within a Region. If one AZ fails, the ALB will automatically shift traffic to the healthy AZs. This failover is automatic and handled by AWS.
2. However, the capacity that your ALB has available is based on the LCUs you've provisioned. If you reserve the absolute minimum number of LCUs required for your normal traffic, and then an AZ fails, the remaining AZs might not have enough capacity to handle the increased load. This is where the confusion arises.
3. The "redundancy option" in the ALB capacity reservation calculator is designed to account for this. It recommends reserving additional LCUs so that your ALB can handle the increased traffic load in the event of an AZ failure. Even though the ALB itself is redundant, the capacity it has might not be. While ALB auto-scaling can eventually add more capacity, it takes time. Capacity reservation ensures that you have the necessary capacity immediately available during an AZ failure, preventing potential performance degradation or connection drops.

Crux: Even though ALBs have built-in redundancy, you still need to reserve extra capacity (LCUs) to handle AZ failures. The "redundancy option" in the calculator helps you determine how much extra capacity to reserve. It's not about the ALB failing, but about the capacity being sufficient after an AZ failure.