AWS Certified CloudOps Engineer Associate SOA-C03 Practice Question

A company runs an auto-scaled web service behind an Application Load Balancer (ALB) in three Availability Zones. Each instance exposes /healthcheck, which executes several database queries before returning HTTP 200. The ALB target group is configured with: protocol HTTP, port 80, path /healthcheck, interval 10 seconds, timeout 5 seconds, healthy threshold 5, and unhealthy threshold 2. During peak traffic some healthy instances take 6-7 seconds to answer the health probe, are marked unhealthy, and are replaced by Auto Scaling, causing additional load on the remaining fleet. Which modification will best prevent unnecessary instance replacement while still allowing the ALB to detect real failures quickly?

Change the health check protocol from HTTP to TCP so that only the listener port needs to be open.
Increase the health check timeout from 5 seconds to 10 seconds and keep all other parameters unchanged.
Keep the current settings but raise the health check interval to 30 seconds to reduce the number of failed probes.
Create a lightweight endpoint such as /status that returns an immediate 200 response and update the target group health check path to use it.

Report Issue

Answer Description

The /healthcheck endpoint performs blocking database work, so its latency rises under load, exceeding the 5-second timeout and triggering unhealthy status after only two failed probes. The most effective fix is to expose a lightweight endpoint (for example /status) that returns a simple 200 response without external dependencies and reconfigure the ALB to probe that path. This keeps health checks fast and reliable regardless of workload, allowing the load balancer to distinguish true instance failure from transient application slowness.

Increasing the timeout or the unhealthy threshold merely delays-but does not remove-the heavy database work, so probes can still exceed the new limits during extreme load. Switching the protocol to TCP would mark the instance healthy as long as the port is open, even if the application stack is non-functional, and changing the interval to 30 seconds slows detection of genuine outages.

Ask Bash

Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.