SOA-C02 Networking and Content Delivery Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Connection draining is set too low

  Why it's wrong here

  Connection draining (also known as deregistration delay) only takes effect when an instance is being deregistered or is marked unhealthy due to registration health checks. During normal operation, in-flight requests are unaffected, so a low draining value cannot be the cause of timeouts. Even if draining were low, the ALB would wait that many seconds before forcibly closing connections to the deregistering instance, which is a distinct mechanism from the idle timeout that governs steady-state request processing.

• ✓

  The ALB idle timeout is set too low

  Why this is correct

  The ALB idle timeout is the maximum time the load balancer allows between successive bytes of a client request or response before closing the connection. If an application takes longer than this interval to process a request and return data, the ALB terminates the connection, which appears to the client as a timeout error. For example, with the default 60-second idle timeout, a backend that runs a 90-second database query will consistently fail unless the timeout is increased to cover the expected processing time.

• ✗

  The target group health check interval is too long

  Why it's wrong here

  The target group health check interval controls how often the ALB sends HTTP health check requests to each registered instance. A longer interval simply means that an unhealthy instance will be detected and replaced more slowly, but it does not alter the behavior of an individual client request. Requests are forwarded to instances regardless of the health check schedule; a timeout occurs only if the ALB itself closes the connection, which is governed by the idle timeout, not the interval between probes.

• ✗

  Cross-zone load balancing is disabled

  Why it's wrong here

  Disabling cross-zone load balancing causes the ALB to only route traffic to targets in the same Availability Zone as the client-facing node, rather than distributing across all AZs. This can produce uneven utilization if instance counts differ per AZ, potentially saturating the smaller AZ and increasing latency. However, saturating the application would result in slow responses, not the immediate connection closing that defines a timeout; the ALB's idle timeout is the component that directly enforces a response deadline regardless of load distribution.

Go deeper

Related to this question