220-1101 Question 102
Single answerLoad balancersA small online retailer hosts its website on two identical web servers behind a load balancer. During peak shopping hours, users report that the site sometimes becomes unavailable even though one of the web servers is still working properly. The technician discovers that the load balancer is still sending some client requests to the failed server for several minutes after it goes offline. Which configuration change would BEST prevent this issue?
- A
Enable health checks on the load balancer so it can automatically stop sending traffic to an unresponsive server
- B
Increase the DHCP lease time for client devices so their IP addresses remain consistent during peak hours
- C
Configure port forwarding on the router to direct all HTTP traffic to only one web server
- D
Replace the switch with a managed switch that supports VLAN tagging for web traffic
Show answer and explanation
Correct answer: A
Explanation
The best answer is to enable health checks on the load balancer. In a real-world environment, load balancers are used to distribute traffic across multiple servers to improve performance and availability. However, to provide fault tolerance, they must also verify that each backend server is actually healthy. Health checks, sometimes called heartbeat checks or monitoring probes, typically test a service such as HTTP, HTTPS, or TCP on a schedule. If a server stops responding or returns an invalid response, the load balancer removes it from the active pool until it recovers. This behavior is a standard best practice in load-balanced environments and is documented by major vendors and cloud providers such as F5, Citrix, AWS Elastic Load Balancing, and Microsoft Azure Load Balancer. The other options involve networking concepts that may be useful in other scenarios, but they do not address the specific problem of continuing to send traffic to a failed server.
- A. Correct.
Correct. Health checks allow the load balancer to regularly test whether each backend server is responding on the expected service or port. If a server fails, the load balancer marks it unavailable and removes it from the rotation, which prevents clients from being sent to a downed system. This is a core load-balancer function used to maintain availability.
- B. Incorrect.
Incorrect. DHCP lease duration affects how long client devices keep IP addresses on the local network. It does not determine whether a load balancer can detect failed web servers or stop forwarding requests to them.
- C. Incorrect.
Incorrect. Port forwarding sends traffic from a public interface to a designated internal host, but directing all HTTP traffic to one server would eliminate load distribution and redundancy. It would not solve the problem of intelligently removing failed servers from service.
- D. Incorrect.
Incorrect. A managed switch with VLAN support can improve traffic segmentation and administration, but VLAN tagging does not provide server health awareness or load-balancer failover decisions. The issue is with backend server monitoring, not Layer 2 segmentation.