Exam Professional Cloud Network Engineer topic 1 question 25 discussion

All of these answers stink. How would a load balancer decrease latency to your application? Latency and distance are related and none of these are decreasing either. B is the best of the worst.

The correct option is B.An HTTP(S) load balancer provides global load balancing, which can direct user traffic to the closest backend, based on the user's geographic location. Since users in Asia are experiencing high latency when the application is hosted in us-central1, configuring an HTTP(S) load balancer will allow the system to route traffic to the nearest available backend (such as one located in an Asian region), significantly reducing the latency. In contrast, a Network Load Balancer operates at the TCP/UDP level and does not optimize for geographic location, leading to high latency for users far from the us-central1 region.

To decrease latency for users in Asia accessing a web application hosted in the us-central1 region, one effective strategy is to leverage content delivery networks (CDNs) that have edge locations in Asia. This helps serve content from a location closer to the users, reducing latency. Therefore, the most suitable option is: B. Configure an HTTP load balancer, and direct the traffic to it.

Cloud CDN works with the global external Application Load Balancer or the classic Application Load Balancer to deliver content to your users https://cloud.google.com/cdn/docs/overview

B: : Dynamic Routing does not work because the application is ONLY in one us-central1 region. Network Load Balancing With a network load balancer, user requests still enter the Google network at the closest edge PoP (in Premium Tier). In the region where the project's VMs are located, traffic flows first through a network load balancer.

NLB - just pass through - the same syn, syn-ack, ack ... GLB (http) - proxy - keeps conn open to backend assuming its premium tier both use closes POP to enter - so no regional aspect is needed to consider https://cloud.google.com/load-balancing/docs/tutorials/optimize-app-latency#network-load-balancing

The question is asking how to route the packets from Asia to US by leveraging Google's private network instead of Internet. Hence to reduce latency. B is correct as HTTP LB - backends can be in any region and any VPC network (Premium tier). On the other hand, Network LB - the backend service must also be in the same region and VPC network as the forwarding rule.

The correct answer is "B"

Answer is : B

B is the one

An HTTP load balancer may help a bit While the speed of light will be unchanged (US <-> asia is a long trip), users will connect to the http load balancer An tcp connection uses a 3 way handshake, so additionnal roundtrip are required But http load balancers uses keepalived, so connections to the origin are kept across requests So, instead of cust <-> US (2 long RTT), you get cust <-> asia (2 small RTT) + asia <-> US (1 long RTT)

Answer is B. Network LB is regional service. This scenario requires global scale type of LB, thus HTTP LB is the correct choice.

B is correct

B is correct answer here.