Exam CAS-004 topic 1 question 4 discussion

A distributed Content delivery network's purpose is to reduce latency for users utilizing the application and reduce the load on the central server by splitting the resources into different areas, regardless of the user's geographical location. Source: Verifying each answer against Chat GPT, my experience, other test banks, a written book, and weighing in the discussion from all users to create a 100% accurate guide for myself before I take the exam. (It isn't easy because of the time needed, but it is doing my diligence)

100% Thanks to AWS examination. CDN is always the choice for static assets such as image or video Read replica can reduce the load to the server

Distributed CDNs for Static Content: Serving static content (like images) via a distributed Content Delivery Network means that users—especially international ones—get the images from servers that are geographically closer to them. This greatly reduces latency and speeds up the initial loading of images.

A. Serve static content via distributed CDNs, create a read replica of the central database and pull reports from there, and auto-scale API servers based on performance. This option provides solutions for each of the identified issues: Serving static content via distributed Content Delivery Networks (CDNs) can reduce latency for international users because it serves content from the closest geographic location to the user. Creating a read replica of the central database and pulling reports from there can reduce the load on the central database, minimizing disruptions to inventory lookups during report processing. Auto-scaling API servers based on performance can ensure that as load increases, new servers are spun up to handle it, distributing the load more evenly and preventing any single server from becoming overloaded. The other options offer some potential improvements but don't address all of the identified issues as effectively.

The infrastructure design changes that would be BEST for the organization to implement to address the highlighted issues are: C. Serve images from an object storage bucket with infrequent read times, replicate the database across different regions, and dynamically create API servers based on load. Explanation: Latency for loading images Issues with inventory during report processing Heavy load on API servers at peak times

Option A is the BEST choice.

Now, let's examine why the other options are not ideal: B. This option partially addresses some issues but falls short in other areas. Increasing the bandwidth and using a CDN would improve image delivery, but changing the database to a non-relational database might not solve the inventory issues during report processing. Additionally, simply splitting the API servers across two load balancers does not provide an auto-scaling solution, which is crucial for handling peak times.

The correct answer is A. Serve static content via distributed CDNs, create a read replica of the central database and pull reports from there, and auto-scale API servers based on performance. This option addresses all three issues as follows: Serving static content via distributed CDNs reduces latency for international users because content is served from locations closer to them. Creating a read replica of the central database and pulling reports from there offloads the reporting workload from the main database, reducing the impact on inventory operations during order placement. Auto-scaling API servers based on performance ensures that resources adjust dynamically to handle varying workloads, preventing heavy load on servers during peak times.

A is the correct answer. B is not correct; changing the architecture of the database would help in some situations, but you would still have a single database in play. Read replicas mean that you can have several databases handling read requests, with only one master handling writes.

B: is good, but A: is better as it scales and has advantages regarding number of databases

When thinking cloud, it opens up the option to autoscale servers depending on resource usage. Therefore A is best since it can account for sudden spikes in load.

A makes more sense.

I believe B isn't right because it states that despite the fact 10 were added, the load was still heavy at times. So you would want and auto-scale API to keep up with performance. Something that will add more API's as needed.

Why wouldn't B. be the best answer?