Exam AWS Certified Solutions Architect - Associate SAA-C03 topic 1 question 145 discussion

I was tempted to pick A but then I realized there are two key requirements: - scale seamlessly - cost-effectively None of A-C give seamless scalability. A and B are about adding second instance (which I assume does not match to "scale seamlessly"). C is about changing instance type. Therefore D is only workable solution to the scalability requirement

I wouldn't run my website on spot instances. Spot instances might be terminated at any time, and since I need to run analytics application it's not an option for me. And using route 53 for load balancing of 2 instances is an overkill. I go with A.

Ans D - Like Konb (1 year, 5 months ago) I almost picked A, but for the same reasons of seamless scaling it has to be D

D aswer is forcing you to use the new "feature" of aws. if you have a performance issue with the ddbb, front and back all in on your ec2 intances. the best way to solve that issue is move de ddbb to RDS and create a new ec2 intances. BUT we need to sell the aws unique service.

Ans D - I was tempted with AAns D - I was tempted with A but then thought that's too obvious and scaling might be an issue... so looking at Spot Fleets "Fleets provide the following features and benefits, enabling you to maximize cost savings and optimize availability and performance when running applications on multiple EC2 instances" it has to be Ans D. https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/Fleets.html

Option A is the best balance between cost-effectiveness and scalability. It allows the application to scale horizontally with minimal changes while ensuring the database is managed and can scale independently, reducing the risk of performance degradation during peak times.

A, not D > Using Spot Instances with an Auto Scaling group adds complexity and risk to the infrastructure

I don't think D is the optimal solution. We certainly can find other solutions that are more cost effective and fulfill the same requirements, but among the provided options, I think D is the most reasonable.

PHP....

I'll pick D because it sounds fun :)

"Scale seamlessly", none of A-C include scaling at all.

spot instance receives 2 minutes interruption notice, it should be enough for requests to finish, it's quite unusal for app to run longer requests only option D allow for seamless scaling with autoscaling group

Option B is a cost-effective choice that combines the benefits of database migration to RDS, horizontal scaling with EC2 instances, and control over traffic distribution with Route 53 weighted routing, making it the best solution for the given requirements.

Scale seamlessly = Autoscaling group, Amazon Aurora MySQL DB instance Cost effective = Spot Fleet

The key reasons are: Migrating the database to Amazon Aurora MySQL provides a scalable, high performance database to support the application. Creating an AMI of the web application and using it in an Auto Scaling group with Spot instances allows cheap and efficient scaling of the web tier. The Application Load Balancer distributes traffic across the Auto Scaling group. Spot instances in an Auto Scaling group allow cost-optimized automatic scaling based on demand. This approach provides high availability and seamless scaling without manual intervention.

D is correct because migrating the database to Amazon Aurora provides better scalability and performance compared to Amazon RDS for MySQL. Creating an AMI of the web application allows for easy replication of the application on multiple instances. Using a launch template and Auto Scaling group with Spot Fleet provides cost optimization by leveraging spot instances. Adding an Application Load Balancer ensures the load is distributed across the instances for seamless scaling. A is incorrect because using an Application Load Balancer with multiple EC2 instances is a better approach for scalability compared to relying on a single instance. B is incorrect because weighted routing in Amazon Route 53 distributes traffic based on fixed weights, which may not dynamically adjust to the changing load. C is incorrect because using AWS Lambda to stop and change the instance type based on CPU utilization is not an efficient way to handle scaling for a web application. Auto Scaling is a better approach for dynamic scaling.

the options that said "launch a second EC2", have no sense ... why 2?, why not 3 or 4 or 5? so options A and B drop. C is no sense (Lambda doing this like a Scaling Group?, absurd) Has to be D. Little extrange cos' Aurora is a very good solution, but NOT CHEAP (remember: cost-effectively). To be honest, the most cost-effectively is B je je