Exam AWS Certified Solutions Architect - Professional topic 1 question 408 discussion

C https://aws.amazon.com/getting-started/projects/build-serverless-web-app-lambda-apigateway-s3-dynamodb-cognito/ A\D: Should be latency based routing the ensure latency is at a minimum. Remember the users are spread globally not to specific regions where you can maybe use geo to spread the load across just a few region. B: Similar to A.

D is correct Below is summary of those discussion A. RDS 99.95%, no across region strategy B. Round robin to different region (latency), Aurora 99.99% meet C. S3 not for dynamic web content, API Gateway & lambda 99.95% D. ECS, EC2, Fargate, NLB, Aurora 99.99% meet, using geolcation-based routing is enough as regional service is 99.99%

C seems correct for me

Will go with D as millions of concurrent request NLB

C seems correct for me

C - Will go with C

Answer is D. The users are worldwide so Geo location based routing is better. Moreover AWS always promotes its services in questions. So aurora multi master and Fargate is better option. Cost for Fargate will be cost efficient.

C is correct

C. Use Amazon Route 53 latency-based routing to route to the nearest region with health checks. Host the website in Amazon S3 in each region and use Amazon API Gateway with AWS Lambda for the application layer. Use Amazon DynamoDB global tables as the data layer with Amazon DynamoDB Accelerator (DAX) for caching.

D ---

It's D

A, C, and D can be eliminated for not fulfilling all requirements. A: SLA not met C: does not allow dynamic web applications, Lambda does not support the required concurrency D: no health checks, Aurora Multi-Master is in one region only B does not look great but fulfills the requirements.

Option-D looks correct. With Option-C two flaws. Lambda concurrency limitations (https://stackoverflow.com/questions/36766553/aws-lambda-concurrent-request-limit-and-how-to-increase-that) and S3 can't be used for Dynamic Website.

D Millions of concurrent requests. NLB is needed here, lambda only can handle 10 000 concurrent requests per sec of I remember correctly

for me it's C

D - Aurora is cheapest and it supports multi region for read-replicas

Read the question: "redesigned" to be "scalable" and allow for "automatic failover". You can only choose between B and C. Roud robin is not suitable due to potential latency happening sometime for some of the users. You should use latency-based routing or geolocation-based routing depending on you application/system usage. I would choose C DAX will lower DynamoDB cost and S3 with Lambda is potentially cheaper than Fargate and load balancers. From experience - you have to run at least one container per task no matter their utilization is. Having millions of user accessing web site running in conatiners it would generate a lot of $ for ECS. Instead redesign your app to use S3 and API Gateway.