AWS Online Assessment Test

Slapchat deployed a new data processing application to the market. The backend part of the application does two specific segments:

- For every new data file submitted by the user, proprietary data cleanup logic is done
- Once data cleanup is complete, proprietary data transformations are done

After launching the application for few months, the metrics indicated that the clean up stage of the application is pretty quick while transformation stage takes longer time to complete. In the next phase, the application is being re-written as 2 microservices (’cleanup’ and ‘transformation’). The deployment strategy is to run both on ECS and scale independently. How should the services be integrated?

A: ‘cleanup’ publishes data to SNS topic, ‘transformation’ subscribes to the SNS topic

B: ‘cleanup’ sends data to S3 bucket, ‘transformation’ is invoked by S3 event notifications

C: ‘cleanup’ sends data to SQS queue, ‘transformation’ processes messages from the queue

D: ‘cleanup’ sends data to Kinesis Data Firehose, ‘transformation’ reads from the Kinesis Data Firehose

ChickTok is a dating app that runs on EC2 behind an ALB. The EC2 instances run in an EC2 Auto Scaling group across multiple AZ. On every Saturday night, the app becomes atleast 5 times slower. Upon investigation, it is found that the CPU utilization reaches 100% within a matter of seconds instead of being a gradual increment. This is happening due to a batch core dating algorithm update that happens on every Saturday 9PM. 

ChickTok wants a solution that

- Tackles the extra workload
- Avoids downtime

Which of the following would you recommend?

One of your friends, Zi Pong created a new app for their startup. They use Amazon S3 exclusively as a data store. The app works perfectly fine. When one of their users raises the frequency at which the app needs to refresh data, they file a bug. The issue is that sometimes they get outdated data when accessing objects in S3. The app logic is correct. What do you think the issue could be?

A: App is reading objects from Amazon S3 using parallel object requests.
B: App is updating records by writing new objects with unique keys.
C: App is reading parts of objects from Amazon S3 using a range header.
D: App is updating records by overwriting existing objects with the same keys.

Flowlife is a tips iOS app that is running on AWS with ~10 million users in the US. Users can post tips about a particular place (restaurants, businesses etc) and other users can search for the tips in the app. Flowlife wants to implement location based alerts on their iOS app. Users will receive relevant tips in proximity to their location. For maximum user engagement and relevant tips, the tips must be in the low minute count. Which of the following solutions is cost effective?