Questions 13 to 15

Question 13

A solutions architect is designing an application on AWS. The compute layer will run in parallel across EC2 instances. The compute layer should scale based on the number of jobs that are to be processed. The compute layer is stateless. The solutions architect must ensure that the application is loosely coupled and the job items are durably stored.

Which design should the solutions architect use?

1. Create an Amazon SNS topic to send the jobs that need to be processed. Create an Amazon EC2 Auto Scaling group for the compute application. Set the scaling policy for the Auto Scaling group to add and remove nodes based on CPU usage.
2. Create an Amazon SQS queue to hold the jobs that need to be processed. Create an Amazon EC2 Auto Scaling group for the compute application. Set the scaling policy for the Auto Scaling group to add and remove nodes based on network usage.
3. Create an Amazon SQS queue to hold the jobs that need to be processed. Create an Amazon EC2 Auto Scaling group for the compute application. Set the scaling policy for the Auto Scaling group to add and remove nodes based on the number of items in the SQS queue.
4. Create an Amazon SNS topic to send the jobs that need to be processed. Create an Amazon EC2 Auto Scaling group for the compute application. Set the scaling policy for the Auto Scaling group to add and remove nodes based on the number of messages published to the SNS topic.

Correct Answer: 3

Explanation: In this case, we need to find a durable and loosely coupled solution for storing jobs. Amazon SQS is ideal for this use case and can be configured to use dynamic scaling based on the number of jobs waiting in the queue.

To configure this scaling, you can use the backlog per instance metric with the target value being the acceptable backlog per instance to maintain. You can calculate these numbers as follows:

• Backlog per instance: To calculate your backlog per instance, start with the ApproximateNumberOfMessages queue attribute to determine the length of the SQS queue (number of messages available for retrieval from the queue). Divide that number by the fleet’s running capacity (which for an Auto Scaling group is the number of instances in the InService state) to get the backlog per instance.
• Acceptable backlog per instance: To calculate your target value, first determine what your application can accept in terms of latency. Then, take the acceptable latency value and divide it by the average time that an EC2 instance takes to process a message.

This solution will scale EC2 instances using Auto Scaling based on the number of jobs waiting in the SQS queue.

INCORRECT: “Create an Amazon SNS topic to send the jobs that need to be processed. Create an Amazon EC2 Auto Scaling group for the compute application. Set the scaling policy for the Auto Scaling group to add and remove nodes based on CPU usage.” is incorrect. Amazon SNS is a notification service that delivers notifications to subscribers. It stores data durably but is less suitable than SQS for this use case. Scaling on CPU usage is not the best solution as it does not relate to the number of jobs waiting to be processed.

INCORRECT: “Create an Amazon SQS queue to hold the jobs that need to be processed. Create an Amazon EC2 Auto Scaling group for the compute application. Set the scaling policy for the Auto Scaling group to add and remove nodes based on network usage.” is incorrect as scaling on network usage does not relate to the number of jobs waiting to be processed.

CORRECT: “Create an Amazon SQS queue to hold the jobs that need to be processed. Create an Amazon EC2 Auto Scaling group for the compute application. Set the scaling policy for the Auto Scaling group to add and remove nodes based on the number of items in the SQS queue.” is the correct answer.

INCORRECT: “Create an Amazon SNS topic to send the jobs that need to be processed. Create an Amazon EC2 Auto Scaling group for the compute application. Set the scaling policy for the Auto Scaling group to add and remove nodes based on the number of messages published to the SNS topic.” is incorrect. Amazon SNS is a notification service, so it delivers notifications to subscribers. It stores data durably but is less suitable than SQS for this use case. Scaling on the number of notifications in SNS is not possible.

References: https://docs.aws.amazon.com/autoscaling/ec2/userguide/as-using-sqs-queue.html