AWS questions EP21

Domain: Design High-Performing Architectures

A manufacturing firm has a large number of smart devices installed in various locations worldwide. Hourly logs from these devices are stored in an Amazon S3 bucket. Management is looking for comprehensive dashboards which should incorporate usages of these devices and forecast usage trends for these devices.

Which tool is the best suited to get this required dashboard?

• A. Use S3 as a source for Amazon QuickSight and create dashboards for usage and forecast trends

• B. Use S3 as a source for Amazon Redshift and create dashboards for usage and forecast trends

• C. Copy data from Amazon S3 to Amazon DynamoDB. Use Amazon DynamoDB as a source for Amazon QuickSight and create dashboards for usage and forecast trends

• D. Copy data from Amazon S3 to Amazon RDS. Use Amazon RDS as a source for Amazon QuickSight and create dashboards for usage and forecast trends

Answer explanation

• Option B is incorrect as Amazon Redshift is a data warehousing service for analyzing structured or semi-structured data. It is not a useful tool for creating dashboards.

• Option C is incorrect as Amazon S3 can be used directly as a source for Amazon QuickSight. There is no need to copy data from Amazon S3 to Amazon DynamoDB.

• Option D is incorrect as Amazon S3 can be used directly as a source for Amazon QuickSight. There is no need to copy data from Amazon S3 to Amazon RDS.

Domain: Design Resilient Architectures

Your company uses an S3 bucket to store data for an application. Sometimes the team also downloads the S3 files for further analysis. As the data is very important, you need to protect against accidental deletions initiated by someone or an application and restore the files when needed.

Which of the following options is appropriate?

• A. Enable the versioning feature in the S3 bucket.

• B. Modify the S3 bucket to be read-only.

• C. Use an S3 Lifecycle policy to transfer objects to a lower cost storage.

• D. Enable the Server-Side Encryption with AWS KMS-Managed Keys (SSE-KMS).

Answer explanation

Correct​ ​Answer​ ​–​ A

• Option​ ​A ​is​ CORRECT:​ The versioning feature in Amazon S3 helps retain prior versions of objects stored in S3. Even if the current version is deleted accidentally, the data can still be recovered from the previous version. Check the reference here https://d1.awsstatic.com/whitepapers/aws-building-fault-tolerant-applications.pdf.

• Option​ ​B ​is​ ​incorrect:​ The application may still need to write data in the S3 bucket. So the option is not appropriate.

• Option​ ​C ​is​ ​incorrect:​ This method helps to save costs but not protect the data.

• Option​ ​D ​is​ ​incorrect:​ Because the Server-Side Encryption cannot protect against accidental deletions.

Domain: Design High-Performing Architectures

You are requested to guide a large Pharma company. They are looking for a solution to save all their R&D test analysis data securely. Daily large numbers of reports are generated; this data would be accessed from multiple R&D centers spread across the globe. The company requires this data to be instantaneously available to all users.

Which of the following is the most suitable way for AWS storage to provide low latency access to users across the globe with the least cost?

• A. Use Amazon EC2 instance with instance store to store data.

• B. Use Amazon EFS volumes to store data.

• C. Use Amazon EBS volumes connected to the EC2 instance to store data

• D. Use Amazon S3 Standard storage class from Amazon S3 to store data.

Answer explanation

Correct Answer – D

S3 Standard offers high durability, availability, and performance object storage for frequently accessed data. Because it delivers low latency and high throughput, S3 Standard is appropriate for a wide variety of use cases, including cloud applications, dynamic websites, content distribution, mobile and gaming applications, and big data analytics.

• Option A is incorrect as the Instance store will provide low latency access from EC2 instances, but it’s a temporary storage option.

• Option B is incorrect. Since this data would be accessed globally, Amazon EFS will not be ideal for this requirement.

• Option C is incorrect as Amazon EBS volumes would be useful storage for single Amazon EC2 instances.

For more information on AWS Storage options, refer to the following URLs-

• https://aws.amazon.com/s3/storage-classes/

• https://docs.aws.amazon.com/efs/latest/ug/storage-classes.html

Domain: Design for New Solutions

A new web application is designed by a developer team using Amazon API Gateway and AWS Lambda. REST APIs will be used for this purpose which will synchronously invoke the AWS Lambda function. The developer team is expecting a huge response to this application and looking for a secure solution for authenticating a large number of clients.            

Which solution can be designed for this requirement in the most effective way?

• A. Create a request parameter-based Lambda authorizer. Configure API Gateway to call Lambda authorizer to authenticate clients making a request using REST API.

• B. Create users in Amazon Cognito user pools. Create an API Gateway authorizer with this user pool and enable the authorizer on the REST API.

• C. Create resource policies matching specific users. Attach this resource policy to Amazon API Gateway which will control access to the APIs.

• D. Create an IAM user policy to invoke API. Attach this policy to Amazon API Gateway which will control access to the APIs.

Answer explanation

Correct Answer: B

Amazon Cognito user pool can be used to control who can access REST APIs in the API Gateway. With this method,

1. Clients sign up with a user pool.

2. Once clients sign in with a user pool, they obtain an identity token.

3. Clients pass this token with API Gateway to get access to the REST APIs created in the API Gateway.

This method is best suited when there is a need to provide access to a large number of global users.

• Option A is incorrect as request parameter-based Lambda authorizers are supported only for WebSocket API and not REST APIs.

• Option C is incorrect because there would be a large number of clients accessing these APIs, and using resource policies will not be an efficient way to control access.

• Option D is incorrect as IAM-based policies are best suited for controlling access to API from within the AWS Infrastructure. These are not suited to control access for external clients.

Domain: Continuous Improvement for Existing Solutions

A Java-based web application is migrated from On-premises to AWS. Amazon Elastic Beanstalk is used to deploy applications using ALB, Amazon EC2 instances in multiple AZ, and Amazon DynamoDB in provisioned capacity mode. Auto-scaling groups are created with general-purpose Amazon EC2 instances. Post-migration, customers observe a delayed response from web applications during busy hours. Further checks observed that the application performs a non-critical zip archive process causing load on the Amazon EC2 instance.

How can a solution be redesigned in a cost-effective way to get a better response from the web application?

• A. Redeploy the application using Elastic Beanstalk worker environments with Amazon SQS and Amazon EC2 instance for the zip archive process. For the front-end, use the web server environment of ALB, Amazon EC2 instance, and Amazon DynamoDB.

• B. Redeploy Auto Scaling groups with memory-intensive Amazon EC2 instance instead of general-purpose Amazon EC2 instance.

• C. Redeploy application on Amazon EC2 instance with Elastic fabric Adaptor (EFA).

• D. Redeploy the database layer with on-demand capacity mode instead of the provisioned capacity mode of the Amazon DynamoDB.

Answer explanation

Correct Answer: A

A worker environment can be deployed with Amazon Elastic Beanstalk for long-running application tasks. Long-running jobs can be sent to the Amazon SQS queue and are further processed by Amazon EC2 instances in the worker environment tier.

In the above case, an application deployed on an Amazon EC2 instance is getting overwhelmed due to an additional non-critical process of generating zip archives. To remediate this issue in a cost-effective way, the zip process can be decoupled to a separate tier, which in turn can increase the performance of the application during busy hours.

Diagram showing typical deployment using a web server and worker environment tiers with Amazon Elastic Beanstalk.

• Option B is incorrect as the ZIP archive is generally a compute-intensive process. Deploying a memory-intensive Amazon EC2 instance won’t be any value addition. Also, using memory-intensive Amazon EC2 instances will be costly. The application can be implemented by decoupling the zip process and decreasing the load on the Amazon EC2 instance.

• Option C is incorrect as Amazon EC2 instances with Elastic Fabric Adaptor are more suitable for applications requiring high-level inter-node communications. In the above case, application response is getting affected due to additional non-critical processes running on the Amazon EC2 instance. 

• Option D is incorrect as since the issue is with overloading of the application due to processing of non-critical tasks of the zip archive, changing capacity mode at the database layer won’t help and in turn will be costly.

Who am I ?

Who am I ?