Exam AWS Certified Machine Learning - Specialty topic 1 question 289 discussion

A - kinessis added unnecessary additional cost and complexity and will add to latency B - wrangler is better suited for daat prep & feature engineering no merging C - Serverless so cost effective, trigger happens immediately so within lambda 15 min window and glue is made for these use cases D - Costly setup and maintenace

hint is 30 to 60 minutes and lambda has 15 minutes. Plus Glue provides many built-in functionality to perform the merge process much easier

A is not needed as we don't need to add Kinesis, it has no purpose here. B is possible but DataWrangler is more expensive then C. C is serverless and cost optimized, so C is correct. D is obviously too expensive.

C. Use an S3 event on the AWS Data Exchange S3 bucket to invoke an AWS Lambda function. Program the Lambda function to run an AWS Glue job that will merge the existing business data with the Athena table. Write the results back to Amazon S3. This solution avoids the need for continuous data streams or provisioning a persistent database cluster, which can incur higher costs. AWS Lambda can trigger cost-effective, short-duration tasks, and AWS Glue is a managed ETL service that can handle the data transformation and merging efficiently. The integration with Amazon S3 and Athena also aligns with the existing data flow and tools.

fix c - > b The most cost-effective solution is to use an S3 event to trigger a Lambda function that uses SageMaker Data Wrangler to merge the data. This solution avoids the need to provision and manage any additional resources, such as Kinesis streams, Firehose delivery streams, Glue jobs, or Redshift clusters. SageMaker Data Wrangler provides a visual interface to import, prepare, transform, and analyze data from various sources, including AWS Data Exchange products. It can also export the data preparation workflow to a Python script that can be executed by a Lambda function. This solution can meet the time requirement of 30-60 minutes, depending on the size and complexity of the data. References: Using Amazon S3 Event Notifications Prepare ML Data with Amazon SageMaker Data Wrangler AWS Lambda Function

he most cost-effective and straightforward solution is C. Use an S3 event on the AWS Data Exchange S3 bucket to invoke an AWS Lambda function. Program the Lambda function to run an AWS Glue job that will merge the existing business data with the Athena table and write the results back to Amazon S3. This approach leverages the serverless architecture of AWS, minimizing operational overhead and cost while ensuring the transformations can be completed within the desired timeframe.