Which combination of steps should a solutions architect take to meet these requirements?
(Choose two.)
Use AWS Glue to process the raw data in Amazon S3.
Use Amazon Route 53 to route traffic to different EC2 instances.
Add more EC2 instances to accommodate the increasing amount of incoming data.
Send the raw data to Amazon Simple Queue Service (Amazon SQS). Use EC2 instances to process the data.
Use Amazon API Gateway to send the raw data to an Amazon Kinesis data stream. Configure Amazon Kinesis Data Firehose to use the data stream as a source to deliver the data to Amazon S3.
Explanations:
AWS Glue is a fully managed ETL (Extract, Transform, Load) service that can efficiently process and transform large datasets in Amazon S3, making it a suitable choice for processing the raw data collected from millions of devices.
Amazon Route 53 is a DNS service and is not designed for managing or scaling EC2 instances directly. It cannot effectively handle the increasing load of incoming data from millions of devices.
Adding more EC2 instances would increase operational overhead and complexity. It is not a scalable solution, especially considering the anticipated increase in the number of devices.
While sending data to Amazon SQS can help decouple the components of the application, using EC2 instances for processing the data introduces additional operational overhead, which the company wants to minimize.
Using Amazon API Gateway to send data to an Amazon Kinesis data stream allows for a highly scalable solution that can handle millions of devices. Kinesis Data Firehose can then deliver the data to Amazon S3 without requiring direct management of EC2 instances, minimizing operational overhead.
I systematize that the answer is:
Use AWS Glue to process the raw data in Amazon S3.