A company ingests large volumes of streaming data from an application running on Amazon EC2. The app sends data to Amazon Kinesis Data Streams configured with default settings. Every other day the application consumes the data and writes it to an Amazon S3 bucket for BI processing. The company finds that S3 is not receiving all the data sent to Kinesis Data Streams. What should a solutions architect do to resolve this?

The correct solution is to update the Kinesis Data Streams default settings by modifying the data retention period. Kinesis Data Streams has a default data retention period of 24 hours. Since the application consumes data every other day (48 hours), data from the first day would have expired and been lost before consumption. Increasing the retention period to at least 48 hours will ensure all data is available when the application processes it. Updating the application to use KPL is not directly related to data loss due to retention; KPL helps with efficient data production but doesn't solve data expiry. Updating the number of Kinesis shards addresses throughput issues, not data loss from retention. Turning on S3 Versioning is irrelevant as the problem is data not reaching S3 at all, not issues with objects once they are in S3.