AWS Certified Data Engineer Associate DEA-C01 Practice Question

A data engineering team needs to ingest telemetry data from thousands of IoT devices. Each device sends small JSON payloads every few seconds. The operations team must detect anomalies within 30 seconds, while data scientists occasionally rebuild models and need to reprocess six months of raw events. Which ingestion pattern best meets these requirements?

Use Amazon API Gateway to invoke an AWS Lambda function that stores only aggregated metrics in Amazon DynamoDB while archiving raw events separately for later analysis.
Buffer events on each device and upload gzipped files to Amazon S3 once per hour, then trigger AWS Glue jobs to process the files.
Send each event directly to an Amazon S3 bucket and run Amazon Athena queries every 15 minutes to detect anomalies.
Implement a streaming pipeline that immediately writes each event to an append-only log (for example, Amazon Kinesis Data Streams) so consumers can process events in real time and replay the stream when needed.

Report Issue

Answer Description

A streaming ingestion pattern provides continuous, low-latency delivery of each event so that operations staff can act on anomalies in near real time. Services such as Amazon Kinesis Data Streams or Amazon MSK keep an ordered, append-only log that can be retained for days or months, allowing downstream consumers to reread or "replay" historical events whenever transformation logic or models change. Hourly batch uploads, periodic Athena queries, or storing only aggregated metrics cannot satisfy both the sub-minute latency and full replayability requirements.

Ask Bash

Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.