ExamPark
Course menuDEA-C01

Exams / DEA-C01 / Study guide / Choosing a data store

Choosing a data store: DEA-C01 study notes and practice questions

What to know about Choosing a data store for the DEA-C01 exam, from the Data Store Management domain (26% of the exam guide), then 1 free question to test it. Part of the DEA-C01 study guide.

Last verified against current AWS documentation ·How we write our questions

What to know for the exam

Store choice questions describe several workloads and ask which store fits each.

The map

Access pattern Store
Cheap, durable storage for raw and curated data, read by many engines Amazon S3 (data lake), with S3 Tables for managed Iceberg tables
Complex SQL analytics over large structured data, BI dashboards Amazon Redshift (provisioned or Serverless)
Single-digit millisecond key-value lookups at any scale Amazon DynamoDB
Transactions with joins and constraints Amazon RDS / Aurora
Full-text search, log analytics, vector search Amazon OpenSearch Service
Highly connected data (fraud rings, recommendations) Amazon Neptune
Time-series data (IoT, metrics) Amazon Timestream
Microsecond caching ElastiCache, DynamoDB DAX
Durable in-memory database Amazon MemoryDB
MongoDB-compatible documents Amazon DocumentDB
Ad hoc SQL on S3 Amazon Athena (a query engine, not a store)

Factors

  • Latency and query pattern: point lookups vs scans vs aggregations.
  • Volume and growth.
  • Schema: fixed, evolving or none.
  • Cost: storage vs compute vs per-query pricing.
  • Concurrency: many dashboard users, or a few analysts.
  • Integration with the tools that will read it.

Storage tiers inside services

  • S3 storage classes and lifecycle rules.
  • Redshift RA3 nodes separate compute from managed storage.
  • DynamoDB Standard-IA table class for rarely read tables.
  • OpenSearch UltraWarm and cold storage for older logs.

Common traps

  • Redshift for millisecond lookups by key. Use DynamoDB.
  • DynamoDB for ad hoc analytics. Export to S3 and query with Athena.
  • Athena treated as a database. It queries data that lives in S3.
  • OpenSearch as the system of record. It suits search and log analytics; keep the source data elsewhere.

Try the 1 practice question

Practice Choosing a data storeTake the free practice exam

Question 1

Easier ·Link to this question (question 1)

A pet insurer's claims portal shows each customer the status of a claim by claim ID. The page must respond in single-digit milliseconds at peak times, and the portal runs about 3,000 lookups per second.

Separately, the actuarial team runs analytics across five years of claim history, and it works in SQL with Athena and Redshift. The two workloads must not slow each other down.

Which storage design meets these requirements?

Choose one answer.

  1. A. Use Amazon Redshift for both workloads, and point the claims portal at a Redshift cluster with a sort key on the claim ID.
  2. B. Store the claim history only in Amazon S3, and let the claims portal query it with Amazon Athena for each lookup.
  3. C. Use DynamoDB keyed by claim ID for the portal, and keep the claim history in S3 for Athena and Redshift analytics, loaded by export or stream.
  4. D. Use Amazon Aurora PostgreSQL for the portal and for the analytics, with read replicas for the actuarial team's reporting.
Show the answer and explanation

Answer: C

  • A. Wrong: Redshift is built for analytic scans and concurrent reporting queries, not for high-volume single-row lookups at millisecond latency. Sort keys help range filters but do not make point reads fast enough for the portal.
  • B. Wrong: Athena queries scan data and take seconds to start, which is too slow for the portal's page loads. The option meets the analytics need but fails the latency requirement.
  • C. Right: Correct. DynamoDB serves single-item reads by key at consistent millisecond latency, which suits the portal. S3 holds the history for Athena and Redshift, so analytics runs without touching the portal's table. Learn more: Best practices for designing and architecting with DynamoDB (opens in a new tab)
  • D. Wrong: Aurora can serve the portal, but the actuarial scans over five years of history would compete with the portal on the same engine. The option does not separate the two workloads as the team asked.

Explanation

Each workload fits a different store. DynamoDB gives single-digit millisecond key lookups for the portal, and S3 with Athena and Redshift handles the multi-year analytics without loading the portal's table. Putting both on one analytic store or one relational engine lets each workload slow the other.

Study

Want more on Choosing a data store? The 5 full-length practice sets have 11 more Choosing a data store questions, 325 in all, none of them in the free exam. One payment of $14.99, no subscription. See the practice sets

Study Choosing a data store

More Data Store Management topics