AWS DEA-C01 data store guide covering storage fit, catalogs, metadata, lifecycle, and schema decisions.
This domain is where AWS tests whether you can choose the right store, keep metadata usable, manage lifecycle transitions, and evolve schemas without breaking downstream systems.
AWS currently weights Data Store Management at 26% of scored content.
flowchart LR
A["Access pattern"] --> S["Choose serving store"]
S --> M["Register metadata and partitions"]
M --> L["Apply lifecycle and retention"]
L --> E["Handle schema evolution"]
E --> O["Optimize indexes, partitions, compression, and vector retrieval"]
The exam rarely asks “where can data be stored?” in isolation. It asks whether the store, catalog, lifecycle rule, schema model, and optimization technique match the way data is read, governed, retained, and changed.
| Lesson | Focus |
|---|---|
| 2.1 Choosing Data Stores for Access Patterns | Learn how performance, cost, access pattern, and migration requirements drive store selection. |
| 2.2 Catalogs, Crawlers & Metadata | Learn Glue Data Catalog, crawlers, partitions, and metadata-management patterns. |
| 2.3 Data Lifecycle, Loads & Retention | Learn load/unload flows, lifecycle policies, versioning, TTL, deletion, and resiliency controls. |
| 2.4 Data Models, Schema Evolution & Optimization | Learn schema change, conversion, lineage, partitioning, indexing, and vector-related optimization ideas. |
| Scenario clue | Strong first instinct | Weak answer pattern |
|---|---|---|
| Analysts need SQL over modeled warehouse data | Redshift, load/unload design, materialized views, Spectrum/federation where appropriate | DynamoDB or raw S3 as the only query layer |
| Data is in S3 but consumers cannot query it | Glue Data Catalog, crawler or controlled table definition, partition sync | Assume files alone are enough |
| Data must age into cheaper storage or expire | S3 Lifecycle, versioning, DynamoDB TTL, retention-aware deletion | Keep everything forever until costs spike |
| Source and target schemas differ during migration | SCT or DMS schema conversion with compatibility checks | Manual rewrite with no conversion or lineage plan |
| Semantic search or RAG retrieval appears | Vectorization, embeddings, vector index choice, metadata filters, access control | Treat vector search as ordinary SQL filtering |
Do not choose a store by brand familiarity. DEA-C01 store answers are access-pattern answers: latency, query shape, update model, retention, catalog visibility, and governance usually decide the service.