AWS Certified Solutions Architect – Associate Databases Syllabus

What authentication does RDS Proxy support and is it publicly accessible?

It supports IAM authentication and enforces credentials via AWS Secrets Manager. RDS Proxy is never publicly accessible; it must be accessed from within a VPC.

What is Amazon DynamoDB?

A fully managed, serverless, key-value and document NoSQL database delivering single-digit-millisecond performance at any scale, with automatic scaling and built-in high availability across 3 AZs.

What is a DynamoDB partition key vs a composite (partition + sort) key?

A partition key alone uniquely identifies an item via its hash. A composite key uses partition key + sort key, allowing multiple items per partition key, sorted by the sort key, enabling range queries.

What is the maximum size of a single DynamoDB item?

400 KB, including attribute names and values.

Difference between a DynamoDB Local Secondary Index (LSI) and Global Secondary Index (GSI)?

LSI shares the partition key but uses a different sort key and must be created at table creation (consumes table capacity). GSI has a different partition and sort key, can be created anytime, and has its own provisioned throughput.

What are DynamoDB's two capacity modes?

Provisioned (you specify RCUs/WCUs, with optional auto scaling) and On-Demand (pay-per-request, auto-scales instantly for unpredictable workloads).

How is one DynamoDB Read Capacity Unit (RCU) defined?

1 RCU = 1 strongly consistent read per second of an item up to 4 KB, OR 2 eventually consistent reads per second up to 4 KB. Larger items round up to the next 4 KB.

How is one DynamoDB Write Capacity Unit (WCU) defined?

1 WCU = 1 write per second for an item up to 1 KB. Larger items round up to the next 1 KB.

What is DynamoDB Accelerator (DAX)?

A fully managed, in-memory cache for DynamoDB that delivers microsecond read latency for cached reads, requires minimal code changes, and is purpose-built only for DynamoDB.

DAX vs ElastiCache for caching DynamoDB: when to use each?

Use DAX for caching individual DynamoDB queries/objects with microsecond latency and no app changes. Use ElastiCache to store aggregation/computation results or share cached data across multiple data sources.

What are DynamoDB Streams?

An ordered, time-ordered flow of item-level modifications (create/update/delete) in a table, retained 24 hours, often consumed by Lambda for triggers, replication, or analytics.

What is DynamoDB Global Tables?

A multi-Region, multi-active (multi-writer) replication feature giving fast local read/write access across Regions, with last-writer-wins conflict resolution; requires DynamoDB Streams enabled.

What is DynamoDB TTL (Time To Live)?

A feature that automatically deletes expired items based on a timestamp attribute (epoch seconds), reducing storage cost and cleanup work; deletions are free and asynchronous.

What is Amazon ElastiCache?

A fully managed in-memory caching service supporting Redis (and Redis OSS/Valkey) and Memcached, used to reduce database load and provide microsecond/sub-millisecond latency.

ElastiCache Redis vs Memcached: key differences?

Redis supports replication, Multi-AZ failover, persistence, backups, pub/sub, sorted sets, and transactions. Memcached is multi-threaded, supports simple sharding, has no persistence, replication, or failover.

What is the cache-aside (lazy loading) caching strategy?

The app checks the cache first; on a miss it reads from the database and writes the result into the cache. Only requested data is cached, but cache misses incur extra latency and data can become stale.

What is the write-through caching strategy?

Data is written to the cache and the database at the same time on every write, keeping the cache fresh, but it adds write latency and caches data that may never be read.

How does ElastiCache Redis provide high availability?

Through Multi-AZ with automatic failover using a primary node and read replicas, plus optional cluster mode for sharding across multiple node groups.

What is Amazon Neptune?

A fully managed graph database service supporting property graph (Gremlin/openCypher) and RDF (SPARQL) models, ideal for highly connected data like social networks, fraud detection, and recommendation engines.

What is Amazon DocumentDB?

A fully managed document database that is MongoDB-compatible, designed for JSON document workloads with Aurora-like storage that scales automatically.

What is Amazon Keyspaces?

A managed, serverless, Apache Cassandra-compatible database service for wide-column workloads, using CQL, with single-digit-millisecond performance and automatic scaling.

What is Amazon QLDB?

Amazon Quantum Ledger Database: a fully managed ledger database with an immutable, append-only, cryptographically verifiable transaction log, ideal for systems of record/audit trails (single trusted authority).

What is Amazon Timestream?

A fully managed, serverless time-series database optimized for IoT and operational metrics, with built-in tiered storage (memory + magnetic) and time-series analytics functions.

What is Amazon MemoryDB for Redis?

A Redis-compatible, durable, in-memory database delivering ultra-fast performance with Multi-AZ durability, usable as a primary database (not just a cache).

What is AWS Database Migration Service (DMS)?

A managed service that migrates databases to AWS with minimal downtime; the source database remains available during migration. It supports homogeneous and heterogeneous migrations.

What is the difference between homogeneous and heterogeneous migrations in DMS?

Homogeneous = same engine (e.g., Oracle to Oracle). Heterogeneous = different engines (e.g., Oracle to Aurora PostgreSQL), which requires the AWS Schema Conversion Tool (SCT) to convert the schema/code first.

What is the AWS Schema Conversion Tool (SCT) and when is it needed?

A tool that converts a source database schema and code (procedures, views) to a target engine. It is needed for heterogeneous migrations; it is not needed for homogeneous (same-engine) migrations.

What are the DMS migration types (replication tasks)?

Full load (one-time copy of existing data), CDC (change data capture for ongoing replication of changes), and Full load + CDC (migrate existing data then continuously replicate changes for near-zero downtime).

How should you choose a database for a relational, ACID-transaction workload with complex joins?

Use a relational database: Amazon RDS or Amazon Aurora (Aurora for higher performance/availability and cloud-native scaling).

Which AWS databases fit key-value, document, graph, in-memory, time-series, ledger, and wide-column needs respectively?

Key-value/document: DynamoDB. Document (MongoDB): DocumentDB. Graph: Neptune. In-memory: ElastiCache/MemoryDB. Time-series: Timestream. Ledger: QLDB. Wide-column (Cassandra): Keyspaces.

How is encryption at rest and in transit handled for RDS/Aurora, and what manages the keys?

Encryption at rest uses AWS KMS keys (must be enabled at creation; can't encrypt an existing unencrypted DB directly—restore an encrypted snapshot). Encryption in transit uses SSL/TLS. IAM database authentication lets you connect using IAM tokens instead of passwords (Aurora/RDS MySQL & PostgreSQL).