Specialist DBs: AWS SA - Associate

Amazon DocumentDB

Managed MongoDB-compatible database. JSON documents. Auto-scales to 128TB.

DocumentDB Cluster Endpoint

Points to primary instance for read/write operations.

DocumentDB Reader Endpoint

Load balances reads across replicas. Up to 15 replicas.

DocumentDB Replication

Data replicated across 3 AZs automatically.

DocumentDB Use Case

MongoDB workloads or JSON document storage migrating to AWS.

Amazon Neptune

Managed graph database. Stores nodes and relationships, not tables.

Neptune Use Case

Social networks, recommendation engines, fraud detection, knowledge graphs.

Neptune Query Languages

Supports Gremlin, openCypher, and SPARQL.

Neptune Streams

Captures graph changes in order. No duplicates. Access via REST API.

Amazon Keyspaces

Managed Apache Cassandra service. Serverless, pay per use.

Keyspaces Use Case

Cassandra migrations or big data distributed NoSQL workloads.

Keyspaces Query Language

Uses CQL (Cassandra Query Language).

Amazon QLDB

Immutable ledger database. Append-only, cryptographically verifiable.

QLDB vs Blockchain

QLDB = centralised ledger (not blockchain). Faster than blockchain frameworks.

QLDB Use Case

Financial transactions, supply chain, audit trails, system of record.

QLDB Query Language

Uses PartiQL.

Amazon Timestream

Serverless time-series database. Stores trillions of events per day.

Timestream Storage Tiers

Recent data = memory tier. Historical data = cheaper magnetic tier.

Timestream Use Case

IoT sensor data, server metrics, stock prices, application monitoring.

Timestream Integrations

IoT Core, Kinesis, QuickSight, Grafana, SageMaker.

Amazon Redshift

Managed data warehouse. OLAP/analytics only. NOT for transactions.

Redshift Architecture

Leader node coordinates queries. Compute nodes execute them.

Redshift Storage

Columnar storage. Better for analytics than row-based storage.

Redshift Serverless

AWS manages capacity automatically. No cluster management needed.

Redshift Spectrum

Query S3 data directly without loading into Redshift. Uses Glue Data Catalog.

Redshift vs RDS

Redshift = analytics/reporting (OLAP). RDS = transactions (OLTP).

Redshift vs Athena

Redshift = load data into warehouse first. Athena = query S3 directly, no loading.