AWS Systems Architect Professional

AWS DocumentDB, Keyspaces, Neptune, and QLDB – SAP-C02 Study Guide

Learn how to choose Amazon DocumentDB, Keyspaces, Neptune, and QLDB from AWS architecture scenarios, including data models, scalability, availability, and exam traps.

AWS Systems Architect ProfessionalAWS Systems Architect ProfessionalUpdated Sep 1, 2026
Study options
WatchComing later
ListenComing later
ReadAvailable
ReviewComing later

Study guide

Technical reference and lesson notes

Purpose of This Lesson

AWS offers several specialized database services beyond Amazon RDS, Aurora, DynamoDB, and the main analytics platforms. For the exam, the most important skill is recognizing the data model or application requirement that points to a particular service.

This guide covers:

  • Amazon DocumentDB for MongoDB-compatible document workloads
  • Amazon Keyspaces for Apache Cassandra-compatible workloads
  • Amazon Neptune for graph relationships
  • Amazon Quantum Ledger Database (QLDB) for immutable, verifiable transaction histories

The service names and use cases matter more than detailed implementation syntax.

Key Concepts

Amazon DocumentDB

Amazon DocumentDB is a fully managed, MongoDB-compatible document database designed for JSON data and document-oriented applications.

Use it when a scenario includes:

  • MongoDB-compatible applications or migration requirements
  • JSON documents with flexible, nested structures
  • Document queries at large scale
  • A managed database that can scale storage without application disruption

Important characteristics include:

  • Storage automatically scales up to 64 TB.
  • Data is replicated across six copies in three Availability Zones.
  • The service is designed for four nines of availability.
  • Up to 15 low-latency read replicas can support read-heavy workloads.
  • AWS Database Migration Service (AWS DMS) can assist with MongoDB migrations.

DocumentDB is compatible with MongoDB APIs and tools, but compatibility should not automatically be interpreted as feature-for-feature identity with every MongoDB deployment. Confirm application compatibility during migration planning.

Amazon Keyspaces for Apache Cassandra

Amazon Keyspaces is a serverless, fully managed, Apache Cassandra-compatible database service. It supports Cassandra Query Language (CQL) and Cassandra-compatible application patterns.

Choose Keyspaces when the scenario emphasizes:

  • Apache Cassandra compatibility
  • High-volume, distributed key-value or wide-column workloads
  • Automatic scaling based on application traffic
  • Consistent single-digit millisecond performance at scale
  • Avoiding Cassandra cluster administration

Key characteristics include:

  • Serverless operation with no clusters to provision or manage
  • Automatic scaling of throughput and storage
  • Thousands of requests per second and virtually unlimited throughput and storage for supported workloads
  • Four nines availability within an AWS Region
  • CQL and Cassandra-compatible APIs

Keyspaces is not a general-purpose relational database. The access patterns, partition-key design, and query constraints associated with Cassandra still matter.

Amazon Neptune

Amazon Neptune is a fully managed graph database. It is designed to model and query relationships among entities rather than primarily storing independent records.

Typical use cases include:

  • Identity and relationship graphs
  • Knowledge graphs
  • Fraud detection
  • Recommendation or social relationship engines
  • Network and dependency analysis

Neptune supports graph technologies and APIs such as:

  • Apache TinkerPop Gremlin
  • openCypher
  • SPARQL

Important characteristics include:

  • Fault-tolerant, self-healing storage
  • Storage that grows in 10 GB increments up to 64 TB
  • More than four nines of availability as described in the service material
  • Up to 15 read replicas

The deciding factor is usually the data model: choose Neptune when traversing relationships is central to the application, not merely because the data happens to contain references between tables or documents.

Amazon Quantum Ledger Database (QLDB)

QLDB was designed as a fully managed ledger database with a transparent, immutable, and cryptographically verifiable transaction history. Its journal records changes in sequence and is append-only: entries can be added, but historical entries cannot be overwritten or deleted.

QLDB was intended for applications that need a trusted history of changes without implementing a distributed ledger network, such as tracking ownership, asset history, or financial records.

The journal uses cryptographic hashing, including SHA-256-based verification, to produce a concise summary of the change history.

#### Current-Service Caveat

Amazon QLDB was retired by AWS in 2025. It may still appear in older training material or legacy question banks, but it should not be selected for new architectures. For current certification preparation, verify whether the current exam guide includes QLDB and distinguish historical knowledge from currently available AWS services.

Exam-Relevant Takeaways

  • MongoDB or JSON documents should prompt consideration of Amazon DocumentDB.
  • Apache Cassandra or CQL should prompt consideration of Amazon Keyspaces.
  • Entity relationships, graph traversal, fraud graphs, or knowledge graphs should prompt consideration of Amazon Neptune.
  • Immutable, append-only, cryptographically verifiable history describes the historical QLDB use case.
  • DocumentDB, Keyspaces, and Neptune are managed services, but managed does not eliminate data-model and access-pattern design.
  • Read replicas are relevant to read scaling; they do not automatically solve write-scaling requirements.
  • AWS DMS can be used in migration scenarios involving MongoDB and DocumentDB.
  • Do not select a relational database simply because the scenario contains structured data. Identify the required data model first.

Architecture Decision Guide

Scenario requirementBest fitWhy
MongoDB-compatible JSON document workloadAmazon DocumentDBManaged document database with MongoDB compatibility
Large-scale Cassandra-compatible applicationAmazon KeyspacesServerless managed Cassandra-compatible service with CQL
Frequent traversal of relationships among entitiesAmazon NeptunePurpose-built graph database
Immutable historical record with verifiable changesQLDB in historical material onlyAppend-only journal and cryptographic verification; retired for new designs
Traditional joins, transactions, and relational constraintsAmazon Aurora or Amazon RDSThese specialized services are not relational databases
General key-value access with DynamoDB data-model patternsAmazon DynamoDBUse when Cassandra compatibility or graph/document requirements are absent

Common Exam Traps

  • Confusing DocumentDB with DynamoDB: DocumentDB targets MongoDB-compatible document workloads; DynamoDB is a different managed NoSQL service with its own data model and APIs.
  • Assuming MongoDB compatibility means complete compatibility: Migration questions may still require application testing and feature validation.
  • Choosing Neptune for any related data: Neptune is appropriate when graph traversal and relationship analysis are core query patterns.
  • Treating Keyspaces as a relational database: Cassandra data modeling is driven by known access patterns and partition keys, not arbitrary joins.
  • Assuming read replicas increase write capacity: Read replicas primarily scale read operations.
  • Ignoring regional availability wording: Keyspaces availability is described within an AWS Region. Do not infer that a single regional deployment automatically provides cross-Region disaster recovery.
  • Using QLDB for a new deployment: QLDB is historical knowledge now that the service has been retired.
  • Selecting a blockchain service for every audit requirement: An immutable audit trail and a decentralized, multi-party ledger are different architectural requirements.

Real-World Engineer Notes

  • Validate DocumentDB compatibility before migration, especially for driver behavior, unsupported MongoDB features, indexing, and query semantics.
  • Design Cassandra-compatible workloads around partition keys and expected queries before selecting Keyspaces. Automatic scaling does not repair poor partition design.
  • Neptune is most valuable when relationship queries would otherwise require complex joins or repeated application-side traversal.
  • Read replicas can improve read throughput and latency, but applications must use an appropriate read strategy and tolerate replica lag where applicable.
  • Service limits such as storage ceilings and replica counts should be checked in the current AWS documentation before production design.
  • For auditability, define retention, export, access control, and independent verification requirements explicitly. “Immutable” alone does not describe the complete compliance architecture.

Quick Reference Summary

  • DocumentDB: MongoDB-compatible JSON document database.
  • Keyspaces: Serverless, Cassandra-compatible wide-column database using CQL.
  • Neptune: Managed graph database for relationship-heavy applications.
  • QLDB: Former managed ledger database with append-only, cryptographically verifiable history; retired in 2025.
  • Migration: AWS DMS can support MongoDB-to-DocumentDB migration scenarios.
  • Scaling: DocumentDB uses replicas for read scaling; Keyspaces scales automatically; Neptune supports read replicas and elastic storage.

Flashcards

  1. Q: Which AWS service is associated with MongoDB-compatible JSON document workloads?

A: Amazon DocumentDB.

  1. Q: Which service supports Cassandra Query Language without managing Cassandra clusters?

A: Amazon Keyspaces for Apache Cassandra.

  1. Q: What is the primary data model of Amazon Neptune?

A: A graph model that represents entities and their relationships.

  1. Q: Name three graph technologies or APIs associated with Neptune.

A: Gremlin, openCypher, and SPARQL.

  1. Q: What is the defining characteristic of the historical QLDB use case?

A: An immutable, append-only, cryptographically verifiable transaction history.

  1. Q: How many Availability Zones are used for DocumentDB’s storage replication in the lesson material?

A: Three Availability Zones, with six copies of the data.

  1. Q: What AWS service can assist with MongoDB migration to DocumentDB?

A: AWS Database Migration Service (AWS DMS).

  1. Q: Does Keyspaces require administrators to provision and manage database servers?

A: No. It is serverless and fully managed.

  1. Q: What is the maximum number of DocumentDB read replicas described in the lesson?

A: Up to 15.

  1. Q: What is the key distinction between Neptune and a document database?

A: Neptune is optimized for traversing relationships and graph queries, while a document database stores self-contained JSON-like documents.

Practice Questions

Question 1

A company is migrating an existing application that uses MongoDB-compatible drivers and stores large, flexible JSON documents. The company wants a managed AWS database and wants to minimize application changes. Which service should the architect evaluate first?

A. Amazon Neptune
B. Amazon DocumentDB
C. Amazon Keyspaces
D. Amazon Aurora PostgreSQL

Correct answer: B. Amazon DocumentDB

Explanation: DocumentDB is designed for MongoDB-compatible document workloads and JSON data. Neptune is for graph workloads, Keyspaces is Cassandra-compatible, and Aurora is relational.

Question 2

A workload uses Cassandra Query Language and requires automatic scaling, consistent single-digit millisecond response times, and no Cassandra cluster administration. Which service best fits?

A. Amazon Keyspaces
B. Amazon DocumentDB
C. Amazon Neptune
D. Amazon Redshift

Correct answer: A. Amazon Keyspaces

Explanation: Keyspaces is a serverless, managed Apache Cassandra-compatible service that supports CQL and automatically scales for traffic.

Question 3

A fraud-detection application must identify indirect relationships among accounts, devices, transactions, and IP addresses. Queries frequently traverse multiple relationships. Which database is the best fit?

A. Amazon RDS for PostgreSQL
B. Amazon DocumentDB
C. Amazon Neptune
D. Amazon Keyspaces

Correct answer: C. Amazon Neptune

Explanation: Neptune is purpose-built for graph data and relationship traversal, making it appropriate for fraud, identity, and knowledge-graph scenarios.

Question 4

An architect is reviewing an old design that recommends Amazon QLDB for a new application requiring an immutable audit history. What should the architect do first?

A. Deploy QLDB across three Availability Zones
B. Replace QLDB with Amazon Keyspaces automatically
C. Verify current AWS service availability and evaluate an active alternative
D. Use Neptune because all audit data is graph data

Correct answer: C. Verify current AWS service availability and evaluate an active alternative

Explanation: QLDB was retired in 2025. Its historical use case involved append-only, cryptographically verifiable history, but it should not be selected for a new deployment. The replacement depends on the audit, compliance, and verification requirements.

Question 5

A team wants to improve the read performance of an Amazon DocumentDB deployment without changing the primary write path. Which capability is most directly relevant?

A. Adding read replicas
B. Converting documents into graph vertices
C. Using Cassandra Query Language
D. Enabling an append-only ledger journal

Correct answer: A. Adding read replicas

Explanation: DocumentDB supports read replicas for read scaling and low-latency access. Read replicas do not directly increase the primary write capacity.