NoSQL Cheat Sheet

Updated 2026-04-21

Next Topic: NoSQL Data Modeling Patterns Cheat Sheet

🧠Study flashcards on this topic129 cards · spaced repetition→

NoSQL databases are non-relational data stores designed for massive volumes of unstructured, semi-structured, and structured data with horizontal scalability and flexible schemas. Born from the needs of web-scale companies like Google, Amazon, and Facebook, NoSQL systems embrace schema flexibility, eventual consistency, and diverse data models — document stores, key-value pairs, wide-column families, graph structures, and specialized types like time-series and vector databases. As of 2026, the NoSQL ecosystem continues to evolve with Redis 8.0 unifying its data platform, Cassandra 5.0 introducing Storage Attached Indexes and vector search, MongoDB 8.0 enhancing queryable encryption, and the GQL ISO standard (ISO/IEC 39075) establishing a universal graph query language. Understanding when to use NoSQL over SQL, how different consistency levels impact performance, and which database type fits your access patterns is crucial for building modern, scalable applications.

What This Cheat Sheet Covers

This topic spans 17 focused tables and 151 indexed concepts, 129 flashcards. Below is a complete table-by-table outline of this topic, spanning foundational concepts through advanced details.

Table 1: NoSQL Database TypesTable 2: MongoDB Document Database OperationsTable 3: Redis Key-Value OperationsTable 4: Cassandra Wide-Column OperationsTable 5: Neo4j Graph Database QueriesTable 6: DynamoDB OperationsTable 7: Elasticsearch Search OperationsTable 8: Data Modeling PatternsTable 9: Consistency LevelsTable 10: CAP Theorem and PACELCTable 11: Replication and ShardingTable 12: Indexing StrategiesTable 13: NoSQL vs SQL When to UseTable 14: Security FeaturesTable 15: Performance OptimizationTable 16: Backup and RecoveryTable 17: Advanced NoSQL Features

Quick IndexSubscribe to unlock

A jump-to index of every table row in this cheat sheet.

Mind MapSubscribe to unlock

An interactive map of every table and concept in this topic.

Table 1: NoSQL Database Types

"NoSQL" isn't one thing — it's a family of databases each built around a different data shape and access pattern. Knowing the categories is the most important decision you'll make, because picking the wrong model fights you forever: document stores for flexible records, key-value for raw speed, wide-column for huge write volumes, graph for relationships, and the specialized time-series, vector, and search engines that have become essential in the AI era.

Type	Example	Description
Document Store	`MongoDB: db.users.find({age: {$gt: 30}})`	• Stores data as JSON/BSON documents with nested structures • ideal for content management, catalogs, and user profiles where schema varies per record.
Key-Value Store	`Redis: SET user:1001 "John"` `GET user:1001`	• Maps unique keys to values (strings, hashes, lists, sets) • fastest NoSQL model for caching, session storage, and real-time counters with sub-millisecond latency.
Wide-Column Store	`Cassandra: SELECT * FROM users` `WHERE user_id = '123'`	• Organizes data into column families with flexible columns per row • excels at time-series data, event logging, and IoT workloads requiring high write throughput.
Graph Database	`Neo4j: MATCH (a)-[:KNOWS]->(b)` `RETURN a, b`	• Stores nodes and relationships as first-class citizens • optimized for traversing connections in social networks, fraud detection, and recommendation engines.
Time-Series Database	`InfluxDB: SELECT mean(cpu) FROM system` `WHERE time > now() - 1h`	• Specialized for timestamped data with efficient compression and aggregation • designed for metrics, IoT sensors, and monitoring systems.
Vector Database	`Pinecone: index.query(vector=[0.1, ...], top_k=10)`	• Stores and queries high-dimensional embeddings for similarity search • powers AI applications like semantic search, RAG systems, and recommendation engines.
Search Engine Database	`Elasticsearch: GET /logs/_search` `{"query": {"match": {"message": "error"}}}`	• Indexes text and structured data using inverted indexes for full-text search with BM25 relevance scoring • built for log analysis, application search, and observability.
Multi-Model Database	`ArangoDB: AQL across document, graph, key-value`	• Supports multiple data models (document, key-value, graph) in one engine • reduces infrastructure complexity for applications needing varied access patterns.

Table 2: MongoDB Document Database Operations

MongoDB is the most widely used document database, storing data as flexible JSON-like documents you query without a fixed schema. These are the everyday operations — inserting, finding, updating, deleting — plus the two that unlock real power: the aggregation pipeline for multi-stage analytics, and $lookup for joining collections when you do need related data.

Command	Example	Description
insertOne	`db.users.insertOne({name: "Alice", age: 28})`	• Inserts a single document into a collection • creates the collection automatically if it doesn't exist.

The rest of this section is part of the full cheat sheet

Table 3: Redis Key-Value Operations

Redis is far more than a simple key-value cache — it's an in-memory data structure server with strings, hashes, lists, sets, sorted sets, and streams, all operated on with single-purpose commands. That's why one tool covers caching, session storage, leaderboards (ZADD), atomic counters (INCR), queues (LPUSH), pub/sub messaging, and event logs (XADD), with Redis 8.0 adding native JSON and Lua scripting for atomic multi-step logic.

Command	Example	Description
SET	`SET session:abc123 "user_data" EX 3600`	• Stores a string value for a key with optional expiration • overwrites existing value if key exists.

The rest of this section is part of the full cheat sheet

Table 4: Cassandra Wide-Column Operations

Cassandra's CQL looks reassuringly like SQL, but the model underneath is very different — everything revolves around the partition key, which decides where data lives across the cluster, and queries that don't include it are slow or forbidden. Inserts and updates are the same upsert operation with last-write-wins, deletes leave tombstones, and Cassandra 5.0's Storage Attached Indexes finally make secondary lookups practical.

Command	Example	Description
CREATE KEYSPACE	`CREATE KEYSPACE store WITH replication =` `{'class': 'NetworkTopologyStrategy',` `'dc1': 3}`	• Defines a namespace for tables with replication strategy • `NetworkTopologyStrategy` recommended for production multi-datacenter deployments.

The rest of this section is part of the full cheat sheet

Table 1: NoSQL Database Types

Type	Example	Description
Document Store	`MongoDB: db.users.find({age: {$gt: 30}})`	• Stores data as JSON/BSON documents with nested structures • ideal for content management, catalogs, and user profiles where schema varies per record.
Key-Value Store	`Redis: SET user:1001 "John"` `GET user:1001`	• Maps unique keys to values (strings, hashes, lists, sets) • fastest NoSQL model for caching, session storage, and real-time counters with sub-millisecond latency.
Wide-Column Store	`Cassandra: SELECT * FROM users` `WHERE user_id = '123'`	• Organizes data into column families with flexible columns per row • excels at time-series data, event logging, and IoT workloads requiring high write throughput.
Graph Database	`Neo4j: MATCH (a)-[:KNOWS]->(b)` `RETURN a, b`	• Stores nodes and relationships as first-class citizens • optimized for traversing connections in social networks, fraud detection, and recommendation engines.
Time-Series Database	`InfluxDB: SELECT mean(cpu) FROM system` `WHERE time > now() - 1h`	• Specialized for timestamped data with efficient compression and aggregation • designed for metrics, IoT sensors, and monitoring systems.
Vector Database	`Pinecone: index.query(vector=[0.1, ...], top_k=10)`	• Stores and queries high-dimensional embeddings for similarity search • powers AI applications like semantic search, RAG systems, and recommendation engines.
Search Engine Database	`Elasticsearch: GET /logs/_search` `{"query": {"match": {"message": "error"}}}`	• Indexes text and structured data using inverted indexes for full-text search with BM25 relevance scoring • built for log analysis, application search, and observability.
Multi-Model Database	`ArangoDB: AQL across document, graph, key-value`	• Supports multiple data models (document, key-value, graph) in one engine • reduces infrastructure complexity for applications needing varied access patterns.

Table 2: MongoDB Document Database Operations

Command	Example	Description
insertOne	`db.users.insertOne({name: "Alice", age: 28})`	• Inserts a single document into a collection • creates the collection automatically if it doesn't exist.

Table 3: Redis Key-Value Operations

Command	Example	Description
SET	`SET session:abc123 "user_data" EX 3600`	• Stores a string value for a key with optional expiration • overwrites existing value if key exists.

Table 4: Cassandra Wide-Column Operations

Command	Example	Description
CREATE KEYSPACE	`CREATE KEYSPACE store WITH replication =` `{'class': 'NetworkTopologyStrategy',` `'dc1': 3}`	• Defines a namespace for tables with replication strategy • `NetworkTopologyStrategy` recommended for production multi-datacenter deployments.