Content Comparison

NoSQL datastore when	Relational database when
High volume of workloads (need latency in milliseconds while performing millions of transactions per second)	Workload volume fits within thousands of transactions per second
Data is dynamic and frequently changes	Data is highly structured and requires referential integrity
Relationships can be de-normalized data models	Relationships are expressed through table joins on normalized data models
Data is expressed without table joins	Data with complex queries and reports
Data is replicated across geographies and requires finer control over performance, availability and consistency.	Data is centralized or can be replicated regions asynchronously
For commodity hardware (app deployed in public clouds)	High-End hardware (app deployed to large and high-end hardware)

Key-Value DB	Store session info User preferences or settings Shopping carts Product recommendations News Feed Publishing & Retrieval …	Dynamo DB, Redis, BerkleyDB.
Document Stores	User Profiles to store different types of info Management of content (video, images, comments, etc.) …	MongoDB, Couch, Lotus Notes.
Wide-Column Store	User preferences or settings Business Intelligence Managing data warehouses Reporting systems …	Google’s Bigtable, Cassandra, HBase.
Graph DB	Social networking site Recommendation engine Logistics Risk assessment Fraud detection	Neo4j, OrientDB, Allegrograph.

Versions Compared