Database - How to choose?
Ref:
| Database Family | When Would You Use It? | Use Case | Optimized For | AWS Related Database Engines or Services | 3rd-party |
|---|---|---|---|---|---|
| Relational | Migrating on-premises relational workloads, ad hoc query patterns | Traditional applications, enterprise resource planning (ERP), customer relationship management (CRM), e-commerce | Structured data in tables, rows, and columns, supports complex ad hoc queries through joins | Amazon Aurora, Amazon RDS | MySQL, Postgres |
| Amazon Redshift | Google BigQuery | ||||
| Key-value | Session stores, shopping carts, scalable for high volumes of state changes | High-traffic web applications, e-commerce systems, gaming applications | Fast access to large amounts of structured, semi-structured, or unstructured data with high read and write throughput | Amazon DynamoDB | Google Bigtable |
| Document | Storing JSON documents with rich querying abilities | Content management, catalogs, user profiles | Storing semi-structured data as JSON documents | Amazon DocumentDB (with MongoDB compatibility) | MongoDB |
| In-memory | Improving read performance through caching | Caching, session management, gaming leaderboards, geospatial applications | Applications requiring sub-millisecond data access | Amazon ElastiCache, Amazon MemoryDB for Redis | Redis |
| Wide column | Migrating on-premises Cassandra workloads, high-speed data processing | High-scale industrial apps for equipment maintenance, fleet management, and route optimization | Workloads requiring heavy reads/writes, high throughput, low latency, and linear scalability | Amazon Keyspaces | Apache Cassandra |
| Graph | Fraud detection, recommendation engines, airline networks | Fraud detection, social networking, recommendation engines | Storing and navigating relationships | Amazon Neptune | neo4j |
| Time series | Managing large amounts of time series data from various sources | Internet of Things (IoT) applications, DevOps, industrial telemetry | Storing and querying data that changes over time | Amazon Timestream | |
| Ledger | Communicating with other entities, maintaining verifiable history of database changes | Systems of record, supply chain, registrations, banking transactions | Maintaining a complete and verifiable history of database changes | Amazon Quantum Ledger Database (QLDB) |
Structured, semi-structured, and unstructured data
| Structured data | Semi-structured data | Unstructured data | |
|---|---|---|---|
| Data pie | 10 to 15 % | 5% to 10% | 80% to 90% |
| Example | RDS table, spreadsheet | HTML, JSON, XML | Multimedia (photo, audio, video), text document, scanned PDF |
Ref: