Skip to main content

Databases

Families​

Relational Databases (RDBMS)​

These use structured tables with fixed schemas, rows, and columns, queried via SQL. They excel in applications requiring complex joins and strong consistency.

  • MySQL / MariaDB
  • PostgreSQL
  • Oracle
  • Microsoft SQL Server

NoSQL Databases​

These provide schema-less or flexible data models for handling large volumes of unstructured or semi-structured data, often distributed across clusters.

They are subdivided into several types / subcategories.

Document Model (Document-Oriented)​

Stores data as semi-structured documents (e.g., JSON, BSON), ideal for hierarchical data in content management or real-time apps.

  • MongoDB
  • CouchDB
  • RavenDB

Key-Value Stores​

Simple databases mapping keys to values, optimized for high-speed lookups like caching or session storage.

  • Redis
  • Amazon DynamoDB
  • Riak

Wide-Column Stores (Column-Family)​

Organize data in flexible columns within rows, optimized for read/write on massive datasets, suited for large-scale analytics.

  • Apache Cassandra
  • HBase
  • ScyllaDB

Graph Databases​

Model data as nodes and edges for relationship-heavy use cases like social networks or fraud detection.

  • Neo4j
  • ArangoDB
  • JanusGraph

Design principles​

ACID​

PropertyRequirement
AtomicityThe writes in a transaction are executed all at once and cannot be broken into smaller parts. If there are faults when executing the transaction, the writes in the transaction are rolled back. So atomicity means “all or nothing”.
ConsistencyUnlike “consistency” in CAP theorem, which means every read receives the most recent write or an error, here consistency means preserving database invariants. Any data written by a transaction must be valid according to all defined rules and maintain the database in a good state.
IsolationWhen there are concurrent writes from two different transactions, the two transactions are isolated from each other.
DurabilityData is persisted after a transaction is committed even in a system failure. In a distributed system, this means the data is replicated to some other nodes.

CAP theorem​

CharacteristicRequirement
ConsistencyEvery read receives the most recent write or an error.
AvailabilityEvery request received by a non-failing node in the system must result in a response.
Partition toleranceThe system continues to operate despite an arbitrary number of messages being dropped (or delayed) by the network between nodes.