Intro

Formally: Collection of vertices and edges

Less formally: set of nodes and relationships connecting them

Expressive structure allowing modelling of all kidns of scenarios e.g. space rocket construction, system of roads, supply chain

• nodes and relationships

• nodes contain properties (key value)

• Nodes can have one ore more labels

• Relationships are named and directed

• Relationships always have a start and end node

• Relationships can contain properties

• Graph Databases

  • Used for transactional online graph persistence

  • Support real-time access from applications

  • Analogous to OLTP systems in relational databases

• Graph Compute Engines

  • Used for offline graph analytics

  • Perform batch processing

  • Comparable to OLAP and data mining tools

GraphDBMS: Online dbms with CRUD methods exposing a graph model

Generally built for use with OLTP systems -> optimized for transactional performance -> engineered with transactional integrity and operational availability in mind

Storage: Some use native graph storage optimized for storing and managing graphs. Some serialize graph data into a relation or object oriented db

Processing engine:

• Some use index-free adjacency (nodes directly point to neighbors) for better performance.

• This is called native graph processing.

  
  

• Performance: Excellent for handling connected data compared to relational and other NoSQL databases.

• Flexibility: Easily add new relationships, nodes, and labels without breaking existing queries.

• Agility: Schema-free model supports iterative development and API-driven testing.