Handling Billions of Edges in a Graph Database
summary
ArangoDB demonstrates a distributed approach that claims to handle billions of edges by combining smart graphs (domain-aware sharding that keeps most neighbors local) with vertex-centric indexes (range lookups that return the latest N neighbors without scanning millions) and a coordinator/DB server split. Traversals run in time proportional to local neighbor counts, edge deletes are constant-time, and the enterprise feature set aims to give near single-server traversal latency while avoiding excessive network hops across a large cluster.
AI-written summary · about
chapters (12)
- 0:00Intro
- 5:27Querying Edges
- 6:24Social Graphs
- 11:55Query Language
- 14:39Query Algorithm
- 21:26Social Graph
- 23:27Graph Query
- 26:48Supernodes
- 29:38Vertex centric indexes
- 31:53Querying your database
- 36:35Parameter View Cluster
- 37:38Summary
related talks
40mDistributed Data on Distributed Infrastructure
20mDomain Driven Design & NoSQL
54mDistributed Graph DBMSs — The Future for Analytics on Big Data?
27mGraphOne: A Data Store for Real-time Analytics on Evolving Graphs
38mDisaster Recovery with a Distributed Database
01h 00mThe Open-Source Distributed Graph Database: Nebula Graph
49mModern Graph Databases
29mAll about graphs in Database