Aerospike Graph: Massively Scalable, High Performance
summary
Aerospike Graph targets real-time OLTP property graphs (identity graphs, fraud detection, recommendations) and claims storage for billions of vertices and trillions of edges with independently scalable compute and storage. Architecture uses an Apache TinkerPop-based graph service front end with a native Gremlin interface and a Spark bulk loader for high-throughput ingestion. Benchmarks claim sub-millisecond one-hop reads, under 3 ms two-hop reads, under 5 ms single-vertex updates and sustained 60k–100k graph QPS, with lower TCO versus in‑memory competitors.
AI-written summary · about
chapters (14)
- 0:00Aerospike Graph intro
- 2:22Graph database apps
- 3:50Realtime property graph
- 5:38Identity graph use case
- 10:18Why Aerospike Graph stands out
- 13:08Graph architecture overview
- 18:52Benchmarking performance
- 20:41Aerospike Graph features
- 21:49Q&A: OLAP vs OLTP
- 22:57Q&A: Data protection
- 24:40Q&A: GDPR and security
- 25:46Q&A: Product structure
- 27:13Q&A: Graph history
- 30:02Closing and booth
related talks
15mUnderstanding the Power of Graph Databases
08mBuilding a Real-time Data Future
28mAerospike's Journey Through Innovations, Sustainability, and Roadmap
43mLarge Scale Online Graph on Aerospike at PayPal Risk
31mAerospike's Innovation Agenda
48mThe real-time database to build your AI future
50mAerospike Technical Overview
49mJourney with Aerospike: Architecture Best Practices