Neo4j

Every relational database can store a foreign key. Ask it to walk three hops of them at scale and watch the query degrade — that concrete failure is the whole reason graph databases exist, and this chapter makes it real with numbers before writing a line of Cypher.

The graph model from Chapter 01 earns its keep here — variable-length paths, shortest-path search, and the two query shapes ("recommend via shared connections," "find the fraud ring") that show up in real systems more than almost any other graph query.

A graph database's traversals are cheap by default (Chapter 01) — but finding your very first node still needs help. This chapter is indexes, uniqueness constraints, and reading a query plan well enough to know which one you're missing.

Where Cypher meets Spring Boot: mapping nodes and relationships onto Java objects with Spring Data Neo4j, and the honest limit of that mapping — the point where a real traversal query needs raw Cypher instead.