One page per concern, each covering the fields you set, what the operator does with them, and the constraints that will bite you if you ignore them.
| Page | Covers |
|---|---|
| 1. Standalone | A single instance: the default shape, and what it cannot do |
| 2. Clustering | Primary and secondary pools, formation, scaling, database allocation |
| 3. Storage | Data and auxiliary volumes, existing claims, extra mounts, retention |
| 4. Connectivity | Listeners, Services, service types, in-cluster and external access |
| 5. Security | Passwords, the opt-in Secret contract, TLS, pod hardening |
| 6. Configuration | neo4j.conf, JVM arguments, APOC, Neo4j logging |
| 7. Plugins | APOC, Graph Data Science, Bloom, and where they may run |
| 8. Monitoring | Prometheus metrics and ServiceMonitor |
| 9. Operations | Sizing, placement, probes, restarts, maintenance, deletion |
If you have not deployed anything yet, start with Your first Neo4j; these pages assume you have a working instance to modify.
The topology mode is immutable. spec.topology.mode cannot change after creation, so a
Standalone instance never becomes a cluster. Choose deliberately, and read
Standalone if you are unsure.
Some settings belong to the operator. Anything the operator has to control to make Kubernetes
work — listen addresses, advertised addresses, discovery, paths — is derived from the spec and
overrides what you put in spec.config.neo4j. The most important ones are rejected outright at
admission so you find out immediately. The full list is in
Operator-owned settings.
The topic pages explain the fields worth explaining. For an exhaustive list, with types, defaults and validation rules, use the API reference.