How to decide between Neo4j causal vs HA clustering

Question

I'm setting up a Neo4j cluster for a greenfield app. What considerations should I take into account when deciding between causal and HA clustering? The docs are very good at describing configuration of each, but not at how to decide which architecture to choose.

At least for now the cluster is only 3 nodes, so read vs. read/write nodes (as described in causal clustering) is not a factor.

Answer 1

There are some differences between HA & CC mode, but by default you should use the CC one : it's the default clustering mode (Neo4j spent a lot of time to develop it).

With CC you :

• will never have any branched data
• can benefit of the official driver functionalities : Load-balancing, routing, bookmark-id(ie. read your own write), ...

Cheers.