I’ve heard about graph databases but never used one in prod before. I recently grabbed a book on Neo4j and then tested migrating my data into it. I have to say, I’m impressed by how much easier and faster it is to construct many of the queries I’m interested in using Cypher.
I have no experience operationalizing Neo4j, and a lot of questions…
Are there any pitfalls I should be wary of? If anyone has experience with both running Neo and Postgres in prod, what are the trade offs? Does it make sense to convert all the application data into Neo, or try to run it side by side with Postgres? if you go with two data stores, where should you draw the line between them?
Thanks for any experience or input anyone can share. I’ve been googling on this for a while and haven’t found many good resources.
You might be able to simplify some of the queries using recursive SQL: https://en.wikipedia.org/wiki/Hierarchical_and_recursive_que...
Neo4J is a very popular DB (ranks #18 here https://db-engines.com/en/ranking) so should be mature.
Side by side would allow you to formulate queries depending on the use case, but of course this comes with quite some cost (keeping data in sync, having another system to take care of).
Neo4J allows you to store properties, so you certainly could model all your info (also the relational aspects) in it.
Hope this helps!
Good luck with what you are up to !