FalkorDB
Replies: 4 comments
-
|
Just discovered the spring-data-falkordb repository while browsing through FalkorDB's Java projects - and it's way more complete than I expected 😄. After reading through the comprehensive README, I see that: Already Production-Ready:
In Progress:
This is MUCH further along than I realized! The migration path from Spring Data Neo4j looks very straightforward - entities and repositories can stay nearly identical, just configuration changes. This changes some of my questions above:
This changes my evaluation significantly - I can start experimenting with spring-data-falkordb immediately instead of building abstractions from scratch! |
Beta Was this translation helpful? Give feedback.
-
|
For what it is worth, I am using FalkorDB with Quarkus via the RedisDataSource extensions (which has Vertx as its starting point). It is brilliant! The RedisDataSource extensions work out of the box with FalkorDB although I did need to write a new wrapper to allow the use of some of the FalkorDB commands that arrived since Redis days, and to facilitate static / block requests for running tests. And, I will eventually rewrite the Vertx and Quarkus classes to ditch all the non-graph Redis stuff to make it a bit leaner and specific to FalkorDB (but that wont be any time soon). The only thing that I am "unhappy" about is the lack of. support for nested maps in FalkorDB (although that might be a more general CYPHER thing, I am not sure since this is my first foray into graph databases). I have some workarounds but am often reminded of this gap in ability which, unfortunately, is not on their roadmap. :-( Other than that, as I say, I think the Quarkus + FalkorDB combination is great. Cheers, |
Beta Was this translation helpful? Give feedback.
-
|
PS, in production, but only in a small way, and soonish will be expanding. |
Beta Was this translation helpful? Give feedback.
-
|
Quick Prototype Update: Spent a few hours testing spring-data-falkordb with my DevGraph app. Got the application running with either Neo4j or FalkorDB backend via Spring profiles. Created a simple abstraction layer + basic falkordb migration framework (inspired by neo4j-migrations). Basic CRUD and custom queries work well (I didn't change anything, same cypher statements - neo4j and falkordb). Discovered some integration issues along the way - happy to discuss if useful. Attaching logs showing migration and startup. Overall impressed with how much works already, though there's clearly more work needed on the Java/Spring side. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Hi FalkorDB team,
I'm exploring FalkorDB and want to understand the Java integration story before deciding if it's worth a deeper investment.
Context
I recently built DevGraph (https://github.com/decebals/devgraph) as a learning project while exploring Neo4j and graph databases. It's a prototype for analyzing developer collaboration patterns using Spring Boot + Spring Data Neo4j.
While DevGraph itself is just a playground, it gave me hands-on experience with graph modeling and made me curious about alternatives like FalkorDB - especially given your performance claims and GraphRAG focus.
I've been working with Java enterprise systems for 25+ years, so I'm approaching this from a practical architecture perspective: "Could I use FalkorDB for a production system?"
My Exploration Idea
I'm considering rebuilding DevGraph (same domain model: developers, projects, contributions) backed by FalkorDB instead of Neo4j. This would let me:
FalkorDB claims significant performance advantages, so having an apples-to-apples comparison from a Java application perspective could be valuable for the community.
The goal isn't just toy examples - I'd focus on patterns you'd actually use in production: proper transaction handling, connection pooling, error management, etc. The abstraction layer design would also make it straightforward to evaluate additional graph databases if the approach proves useful.
But first, I need to understand if Java/Spring Boot is even your intended audience.
Part 1: Java Ecosystem Strategy
Core Question: Where does Java fit in FalkorDB's vision?
Current State
Future Direction?
Is there interest in:
Community vs Official
My concern: I don't want to spend weeks building abstractions if FalkorDB's focus is primarily Python/JavaScript for AI workloads and Java is secondary.
Part 2: Technical Integration Questions
Assuming Java IS a valid path forward:
Driver & Connectivity
Spring Boot Integration Pattern
What's the recommended approach for Spring Boot applications?
Option A: Low-level JFalkorDB
Option B: Build abstraction layer
Option C: Wait for Spring Data FalkorDB (if planned?)
Data Migration
Part 3: GraphRAG vs General-Purpose
I notice FalkorDB heavily emphasizes GraphRAG and AI workloads.
Use Case Fit
A typical use case like DevGraph involves:
Question: Is FalkorDB optimized primarily for AI-driven queries, or is it equally strong for traditional graph analytics?
Positioning
Neo4j also added GraphRAG capabilities recently, but their core strength remains general-purpose graph databases.
Where does FalkorDB position itself:
This matters because if you're optimizing specifically for AI workloads, traditional queries might not see the claimed performance gains.
Part 4: Production Readiness
For anyone considering FalkorDB for production:
Multi-tenancy
High Availability
Monitoring & Operations
Memory Management
What I Can Contribute
If Java/Spring Boot is strategically important to FalkorDB, I'm happy to help:
Note: I've built open-source infrastructure before (PF4J with 2,600+ stars, used by Netflix/Facebook/Eclipse Foundation) so I understand ecosystem building.
Previous evaluation example: FerretDB/FerretDB#4161
Summary
What I need to understand:
What I can offer:
Thanks for your time! Looking forward to understanding where FalkorDB is headed.
Best,
Decebal
Beta Was this translation helpful? Give feedback.
All reactions