Recommendation for implementing a distributed cache in a Vert.x application?

[Nic Sheard]

I'm having great success with Vert.x in various applications so thanks all.

I'm now looking at how best to implement a distributed cache in Vert.x.  I've seen various posts about this being on the roadmap and see that a distributed Map is listed under Vert.x 3.0.  Trouble is I would like to implement something very soon and was wondering if anyone has already done something similar or has any recommendations?

This would be for a high throughput distributed decision engine.  The idea would be to scale through multiple vert.x instances distributed across servers (~4) each running multiple verticle instances.  The whole lot will sit behind a loadbalancer handling updates and decision queries based on an underlying persistent data structure.  There will be relatively few updates compared to queries so a distributed cache is going to be important to maximise efficiency.

I'm thinking possibly of experimenting with a simple non-distributed cache, overlayed with a distributed cache entry flush mechanism triggered on updates.  If so potentially I could implement the distributed cache flush either with the event bus or using NetServer.  Any thoughts on what would be most appropriate or other alternatives?  I would like to keep the whole thing simple and light weight, and potentially migrate to whatever is implemented in the core in the future... 

Cheers,

Nic

[Krzysztof Makowski]

Have you thought about using redis?

[Jamie B]

Hazelcast - the same technology used in Vert.x for EventBus and Shared Data.

[Nic Sheard]

Thanks for the suggestions - I'm already using redis pub/sub elsewhere, and I'm not so familiar with Hazelcast but seems like a good candidate - time to do a little experimenting I think!