Hazelcast Cache Connector

Learn how to use Hazelcast with deepstream

* Hazelcast not upgraded to V4/V5. Please refer to this github issue if you require this *

What is Hazelcast?

Hazelcast is a distributed caching layer, organized as a grid of independent nodes that sync their state. On top of that, Hazelcast allows to perform computations based on the stored data and even supports basic server side messaging.

Why use Hazelcast with deepstream?

Hazelcast can be a good choice as a fast and scalable caching layer for deepstream. It can outperform other caches like Redis in cluster-mode, but is a bit trickier to set up. Where Hazelcast really comes into its own is the additional Map-Reduce functionality built on top of its caching capabilities. This makes it possible to perform simple, distributed computations based on record data and feed the results back to deepstream.

Hazelcast Diagram

Using Hazelcast with deepstream

deepstream comes with a cache connector for Hazelcast preinstalled

Hazelcast install console output

or, if you’re using deepstream in Node, get it from NPM

Configuring the Hazelcast connector

You can configure the Hazelcast cache connector in the plugins section of deepstream’s config.yml file. Please find a full list of configuration options here

plugins:
    cache:
      name: hazelcast
      options:
        networkConfig:
          addresses:
            - host: hostname
              port: 1234