Guide on how to customize services using extensions/plugins

[duribe-vibrent]

In the documentation, there is a mention that it would be possible to use other data stores by using some extension points, but there are no details on how this could be accomplished. I found the class that implements the interface for Cassandra, but not sure what would need to be done to do it with an RDBMS like MySQL.

Also, is there an alternative to Hazelcast? We don't have that in our infrastructure, but we do have Zookeeper and Redis (which also offers distributed locks).

[sandeepmalik]

Hi @duribe76, Good point. I'll update the documentation to that effect. Essentially, you will need to implement two interfaces defined in this file.
https://github.com/walmartlabs/bigben/blob/master/lib/src/main/kotlin/com/walmartlabs/bigben/entities/EntityProvider.kt

And put that class as "domain" module in
https://github.com/walmartlabs/bigben/blob/master/app/src/main/resources/bigben.yaml

As far as Hazelcast is concerned, BigBen uses it in embedded mode so there's really no infrastructure required to set it up separately. There're three places where Hz is used:

1. Cluster management and singleton lock (ZK can be used)
2. Distributed counter (for sharding) (redis/memcache/Couchbase can be used)
3. Distributed executor for parallel processing / triggering of events across the cluster (Spark, Ignite, etc can be used).

So, there're three different functions and Hz allows us to tackle each one of them in embedded mode without any external dependency, so that's why we chose it.