Horizontal scaling has a lot of benefits, like higher throughput and fault tolerance, but your data needs to be 'big' enough to take advantage of them.
We look at the processing of stateful streaming data, state interfaces, and the implementation of state management and fault tolerance in Apache Flink.