When you create a MapR Streams stream, it logically exists within the cluster and grows, as needed, to span as many servers in the cluster, as needed. This is all transparent to a user. For example, there's not the concept of assigning partitions to brokers. Generally, this is a advantageous because, if necessary, one can limit the servers for a given stream using volume topology.
You can read more here
High Speed Kafka API Publish Subscribe Streaming Architecture: How it works at the message level | MapR
MapR Streams Concepts
Retrieving data ...