Where are partitions distributed?

In systems that use partitioning (for example, Apache Kafka), partitions are distributed across the brokers (nodes) of the cluster. Each partition is a separate log stored on one of the brokers, which allows for scaling the load and ensuring fault tolerance.

Partition distribution is usually done automatically using built-in balancing algorithms, but it can also be configured manually. This helps to evenly distribute data and load across servers, improving system performance and scalability.