September 2018
Intermediate to advanced
398 pages
9h 43m
English
In order to process the messages exchanged between the producers and the consumer, Kafka defines three main components—topics, partitions, and offsets.
A topic groups messages of the same type for streaming. It has a name and a number of partitions. You can have as many topics as you want. As Kafka can be distributed on multiple nodes, it needs a way to materialize the stream of messages on these different nodes. This is why the message stream (topic) is split into multiple partitions. Each partition contains a portion of the messages sent to a topic.
Each node of the Kafka cluster manages several partitions. A given partition is assigned to several nodes. This avoids losing data if a node is lost, and allows ...
Read now
Unlock full access