How does Kafka work?


Answers

J
Jeanie.Ullrich

Kafka is ultimately software you run on a cluster of commodity hardware servers so that you can efficiently stream and store data in real time.

You write applications called producers which publish messages to topics. These topics are a log stream of messages grouped together like tables would be in a MySQL database.

You write separate applications called consumers which read from these topics. A group of consumers reads from a given topic in parallel with other consumer groups.

Kafka keeps data around for as long as you specify via a retention policy.

A
Andres_Kerluke

Kafka is a messaging system. Applications called "producers" send messages to specific "topics". Other applications called consumers then read from these topics.

A
Aurore.DAmore87

Kafka works as a messaging queue.

Consumers read from topics

publishers write to topics

C
Clementina_Ziemann68

loaded question ...check the docs

V
Verner_Langosh

The key to Kafka is how the log stream is partitioned and distributed across a cluster of machines. With only one consumer in a given group exclusively reading from a topic, Kafka offers parallel processing of message consumption.