Is Kafka push or pull?

Is Kafka Pull-based or Push-based

Apache Kafka is often associated with message queues like AMQP or Rabbit MQ. However, Kafka architecture takes a different approach while defining how messages are flowed across the system. With millions of messages being served around the clock, its core design is based not only on a traditional Push/ Pull model but also on a distributed streaming platform structure, which makes it inherently capable of handling large sets of data.

Main Objective

The primary key performance indicator (KPI) for Kafka is exactly on-time delivery, considering exactly-once-semantics. One more crucial aspect is durability because Kafka stores messages independently using a distributed storage log technology.

Now is Kafka Push or a pull-based system? Pull model

Kafka fundamentally stands on a ‘PULL’ operation approach. It means consuming groups, or consumers’, initiate the process by requesting messages from specific consumer-group offsets. In various configurations, this process offers versatility and adaptability among nodes. This flexibility might enhance the system’s reaction pace and responsiveness. One clear advantage of the Poll and Consume pattern seen by Kafka is the prevention against data loss, or so-called ‘Failsafe’ capabilities since messages are stored once published.

How Message Requests Work

1.-In Kafka, a requesting poll is made by active brokers to fetch messages starting in a specified offset area known as the target set.
2.Proxies, also known
This means that the communication always starts from the Broker when a consumer wants pull any message.