How did data flow through Kafka? How did events arrive?

In Kafka, data is transmitted in the form of events (messages) that are published to topics. Each topic is divided into partitions, and events within partitions are ordered by offsets. Producers send messages to a specific topic, and consumers subscribe to topics and read events sequentially, tracking offsets.

Events are received by consumers via a pull model: consumers request new messages, specifying the offset from which they want to read. This allows them to control which events have already been processed and to re-read data if necessary.

Example in Go using the sarama library:

consumer, err := sarama.NewConsumer(brokers, nil)
if err != nil {
    panic(err)
}
partitionConsumer, err := consumer.ConsumePartition("topic", 0, sarama.OffsetNewest)
if err != nil {
    panic(err)
}

for msg := range partitionConsumer.Messages() {
    fmt.Printf("Received message: %s\n", string(msg.Value))
}

Thus, the data stream through Kafka is organized as a sequence of messages in topics, and events are delivered to consumers upon request with control over the read position.