What message brokers have you worked with? Compare Kafka and RabbitMQ

I have worked with message brokers such as Apache Kafka and RabbitMQ.

Kafka is a distributed data streaming platform focused on high throughput and scalability. It is well-suited for processing large volumes of data in real-time, provides fault tolerance, and stores messages in topics with the ability for re-reading.

RabbitMQ is a message broker based on queues and the AMQP protocol. It is convenient for organizing asynchronous interactions between services, supports complex message routing, guarantees delivery, and is suitable for systems with more traditional queue logic.

Main differences:

• Kafka is oriented towards high speed and storing data streams, RabbitMQ towards flexible routing and guaranteed delivery.
• Kafka stores messages in topics with re-reading capability, RabbitMQ deletes messages after acknowledgment.
• Kafka scales better horizontally for large loads, RabbitMQ is easier to set up for classic queues.

Example of using Kafka in Go with the sarama library:

producer, _ := sarama.NewSyncProducer(brokers, nil)
msg := &sarama.ProducerMessage{Topic: "topic", Value: sarama.StringEncoder("message")}
producer.SendMessage(msg)

Depending on the task, I choose the appropriate broker.