What are the reasons why Kafka messages do not get lost?

There are several reasons why Kafka ensures messages are not lost:

1. Persistence storage: Kafka uses persistence storage to write messages to disk, ensuring that messages will not be lost after being sent.
2. Redundant backup: Kafka utilizes a replication mechanism to copy messages from each partition to multiple brokers, ensuring that at least one replica is available. In the event of a broker failure, data can be retrieved from other replicas to prevent message loss.
3. Batch sending: Kafka allows multiple messages to be sent to the server in batches, reducing network transmission costs and disk write times, improving message reliability.
4. Sequential write: Kafka writes messages to disk in a sequential manner, avoiding the performance issues associated with random disk writes.
5. Message Replication Confirmation Mechanism: Kafka uses a replication confirmation mechanism to ensure that messages are successfully written to multiple copies. The producer will only receive a confirmation once all copies have been successfully written.
6. Batch fetching on the client side: Kafka clients can fetch multiple messages in batches, reducing network transmission overhead and improving the reliability of messages.
7. Quickly duplicating: Kafka utilizes zero copy technology for message duplication, reducing the number of data copies between memory and disk, improving duplication efficiency and message reliability.

In conclusion, Kafka ensures message reliability by implementing various mechanisms like persistent storage, redundant backups, batch processing, sequential writing, confirmation replication, client batch fetching, and fast copying, aiming to minimize the risk of message loss.