Have you used DLQ/DLT topics? How were retries organized — synchronously or asynchronously?

Yes, in projects using Kafka, we employed DLQ (Dead Letter Queue) for handling messages that could not be processed correctly after several attempts. Messages that caused errors were sent to a separate DLQ topic for subsequent analysis and manual processing.

Regarding retries, we implemented asynchronous retry attempts. When an error occurred, the message did not block the main processing thread but was re-queued with a delay (for example, using delayed topics or timers) to try processing it later. This approach reduced load and prevented system blockages.