I have just read an interesting article about patterns for handling errors in Kafka. The article presents four patterns for handling errors in Kafka applications:

  • Stop on error
  • Dead Letter Queue
  • Retry topic and retry application
  • Maintain order of redirected events

I have already used the spring-kafka @RetryableTopic annotation where it handles the creation of the retry and dead letter topics automatically for Java and Kotlin backend applications. It is very useful and saves a lot of time in these scenarios.

How do you handle errors in Kafka applications? Drop you comments below about what you do in these situations.

Tags: , , , , , ,