RabbitMQ Interview Questions

1. What is RabbitMQ and what problems does it solve?

2. What is a message broker and why do systems use one?

3. What is the difference between RabbitMQ and Apache Kafka?

4. What is the AMQP protocol and how does RabbitMQ use it?

5. What are exchanges in RabbitMQ and what types exist?

6. What is a direct exchange and when do you use it?

7. What is a fanout exchange and how does it broadcast messages?

8. What is a topic exchange and how does routing key pattern matching work?

9. What is a headers exchange in RabbitMQ?

10. What are queues and bindings in RabbitMQ?

11. What is the difference between a producer and a consumer in RabbitMQ?

12. What is message acknowledgement (ack/nack) in RabbitMQ and why does it matter?

13. What is a dead letter exchange (DLX) and when do you use it?

14. How does RabbitMQ handle message durability and persistence?

15. What is the difference between a durable and a transient queue?

16. What is prefetch (QoS) in RabbitMQ and how does it affect throughput?

17. How does RabbitMQ handle message ordering?

18. What is a work queue (task queue) pattern in RabbitMQ?

19. What is the publish/subscribe pattern in RabbitMQ?

20. What is the request/reply (RPC) pattern in RabbitMQ?

21. How does RabbitMQ handle high availability with mirrored and quorum queues?

22. What is RabbitMQ clustering and how does it work?

23. How do you handle message retries and back-off in RabbitMQ?

24. What is the difference between at-most-once, at-least-once, and exactly-once delivery?

25. How does RabbitMQ handle flow control and back pressure?

26. What are RabbitMQ virtual hosts and why are they useful?

27. How do you monitor and manage RabbitMQ in production?

28. How do you secure RabbitMQ with authentication and TLS?

29. When should you choose RabbitMQ over Kafka or a database queue?

30. What are common RabbitMQ best practices and anti-patterns?

31. How do you handle poison messages using the quorum queue delivery limit?

32. How do you use asynchronous publisher confirms without losing throughput?

33. How do you diagnose and fix connection and channel churn in RabbitMQ?

34. What are RabbitMQ memory and disk alarms, and how do you tune them?

35. How does RabbitMQ cope with very long queues and paging messages to disk?

36. How does RabbitMQ detect and handle network partitions in a cluster?

37. When do you use the Shovel plugin versus Federation in RabbitMQ?

38. What are RabbitMQ Streams and how do they differ from classic and quorum queues?

39. What is single active consumer in RabbitMQ, and how does it differ from consumer priorities?

40. How do you test RabbitMQ producers and consumers reliably?

41. How do you plan and execute a rolling upgrade of a RabbitMQ cluster without dropping traffic?

42. How do you migrate live classic mirrored queues to quorum queues without losing messages?

43. A publisher gets confirms but the messages never arrive. How do you detect unroutable messages?

44. Consumers keep getting disconnected with a delivery acknowledgement timeout. How do you handle long-running tasks?

45. A broker node restarts. What must a client library do to recover connections, channels and topology correctly?

46. After a node restart all quorum queue leaders end up on two nodes. How do you control leader placement?

47. Your service needs to send large payloads through RabbitMQ. How should you design for that?

48. You need to change a queue's TTL and dead-letter settings, but redeclaring fails with PRECONDITION_FAILED. What do you do?

49. A specific message never reached its consumer. How do you trace it through the broker?

50. A team wants priority queues so urgent jobs jump the backlog. What are the operational trade-offs?

51. One team's service keeps destabilising a shared RabbitMQ cluster. What controls do you put in place?

52. How do you back up a RabbitMQ cluster, and what does disaster recovery actually look like?

53. What does a basic.nack from the broker on a publisher confirm channel actually tell you?

54. What happens to unacknowledged messages when a consumer's channel or connection closes?

55. Why does an unbounded prefetch let one consumer starve the rest, and how do you size it?

56. Why do messages sometimes loop endlessly between a dead-letter exchange and the source queue, and how do you stop it?

57. How do queue length limits and overflow behaviour work, and which overflow mode should you choose?

58. What happens to a quorum queue when it loses quorum, and how does that differ from a classic queue failing?

59. How should an application share connections and channels across threads, and what does churn cost?

60. A queue keeps growing in production. How do you diagnose it systematically?

61. When should you use basic.get polling instead of basic.consume, and why is polling usually wrong?