RabbitMQ

 rabbitMQ Publishing Issue

Mohamed Mohsen's profile image
Mohamed Mohsen posted Jan 04, 2025 06:44 AM
We are using rabbitMQ (bitnami) with the following details:
  • RabbitMQ Version: Bitnami 3.12.14
  • Number of Nodes: 3 (hosted on the OCP platform)
  • Type of queues is classic mirrored.
  • At certain time, the application client asynchronously published several messages to the RabbitMQ system but did not receive acknowledgment (Ack) messages.
  • After 3 minuets, the application client logged that the messages were negatively acknowledged (N-Ack). However, the messages were successfully published on the system and consumed by another service. Despite this, a false rejection message was communicated to the client based on the N-Ack signals.
  • Simultaneously, we observed a memory leak on the leader node managing the relevant queue. This memory issue caused the leader node to crash loop at same time, even though the hosting platform (infrastructure) had sufficient available resources.

We would appreciate your assistance in identifying the root cause of this issue and any recommendations for preventing similar issues in the future.