rabbitmq trigger with invalid queueName creates connections that persist

[mauvilsa]

Report

When a scaledObject is created with rabbitmq trigger and queueName pointing to a queue that doesn't exist in that rabbitmq, the keda-operator pod increasingly creates new connections that persist. Eventually a limit can be reached and new connections to rabbitmq are rejected.

Expected Behavior

When a rabbitmq queue doesn't exist, the scaledObject fail and close any opened connections to rabbitmq.

Actual Behavior

The scalerObject fails periodically, and the number of connections to rabbitmq keep increasing until creating new connections fail.

Steps to Reproduce the Problem

1. Create a scaledObject with proper details to connect to rabbitmq
2. Set a queueName to a queue that doesn't exist in that rabbitmq
3. Wait for failures and see in the rabbitmq web ui how number of connections keeps increasing

Logs from KEDA operator

ERROR scale_handler error getting metric for trigger {"scaledObject.Namespace": "...", "scaledObject.Name": "...", "trigger": "rabbitMQScaler", "error": "error inspecting rabbitMQ: Exception (404) Reason: \"NOT_FOUND - no queue '...' in vhost '/'\""}
ERROR scale_handler error getting scale decision {"scaledObject.Namespace": "...", "scaledObject.Name": "...", "scaler": "rabbitMQScaler", "error": "error establishing rabbitmq connection: dial tcp ...: i/o timeout"}

KEDA Version

2.14.0

Kubernetes Version

1.30

Platform

Microsoft Azure

Scaler Details

rabbitmq

Anything else?

No response

[JorTurFer]

Hello
Are you using AMQP or HTTP?

[mauvilsa]

Hello Are you using AMQP or HTTP?

amqps://