RabbitMQOperator task stuck on queued when using CeleryExecutor as core executor

[abdalhalimalzohbi]

Apache Airflow Provider(s)

redis

Versions of Apache Airflow Providers

airflow-provider-rabbitmq 0.6.1

Apache Airflow version

2.10.2

Operating System

airflow bitnami image

Deployment

Official Apache Airflow Helm Chart

Deployment details

No response

What happened

When using CeleryExecutor as the core executor for Airflow and utilizing the RabbitMQOperator task from rabbitmq_provider.operators.rabbitmq, the task remains stuck in the queued state and is not recognized by Celery workers. The issue arises when attempting to use RabbitMQOperator to publish messages to an external RabbitMQ queue, which is not related to the Airflow instance.

The root cause of the conflict is that the RabbitMQOperator takes a queue field as a parameter, and this value is mistakenly treated as the name of the queue that the Celery workers will use to execute the task. This creates confusion, as the queue parameter is intended to represent the destination queue for message publishing, not the execution queue for Celery tasks.

What you think should happen instead

No response

How to reproduce

NA

Anything else

No response

Are you willing to submit PR?

• [ ] Yes I am willing to submit a PR!

Code of Conduct

• [X] I agree to follow this project's Code of Conduct

[boring-cyborg[bot]]

Thanks for opening your first issue here! Be sure to follow the issue template! If you are willing to raise PR to address this issue please do so, no need to wait for approval.

[eladkal]

airflow-provider-rabbitmq is not official Airflow provider. There is no RabbitMQOperator in the Apache-Airflow project. Unless you can explain what is the bug in Airflow code base I suggest to submit this report to the maintainers of airflow-provider-rabbitmq