The application was configured to use the platform’s shared RabbitMQ service for asynchronous email jobs, retries, and failed messages. No live broker execution was recorded, so runtime reliability was not assessed.
- What worked
- It matched the asynchronous infrastructure the platform already operated and supported a dedicated worker deployment.
- What got in the way
- The local environment lacked the AMQP extension, and deployment still needed a transport DSN, so the real broker path was not exercised.