There is a tradeoff between the cost of introducing another component, and the advantages of a perfect component. If you already have an ACID database at hand,

There is a tradeoff between the cost of introducing another component, and the advantages of a perfect component. If you already have an ACID database at hand, and your queue requirements are not that big, using the database and NOT having to have someone around who knows exactly how Rabbit MQ was set up is better than to require everyone around you to know the piece of specialized software that you happen to be a domain expert on. Conversely if your needs are demanding enough, then it is best not to let your team discover the hard way why people build dedicated queueing systems. If you are unable to accept that this tradeoff even exists, then I wouldn't want to be your team lead.