Autoscale workers depending on number of messages in queues
Also dedicate loaders per queue
Related to T3592