Comment by sknzl

If you start scaling and running more than one instance, each instance would run the same periodic jobs. Shouldn't periodic jobs be off loaded to a queue and worker(s) to process to the messages?