Distributed job scheduler is, in a larger sense, a workflow engine. It will need to understand what is the next job to trigger, handle different situation where job fails\timeouts and do the retry if specified.