There's a little bit of similarity in a few of the Restart actions in the API. Here's the breakdown:
+ RestartMedia: This will completely restart the entire job, including all tasks.
+ RestartMediaErrors: This action will retry specific tasks within the request that produced an error. For example, if you received a 404 error on a file and remedied the issue, this action will successfully rerun only that task.
+ RestartMediaTask: This action requires an additional parameter. You can specify to rerun a specific task within a MediaID. For more information on the difference between MediaIDs and TaskIDs, please go here.