They have to be stored somewhere, and they do indeed persist after a reboot (it would be pretty bad if every schedule job was lost on a restart...), and they should be included in backups and restores, which are the hub database. So, they're probably stored in the database, though that's technically an implementation detail we don't really know or normally need to. In any case, this is a possible symptom of a corrupt one, with a soft reset and restore (or just a restore, though I don't know if it works that way in the current hub platform version) doing some "clean up" that should fix things if whatever the hub does on its own doesn't help.
Download the Hubitat app
