had a similar situation and i killed the jobs then rebuilt them the exact same way.  worked fine afterwards. 

may be an issue in the SQL builds the node after...