Found that the "differential protection" circuit for the ops center was tripped when I got here at 1830. From what I see in the data files, this might have happened about 1630, with ustar losing data (but of course, the DSMs stilll haven't been rsync'd).