Yes, I understood that from reading the article, the point is that they still wasted an hour in trying to see if that would fix it or not because the symptoms were similar (crash of the master GFS server).