Some issues have been detected on the manager, our teams are trying to correct the problem.
Update(s):
Date: 2015-12-02 15:25:40 UTC Optimization at the TCP of the web farm seem to have corrected the problem.
We will continue to monitor and to investigate.