I setup a traffic shaping/QoS policy to limit outbound SMTP traffic to 17% of the bandwidth of our T1 and to also divy up some other traffic into roughly equal parts. However it doesn't seem to be working as planned. Remote users of RDP protocol were complaining of slow response from the terminal servers although I have tried to set aside 25% of the BW for RDP class of traffic. Please see attached bandwidth graph of the T1 and also the output of "show service-policy" to see what I mean. The graph shows the T1 pegged for almost two hours and the "show service policy" shows the SMTP class apparently getting all the BW. What am I doing wrong? Should I be policing instead?