Traffic Shaping with MQC/HQF

This is how we configure traffic shaping on Cisco IOS..

To learn about Traffic Shaping Terminology click here.
To learn about MQC/HQF click here.
To learn how to use Iperf on Windows click here.

Ethernet interfaces on diagram are Gig interfaces actually in this lab scenario because I am doing it on actual devices and the diagram is drawn in EVE-NG.

MQC/HQF configuration is pretty straight forward..

I am going to configure a policy on R1;

R1(config)#policy-map SHAPE
R1(config-pmap)#class class-default
R1(config-pmap-c)#shape average 200m

I am going to shape all the traffic because this is a lab. In real world you may need to configure a match criteria other than class-default.

The policy map says to shape the traffic to 200 Mbps..

Before applying it to an interface, let's see how the traffic going through naturally..

Seems like about 350-550 Mbps traffic is going through..

Now let's apply the policy map on Gig0/1 interface of R1;

R1(config)#int g0/1
R1(config-if)#service-policy output SHAPE

Now as you can see, the traffic is not exceeding 200 Mbps..

If we take a look at a I/O graph of Wireshark for both the scenarios, captured on Server interface;

The policy applied can be seen by show policy-map interface g0/1

Peak Shape Configuration

I am removing the shape average 200m command from the policy map and replace it with shape peak 200m command..

The thing to remember in peak shaping is that the value you enter on the command is not the value of PIR (Peak Information Rate). It's the value of CIR.

So the policy map will be like the following..

R1(config)#policy-map SHAPE
R1(config-pmap)#class class-default
R1(config-pmap-c)#shape peak 200m

As you can see the Target Shape Rate is 400 Mps now because b default Bc and Be takes equal values. Throughput measurements from Iperf will be like the following..