Question on how does a QoS policy mapped to a sub-interface know what the logical bandwidth is? For example, if I have a parent policy that is shaping traffic at a rate of 1GB on a 10GB link and has a nested policy with bandwidth guarantees for multiple classes how does it know to calculate that the logical bandwidth is 1GB vs 10GB when it’s time to guarantee bandwidth percents?
If you don’t configure bandwidth under the parent policy then the interface bandwidth is divided by each class under the parent policy. For example:
class-map match-all CHILD_CLASS class-map match-all PARENT_CLASS ! policy-map CHILD class CHILD_CLASS policy-map PARENT class PARENT_CLASS shape average 700000000 service-policy CHILD interface GigabitEthernet0/1 service-policy output PARENT
There is no
bandwidth command under the parent policy so the
class-default classes would have a minimum guarantee of 500 Mbit (50% of Gigabit interface). The
PARENT_CLASS shapes up to 700 Mbit.
On a 10 Gigabit interface with a 1GB shaper, you should set the bandwidth to 1 GB so the child policy inherits 1GB as the bandwidth.
Does this help?
It does help and thank you.