Showing results for 
Search instead for 
Did you mean: 

Shaping per VLAN with Q-in-Q between IAD and Cat 3560

We have a number of IAD 2435's deployed at remote locations using Metro E at the client site.  We're running Q-in-Q, so each client site comes in as a VLAN tag on a trunk port that is terminated on a Catalyst 3560 at our central location.  Let's assume the client site each has 10 Mbps of bandwidth, capped by the service provider providing the Metro E Q-in-Q service.  All port speeds are 100/Full.

I need to know the best way to shape both inbound and outbound for voice for the client.  I know that I can shape outbound traffic on the IAD using a parent/child policy.  That policy is below and seems to be the best solution for IAD outbound traffic.  Can you confirm that this is best and that I need to shape it below 10 Mbps to ensure the child policy is applied?

IAD 2435:

policy-map AutoQoS-Policy-Trust

class AutoQoS-VoIP-RTP-Trust

  priority percent 70

class AutoQoS-VoIP-Control-Trust

  bandwidth percent 5

class class-default


policy-map SubrateEthernet

class class-default

  shape average 9000000

  service-policy AutoQoS-Policy-Trust

interface FastEthernet0/0

description Metro Ethernet

ip address x.x.x.x x.x.x.x

speed 100


service-policy output SubrateEthernet

Second, what is the best way to handle the inbound traffic to the IAD?  The Catalyst 3560 can only do per-vlan policing on an SVI, and I believe that I can only do inbound policing on the IAD.  So, I assume my only option is to police data traffic to ~8 Mbps to ensure that there's always ~2 Mbps reserved for voice?

CreatePlease to create content
Content for Community-Ad
July's Community Spotlight Awards