We have a few issues wwith QoS on our BIP link between the Carrier's MPLS and our Call Centre.
1. There is no QoS service policy applied to the ATM VC interfaces of the BIP routers in both HQ WAN Router (Link 2 in Diagram)and the HA WAN Router (Link 1)even though the service-policy output command was entered. This is because the service policy being applied attempts to reserve more bandwidth than the maximum allowed (75%). This results in the command being accepted but not applied to the interface.
2. The QoS service policy on the BIP router in the Call Center WR (Link 4) is applied to the physical GigabitEthernet interface (not the subinterface) and no shaping is configured. This is likely to result in queuing not taking place as the physical interface is almost certainly never congested.
3. The WAN-OUT service-policy configured on all the BIP routers would (if applied) oversubscribe the class of service provisioned on all 3 BIP circuits. The BIP circuits have Premium or Premium 2 class of service, which only allow for 60% and 75% (respectively) of reserved bandwidth (in line with Cisco default maximum of 75%).
4. The classification and markings applied by the LAN-IN service-policy configured on all the BIP routers do not match the class of service of BIP circuits. This could (and more than likely does) result in some packets being either remarked or dropped by the inbound service-policy on the PE routers.
Link 1 Premium Class
Link 2 Premium 2 Class
Link 3 10MB Leased Line
Link 4 Premium Class.
I need help doing this so I can learn alot about QoS.
I suspect that I will have you re-write the policy so that they are not over subscribed. I think some of the links may be set to be more than 60% and 75%.
Is this just a matter of changing the % so that it totals 60% and 75%.
I hope I am not barking in the wrong direction.