IEOC - Internetwork Expert's Online Community

Hi,

I noticed in the solutions for Task 7.2 that 'qos-groups' have been used on R2 and R3.

My understanding was that these were only needed on the PE routers as the MPLS EXP cannot be matched outbound to the CE as it is stripped, hence need to match on a qos-group for any outbound to CE policy.

Can you explain why qos-groups were used on R3 and R2 when they could have just simply matched on the topmost MPLS EXP bit outbound? Also they are not used at all on the PE routers R4 and R6 where I believe they should be in order to apply correct outbound policy

Also, I noticed that FRTS was used to shape rather than drop but again implementation is inconsistent:

• R1 has shaping outbound on all its FR interfaces
  
• R2 has shaping configured to both R1 and to R3.
• R3 has no shaping configured anywhere.
• R2 has no shaping configured to R6.
• R6 has no shaping configured outbound to R2.

Could you provide some clarification of the reasoning behind the sporadic use of FRTS here - I would have thought all should use it to meet the question requirements (and on the non-FR links to use 'shape' command in CBWFQ)?

Thanks in advance

Just noticed that the old forum entries have been moved to the IEOC so please ignore the above - these issues have been raised and answered in the relevant section of the old forums

Hi PKG,

Can you post the link of old discussion about the topic?

thanks

Dev

Here is the link: http://ieoc.com/forums/t/766.aspx

Brian's answer is the following: "There can be cases where the transport label has a different EXP value than the VPN label, so using the QoS group ensures that you are doing the QoS based on the transport label."