QoS in the Core?

creamy_stew

I just started reading the BCMSN study guide and in the first chapter they say that "Advanced QoS" happens at the core layer. What do they mean by that? I thought the idea was to keep the Core as basic as possible, avoiding anything that might slow down the raw packet/l2 switching?

szmarcek

I agree that the Core should be as basic as possible.


According the Cisco official exam guide the qos classification should be done as close to the edge as possible – preferably at VOIP phone or access layer switch.


The reason for trust boundary being at the VOIP phone, access layer or distribution layer (if necessary) is that we do not want every device which will process the packets to go over classification process because it is very costly for processors etc. So once the packets have been classified (at the trust boundary) other devices will only mark packets accordingly to what policies they have set up - that is how I understand it.
I am not sure what features they mean by advanced qos.

joshgibson82

creamy_stew wrote: »

I just started reading the BCMSN study guide and in the first chapter they say that "Advanced QoS" happens at the core layer. What do they mean by that? I thought the idea was to keep the Core as basic as possible, avoiding anything that might slow down the raw packet/l2 switching?

I haven't read exactly what you're referring to, but I would imagine they mean sometimes more advanced things like traffic shaping occur at the core of a campus because the WAN circuits connect there, and the company may shape, police, etc. at that layer of the network due to the huge bandwidth bottleneck that typically occurs on WAN circuits.

Hope this helps...