I need to know if under severe congestion i.e 100Meg LAN to 64K WAN if its at all possible for the packets going from the LAN onto the Bottlenecked WAN to not even reach the Output Buffer or even the TX-Ring on the interface So that the IOS is unable to determine the marking on the packets and thus not give the packet its required QOS?
The reason I am asking is that exact senorio above 100Meg LAN to 64K WAN.
Traffic that was supposed to get 30% of the bandwidth was extremely slow and output drops was very high
What you are describing is exactly what happens under congestion. Under congestion conditions, packets will first be queued to the configured software queues, and then de-queued from there into the tx-queue.
If you are using CBWFQ on the interface, that will kick in and determine packet scheduling order before the packets reach the tx-queue.
We ran into a similar problem when we were doing some testing at the Cisco Proof-of-Concept lab at the San Jose campus. We were actually only blasting about 8 megs of non-stateful traffic along with voice through a little T1 WIC in a 2600, and we were dropping traffic inconsistent with the service-policy. The show policy-interface displayed the correct behavior, but we were dropping voice and BGP, which should have received priority and guaranteed bandwidth, respectively.
The Cisco lab guys got one of their premier QoS experts on the line, and he explained that at a certain point, you will simply overrun the WAN interface hardware. The bigger the blade/card, the more you will be able to blast it. But for the smaller cards, you can have the most beautiful QoS policy ever created put into place, but if you blast it with more and more traffic, at some breaking point it will just start dropping a bit of everything. He had seen it many times before, and it helped us to know that it wasn't a misconfig or bug.
Our solution was to police inbound at the LAN interface that was supplying the huge volume of traffic, and it worked like a champ. The outbound interface was still receiving more traffic than it could send across the WAN link, but due to the inbound policing that had occurred on the LAN interface, it was not overrun beyond processing capacity.
We are pleased to announce availability of Beta software for 16.6.3.
16.6.3 will be the second rebuild on the 16.6 release train targeted
towards Catalyst 9500/9400/9300/3850/3650 switching platforms. We are
looking for early feedback from customers befor...
Introduction Featured Speakers Luis Espejel is the Telecommunications
Manager of IENova, an Oil & Gas company. Currently he works with Cisco
IOS® and Cisco IOS XE platforms, and NX to some extent. He has also
worked as a Senior Engineer with the Routing P...
In this session you can learn more about Layer 3 multicast and the best
practices to identify possible threats and take security measures. It
provides an overview of basic multicast, the best security practices for
use of this technology, and recommendati...