[c-nsp] ASR9k Bundle QoS in 6.0.1

[Saku Ytti]

On 12 May 2016 at 14:16, Robert Williams <Robert at custodiandc.com> wrote:

> However - This text below, which is in the updated 6.0.1 release notes, sounds remarkably like the QoS policy can now be applied for a bundle in an 'aggregate' nature:

Sightly relevant. How could such feature work, in principle, if
different ports are offering significantly different rates of traffic?
Imagine aggregate shaper being 2Gbps, NPU1 at LC1 is being offered
1.5Gbps and NPU2 at LC2 is being offered 1.5Gbps. Which will drop, and
how much? And how will they know it?
Only thing I can think of is NPU exporting counters over EOBC and
other NPU adding it to its own counters, but this seems immensely
fragile and reaction time would be very slow (considering some devices
have buffers in microseconds, how will you guarantee at leasts 100s of
microseconds reaction time, so that you won't burst more than far-end
can handle?)

>
> http://www.cisco.com/c/en/us/td/docs/routers/asr9000/software/asr9k_r6-0/general/release/notes/reln-601a9k.html#concept_D31CCC2886D343A499A6D458B9980E97
>
> Aggregated Bundle QoS
> Aggregated Bundle QoS feature allows the shape, bandwidth, police rates and, burst values to be distributed between the active members of a bundle, where a QoS policy-map is applied.
> For information on Aggregated Bundle QoS, see the chapter Configuring Modular QoS on Link Bundles in the Cisco ASR 9000 Series Aggregation Services Router Modular Quality of Service Configuration Guide.




-- 
  ++ytti