[c-nsp] service-policy on virtual interface

[Peter Rathlev]

On Wed, 2009-09-09 at 08:18 -0500, Pete Templin wrote:
> Tony wrote:
...
> >> will progressively discard more. If the queue gets to 70% then ALL
> >> COS0 traffic will be discarded. In queue 1 this 30% at the top of
> >> the queue is for COS 2 & 3 packets in threshold 2.
> > 
> > This theoretically means that the maximum throughput you would see
> > on the above card with just enabling QOS and not configuring
> > anything is around 700Mbps (70% of 1Gbps). in reality it would
> > probably be a bit less, but that would be absolute maximum
> 
> Can you clarify this for me?  If 999Mbps of traffic is arriving on an 
> 'ingress' port and is the only traffic leaving on an 'egress' port, 
> there shouldn't be much if any queueing, right?

Correct. The 70% WTD doesn't prevent the interface from carrying 1 Gbps.

> Queueing should only be occurring when the (instantaneous) packet
> arrival rate exceeds the rate at which the device can dispatch the
> packets, and as those queues get to various thresholds of %-full,
> they'll execute WRR to manage the queue depth, correct?

I think you mean RED instead of WRR, but yes.

This means that even an interface with no buffers could throw out full
line rate if the traffic arrives correctly.

It is of course still a problem to have a default WTD of 70% for
CoS/DSCP 0 traffic, which in OPs case is all traffic. The switch is
cannot handle bursts as well and you waste 30% of the expensive RAM used
for buffering.

Depending on burstiness of traffic this might or might not be a problem.
But for most practical purposes, enabling QoS and adjusting the first
WTD for the default queue (typically queue 1) up to 100% should give the
same results as no QoS.

-- 
Peter