so I am configuring L2 qos currently on one of our switches. what I want to know is the following
If I use priority-queue out on an interface I know that the srr-queue bandwidth shape command will have no effect, and that in times of congestion the prioirty queue will be serviced before anything else.
Now, if I have a share policy like this 'srr-queue bandwidth share 10 20 50 20' and I use 'srr-queue bandwidth limit 20' on a 1 gig interface (which means I am limiting the link to 200mb)
Does this mean that -
Queue 1 - prioritiy queue will always get serviced first
Queue 2 - will have 20% bandwidth reserved in times of congestion provided that priority queue is not congested. bandwidth = 20% of 200mb = 40mb
Queue 3 - will have 50% bandwidth reserved in times of congestion provided that priority queue is not congested. bandwidth = 50% of 200mb = 100mb
Queue 4 - will have 20% bandwidth reserved in times of congestion provided that priority queue is not congested. bandwidth = 20% of 200mb = 40mb
I also know with the config above if I change the srr-queue bandwidth share command to 'srr-queue bandwidth share 50 100 250 100' it is the same as using 10 20 50 20. The way I calculate this is Queue_Weight_qX/(q1+q2+q3+q4) * Bandwidth. It gives same results, no surprise there really.
The only other quesiton I have is this ---
If I dont set priority queue out on the interface but configure
srr-queue bandwidth limit 20 (1 gig port)
srr-queue bandwidth shape 10 0 0 0
srr-queue bandwidth share 10 20 50 100
In theory this should be the outcome
200mb port due to limit bandwidth 20
for Queue 1 shape reservation - 1/10th of total bandwidth = 20mb - reserved at all times
Remaining bandwidth = 180mb
Shared amongst Q2,Q3,Q4 - now according to the weights I have assigned
Cisco DNA Software Demo Series - Cisco ThousandEyesRegister nowWednesday, May 12, 202110:00 am Pacific Daylight Time(San Francisco, GMT-07:00)SaaS applications and cloud-based services are increasingly critical for on-campus users, but they can be challen...