We have a 100Mbps WAN circuit connected to e1/1. We would like the queuing on this interface to follow the 4 queue model percentages we have setup on the chassis.
From my days working on 7K's we could simply set the bandwidth on e1/1 to 100Mbps and it would queue it correctly using that BW statement. ( i.e. when it approached 100Mbps it would start queuing to the percentages we had setup). Simple and straight forward. Does the 9K behave the same way? Or will we need to jump through different hoops?
We looked at traffic shaping but the min we can set for NXOS is 100Mbps on the shaping, so we cannot traffic shape each queue based on our needs.
Would setting the interface BW statement help at all here? Does the 9K queuing even pay attention to it? Basically we need our critical traffic to be guaranteed the percentages in the queuing model.
Can I infer the configured bandwidth statement on the physical interface would be used? and not the real line speed?
Not 100% sure, but I would hope that the speed interface configuration comes into play, if not all the calculations for the QOS will be wrong unless QOS is only used with line speed. BTW, is this a module or a 9300 switch ( N9K-X97160YC-EX)?
If it is a module, it doesn't appear to support 100Mb