The four ports on the Cisco Nexus 7000 Series switch that connect to the uplink ports are fabric ports. Only QoS policies can be configured on the server-facing FEX ports. Currently, queuing on the FEX interfaces is not supported.
http://www.cisco.com/en/US/docs/switches/datacenter/sw/6_x/nx-os/qos/configuration/guide/overview.html Since it's not possible to apply QoS config on the FEX fabric ports, by applying DSCP marking on the FEX host ports, packets ingressing on those ports will retain the DSCP value while going to the parent N7k. In any case, you might want to confirm with a SPAN capture. Best regards, Andras On Mon, Jan 14, 2013 at 4:26 PM, Pavel Vraštiak <[email protected]>wrote: > Hi list, > suppose you have two n7ks (F2 modules) connected to each other (vpc) and > several FEXes directly connected to them. Now, is it neccesary/possible to > configure any qos-related stuff on the link between fex and n7k? I was not > successful with that. I read somewhere that there are pause frames which > push congestion to other links. > > Would setting DSCP values on ingress n2ks and prioritizing traffic on > egress from n7k based on these DSCP values be sufficient and working as > expected? > > I mean, I found tons of stuff on that for n5k/n2ks, but n7k is way too > different. > > Paul > ______________________________**_________________ > cisco-nsp mailing list [email protected] > https://puck.nether.net/**mailman/listinfo/cisco-nsp<https://puck.nether.net/mailman/listinfo/cisco-nsp> > archive at > http://puck.nether.net/**pipermail/cisco-nsp/<http://puck.nether.net/pipermail/cisco-nsp/> > _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
