Re: [c-nsp] Nex3K and MTU
Thanks very much Christophe - why does "sh int port-chan 10" show mtu of 9216? Shouldnt it show 1500bytes? # sh int port-channel 10 port-channel10 is up Hardware: Port-Channel, address: 30f7.0da0.7378 (bia 30f7.0da0.7378) Description: 20Gb_VPC_PEER_LINK_TO_TORSW02-EQ-SY3-1806B MTU 9216 bytes, BW 2000 Kbit, DLY 10 usec And Id love to know the logic behind the decision to use qos to "change" mtu, on this platformseems the larger Nexus's you can change it on a per Int basis? lolOr is logic a paradox when it comes to the differences between Cisco platforms/OS's/Feature disparities... From: Christophe Fillot <c...@utc.fr> Sent: Thursday, 3 March 2016 7:34 PM To: CiscoNSP List; cisco-nsp@puck.nether.net Subject: Re: [c-nsp] Nex3K and MTU On 03/03/2016 02:10 AM, CiscoNSP List wrote: Yet "other" ports show 1500bytes: # sh int eth 1/1 Ethernet1/1 is down (Link not connected) Dedicated Interface Hardware: 10/100/1000 Ethernet, address: 547f.ee7b.ff28 (bia 547f.ee7b.ff28) MTU 1500 bytes, BW 100 Kbit, DLY 10 usec reliability 255/255, txload 1/255, rxload 1/255 Encapsulation ARPA Port mode is access To see the real MTU you have to use "show queuing interface ". For example: # sh int e1/45 Ethernet1/45 is up Dedicated Interface Hardware: 1000/1 Ethernet, address: 8c60.4f96.2cf4 (bia 8c60.4f96.2cf4) Description: storage2-01 MTU 1500 bytes, BW 1000 Kbit,, BW 1000 Kbit, DLY 10 usec [...] # sh queuing int e1/45 n5k-pg2-a# sh queuing int e1/45 Ethernet1/45 queuing information: TX Queuing qos-group sched-type oper-bandwidth 0 WRR100 RX Queuing qos-group 0 q-size: 100160, q-size-40g: 100160, HW MTU: 9216 (9216 configured) drop-type: drop, xon: 0, xoff: 0 [...] Hope this helps Christophe ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
Re: [c-nsp] Nex3K and MTU
On 03/03/2016 02:10 AM, CiscoNSP List wrote: > Yet "other" ports show 1500bytes: > > # sh int eth 1/1 > Ethernet1/1 is down (Link not connected) > Dedicated Interface > Hardware: 10/100/1000 Ethernet, address: 547f.ee7b.ff28 (bia 547f.ee7b.ff28) > MTU 1500 bytes, BW 100 Kbit, DLY 10 usec > reliability 255/255, txload 1/255, rxload 1/255 > Encapsulation ARPA > Port mode is access > To see the real MTU you have to use "show queuing interface ". For example: # sh int e1/45 Ethernet1/45 is up Dedicated Interface Hardware: 1000/1 Ethernet, address: 8c60.4f96.2cf4 (bia 8c60.4f96.2cf4) Description: storage2-01 *MTU 1500 bytes*, BW 1000 Kbit,, BW 1000 Kbit, DLY 10 usec [...] # sh queuing int e1/45 n5k-pg2-a# sh queuing int e1/45 Ethernet1/45 queuing information: TX Queuing qos-group sched-type oper-bandwidth 0 WRR100 RX Queuing qos-group 0 q-size: 100160, q-size-40g: 100160, *HW MTU: 9216 (9216 configured)* drop-type: drop, xon: 0, xoff: 0 [...] Hope this helps Christophe ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
[c-nsp] Nex3K and MTU
Ok - Well another "surprise" for me on these boxes, you cant set MTU the "normal" way (Well normal for me, coming from Cat/routers)...i.e. you cant set it globally (system mtu), nor per Int...you have to do it via QOS(Interesting decision by NXOS developers lol) Anyway - What Im a little confused about is thisthe 3Ks automatically place a rather extensive "default" QOS policy (Copp/control plane), which doesnt have any reference to an MTU of 9216. QOS to "add" MTU of 9216 is this (From what Ive researched): policy-map type network-qos jumbo class type network-qos class-default mtu 9216 system qos service-policy type network-qos jumbo Without this added though, VPC link (20Gb portchan), it is stating MTU of 9216: # sh int port-channel 10 port-channel10 is up Hardware: Port-Channel, address: 30f7.0da0.7378 (bia 30f7.0da0.7378) Description: 20Gb_VPC_PEER_LINK_TO_TORSW02-EQ-SY3-1806B MTU 9216 bytes, BW 2000 Kbit, DLY 10 usec reliability 255/255, txload 1/255, rxload 1/255 Encapsulation ARPA Port mode is trunk Yet "other" ports show 1500bytes: # sh int eth 1/1 Ethernet1/1 is down (Link not connected) Dedicated Interface Hardware: 10/100/1000 Ethernet, address: 547f.ee7b.ff28 (bia 547f.ee7b.ff28) MTU 1500 bytes, BW 100 Kbit, DLY 10 usec reliability 255/255, txload 1/255, rxload 1/255 Encapsulation ARPA Port mode is access Where is the switch getting this 9216mtu for the VPC/Portchan link from? (It def. isnt set on the member Ints(You are unable to set MTU on Ints on these boxes), and you cant set it globally)? Cheers ___ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/