[c-nsp] traffic shape on 87x/88x/18xx SVI interfaces

[Shimol Shah]

GTS is the old way of doing Qos. It is not supported in the CEF path. 
MQC is the new and recommended way. With GTS if you do "sh cef int <>" 
you will see something like below

DUT#sh run int g3/2
Building configuration...

Current configuration : 168 bytes
!
interface GigabitEthernet3/2
  ip address 192.168.3.1 255.255.255.0
  tag-switching ip
  traffic-shape rate 20000000 500000 500000 1000
  max-reserved-bandwidth 100
end


DUT#sh cef int g3/2 int
GigabitEthernet3/2 is up (if_number 7)
   Corresponding hwidb fast_if_number 7
   Corresponding hwidb firstsw->if_number 7
   Internet address is 192.168.3.1/24
   ICMP redirects are always sent
   IP unicast RPF check is disabled
   Inbound access list is not set
   Outbound access list is not set
   IP policy routing is disabled
   BGP based policy accounting is disabled
   Packets switched to this interface are dropped to the next slow path: 
Traffic shaping


Notice the last line ^^^^^^^^^^^^

Without it you will see,

DUT#sh cef int g3/2 int
GigabitEthernet3/2 is up (if_number 7)
   Corresponding hwidb fast_if_number 7
   Corresponding hwidb firstsw->if_number 7
   Internet address is 192.168.3.1/24
   ICMP redirects are always sent
   IP unicast RPF check is disabled
   Inbound access list is not set
   Outbound access list is not set
   IP policy routing is disabled
   BGP based policy accounting is disabled
   Hardware idb is GigabitEthernet3/2 (7)
   Software idb is GigabitEthernet3/2 (7)
   Fast switching type 22, interface type 141
   IP CEF switching enabled
   IP Fast switching turbo vector
   IP CEF switching with tag imposition turbo vector
   Input fast flags 0x0, Output fast flags 0x0
   ifindex 4(4)
   Slot 3 Slot unit 2 VC -1
   Transmit limit accumulator 0x0 (0x0)
   IP MTU 1500
  Subblocks:
   MPLS subblock present

Hope that helps

Shimol



On 4/27/10 10:06 AM, Chris Flav wrote:
>
> Hello,
>
> We are trying to do a simple "traffic-shape rate" command on a variety of router platforms (871,881,1811) and have determined that the traffic-shape does not actually take effect unless "no ip route-cache cef" is applied to the Fe4 interface (or Fe0 or Fe1 on the 1811).  Traffic shape commands applied to the Fe4 interface however work properly, and on the 1811 Fe0 and Fe1 interfaces behave correctly.
>
> Is there any other way to have shaping applied on VLANs configured on the SVI interfaces without sacrificing CEF?
>
> bug toolkit didn't show any bugs relating to shaping on the platforms I checked.
>
> Thanks,
>
> C.
>
>
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/