[c-nsp] ASR 920 - H-QoS across 2+ EFP interfaces

Wed Jul 29 20:25:58 EDT 2015

Steve,
Please refer to the QOS section in the following Cisco Live deck. ASR920 has the same behavior.  It will give you an idea about the QOS architecture and how to construct your policy in Cisco ASR920.
http://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnxtd2FyaXN8Z3g6NzI1MTc2YzdjNGI2YmQ1NA

Best Regards,

[http://www.cisco.com/web/europe/images/email/signature/horizontal06.jpg]

Waris Sagheer
Technical Marketing Manager
Service Provider Routing Segment
waris at cisco.com<mailto:waris at cisco.com>
Phone: +1 408 853 6682
Mobile: +1 408 835 1389

CCIE - 19901

<http://www.cisco.com/>

This email may contain confidential and privileged material for the sole use of the intended recipient. Any review, use, distribution or disclosure by others is strictly prohibited. If you are not the intended recipient (or authorized to receive for the recipient), please contact the sender by reply email and delete all copies of this message.

For corporate legal information go to:http://www.cisco.com/web/about/doing_business/legal/cri/index.html

From: "cisco-nsp-bounces at puck.nether.net<mailto:cisco-nsp-bounces at puck.nether.net>" <cisco-nsp-bounces at puck.nether.net<mailto:cisco-nsp-bounces at puck.nether.net>> on behalf of Steve Margelos <smargelos at atlantech.net<mailto:smargelos at atlantech.net>>
Date: Monday, July 20, 2015 at 10:51 AM
To: "cisco-nsp at puck.nether.net<mailto:cisco-nsp at puck.nether.net>" <cisco-nsp at puck.nether.net<mailto:cisco-nsp at puck.nether.net>>
Subject: [c-nsp] ASR 920 - H-QoS across 2+ EFP interfaces

Hello,

I have an ASR 920 that I will be using to form an NNI with an upstream provider.  We typically have customers that purchase not just an internet circuit, but also P2P/VPLS services all over the same single physical circuit.  We then separate these services out into logical units on our NNI so that we can take appropriate action, depending on the inner  C-VLAN that we see coming from the remote CPE.  The problem I am having is designing/implementing a QoS solution that will take into account multiple logical circuits, while still allowing me to prioritize certain types of traffic.  Normally, we use Juniper to accomplish this, via interface-sets and applying a multi-level class-of-service configuration. As an example:

set interfaces interface-set Customer-XYZ interface ge-1/0/5 vlan-tags-outer 725

set class-of-service interfaces interface-set Customer-XYZ output-traffic-control-profile LEVEL_2-svlan_100m

set class-of-service interfaces ge-1/0/5 unit 123 output-traffic-control-profile LEVEL_3-cvlan_data-50m       **CKT #2**
set class-of-service interfaces ge-1/0/5 unit 725 output-traffic-control-profile LEVEL_3-cvlan_data-50m       **CKT #1**

set class-of-service traffic-control-profiles LEVEL_2-svlan_100m scheduler-map svlan-data-voice
set class-of-service traffic-control-profiles LEVEL_2-svlan_100m shaping-rate 100m
set class-of-service traffic-control-profiles LEVEL_2-svlan_100m guaranteed-rate 100m

set class-of-service traffic-control-profiles LEVEL_3-cvlan_data-50m scheduler-map cvlan-data
set class-of-service traffic-control-profiles LEVEL_3-cvlan_data-50m shaping-rate 100m
set class-of-service traffic-control-profiles LEVEL_3-cvlan_data-50m guaranteed-rate 50m

Trying to translate this type of configuration over to the Cisco ASR 920 has been.... Difficult.

I believe the only thing close to the Juniper interface-set is the Cisco 'service-group'.  The problem I am having is that the service-group    service-policy restrictions are  not allowing me to ensure certain traffic (voice) is given priority queue access.    Per the ASR documentation, I understand that queuing and marking is not supported at the service-group policy level, but I am hoping to get this to work by having a service-policy on the EFP itself and having the ASR "merge" the two together, as described in the documentation - of course, that is not working as designed either.

Here is what I have so far:

**setting up the service-group and calling a rather simple class-default policy-map**
service-group 1234
description CID-1234 EVC group
service-policy output 50Mbps_Egress_Service_Group
end

Router#show policy-map 50Mbps_Egress_Service_Group
  Policy Map 50Mbps_Egress_Service_Group
    Class class-default
     police cir 50000000 bc 1562500
       conform-action transmit
       exceed-action drop

interface GigabitEthernet0/0/0
no ip address
load-interval 30
negotiation auto
storm-control broadcast level 5.00 4.00
storm-control action shutdown
no lldp transmit
no lldp receive
spanning-tree portfast trunk
spanning-tree bpdufilter enable
spanning-tree bpduguard enable

service instance 30 ethernet
  description CID-XYZ-Internet
  encapsulation dot1q 30
  rewrite ingress tag pop 1 symmetric
  group 1234
  service-policy output EFP_Egress_Level3

Router#show policy-map EFP_Egress_Level3
  Policy EFP_Egress_Level3
    Class EGR-EF-CLASS
      priority

Router#show class-map EGR-EF-CLASS
Class Map match-all EGR-EF-CLASS (id 4)
  Description: Matches VoIP EF traffic
   Match ip  dscp ef (46)

Once I apply the service-policy in the EFP instance above, I get a console log message stating the following:

"Cannot attach queuing-based child policy to a non-queuing based class"

So my question is, can anyone point me in the right direction in getting as closed to the desired Juniper solution posted above?

Thank you.

Sincerely,
Steve Margelos
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Atlantech Online, Inc.  | http://www.atlantech.net
phone: 301-755-2260 | 866-755-2260
fax: 301-589-3936
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
view our Network status | http://noc.atlantech.net

_______________________________________________
cisco-nsp mailing list  cisco-nsp at puck.nether.net<mailto:cisco-nsp at puck.nether.net>
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/