[c-nsp] Re ASR9000 Multicast over P2MP TE tunnels

Catalin Petrescu cpmarvin at gmail.com
Wed Jul 2 03:18:43 EDT 2014


Hi Cydon ,

We had a similar problem with XR . The setup was ng-mvpn , sender site xr
receiving site juniper.

What we found is the traffic is switched to pmp-te but the counters are not
reflecting this , so you might run into the same issue.

check:

-sh mrib mpls forwarding
- sh mrib route 232.11.11.11 detail  , look for enc id
- sh mfib  hardware encap id X  location 0/0/CPU
- sh mpls forwarding p2mp < -- counter increasing , wireshark on confirms
traffic is mapped to lsp

Relevant config in our setup :

multicast-routing
 address-family ipv4
  mdt source Loopback0
  rate-per-route
  interface all enable
  accounting per-prefix
 !
 vrf CST1
  address-family ipv4
   mdt source Loopback0
   mdt static p2mp-te tunnel-mte534
   rate-per-route
   interface all enable
   bgp auto-discovery p2mp-te
   !
   accounting per-prefix
  !
 !


router igmp
 vrf CST1
  interface tunnel-mte534
   static-group 232.100.100.15 10.3.183.3 < route to src over
 GigabitEthernet0/0/0/2.158
  !
 !
!

interface tunnel-mte534
 ipv4 unnumbered Loopback0
 destination 10.20.3.3
  path-option 2 dynamic
 !
 destination 10.20.4.4
  path-option 1 dynamic
 !
!

sh pim vrf CST1 topology

(10.3.183.3,232.100.100.15)SPT SSM Up: 09:02:31
JP: Join(00:00:06) RPF: GigabitEthernet0/0/0/2.158,10.3.58.8 Flags: MT  <
join sent to source
  tunnel-mte534               09:02:31  fwd LI LH

Traffic mapping to lsp is working.

Another test we did is to have the xr as a receiver , the p2mp tunnel is
building , but as you said no LSP-VIF interface , thus the rpf check will
fail. I never got it to work as advertised and ended up using mldp.

Hope this helps.

Catalin Petrescu

From: Cydon Satyr <cydonsatyr at gmail.com>
To: "cisco-nsp at puck.nether.net" <cisco-nsp at puck.nether.net>
Subject: [c-nsp] ASR9000 Multicast over P2MP TE tunnels
Message-ID:
        <CAF0PUwfayEE-KQUPPDmp6+N7UFs8ySDR+VOFPUUVnh5uA29NHQ at mail.gmail.com>
Content-Type: text/plain; charset=UTF-8

Hello experts.

I'm trying to map multicast traffic to RSVP-TE tunnel, without success. I'm
not quite sure what I might be missing (I think it's all there).

Topology is simple:


155.1.1.0/24------Gi0/0/0/0.110----XR1--------MPLS----------XR2-------Gi0/0/0/0.150


Here's the config (the relevant stuff - there's whole RSVP-TE topology
already working under it):

HEADEND (asr9k)
---------------------------

interface tunnel-mte1000
 ipv4 unnumbered Loopback0
 load-interval 30
 destination 2.2.2.2
  path-option 10 dynamic
 !
 destination 3.3.3.3
  path-option 10 dynamic
 !
 path-selection metric igp
 affinity ignore
!

multicast-routing
 address-family ipv4
  interface tunnel-mte1000
   enable

router igmp
 interface tunnel-mte1000
  static-group 232.11.11.11 155.1.1.104
 !

TAILEND (asr9k)
-------------------------

multicast-routing
 address-family ipv4
  core-tree-protocol rsvp-te group-list acl-232
 static-rpf 155.1.1.104 32 mpls 1.1.1.1

router igmp
 interface GigabitEthernet0/0/0/0.150
  static-group 232.11.11.11 155.1.1.104


Tunnel is up, I'm trying to ping from 155.1.1.104 which is connected
directly to HEADEND router.

show mrib route 232.11.11.11 | begin \\(
Tue Jul  1 17:29:01.984 UTC

*(155.1.1.104,232.11.11.11) RPF nbr: 155.1.1.104 Flags: EID*
*  Up: 05:35:59*
*  Incoming Interface List*
*    GigabitEthernet0/0/0/0.110 Flags: A, Up: 01:19:08*
*  Outgoing Interface List*
*    tunnel-mte1000 Flags: F NS LI LVIF, Up: 01:18:30*

However, no traffic seems to pass trough tunnel.

*show mfib interface tunnel-mte 1000*
*Tue Jul  1 17:30:41.257 UTC*

*Interface : tunnel-mte1000 (Enabled)*
*SW Mcast pkts in : 0, SW Mcast pkts out : 0*
*TTL Threshold : 0*
*Ref Count : 5*


Any chance you could give me a pointer to where I might be wrong? As a side
note, I don't see LSP-VIF interface coming up on tailend router.

Best Regards!


More information about the cisco-nsp mailing list