Narbik / Dan TS Lab 8 - question 12 (MPLS)

    - S0/0 - (R1) - F0/0 -
  | |
S0/0 F0/0 EIGRP

(R2) (R6) -----------------

F0/0 F2/0 OSPF
  | |
      - F0/0 - (R3) - F0/0 -

Hello,

This one is stumping me. R2, R3, and R6 make up the MPLS network with an
MPLS VPN between R2 and R6 with their local loopbacks being members of vrf
V26.

 =======================================
   ((R2))
   interface Loopback26
    ip vrf forwarding V26
    ip address 26.26.0.2 255.255.255.255

   ((R6))
   interface Loopback26
    ip vrf forwarding V26
    ip address 26.26.0.6 255.255.255.255
  =======================================

Both R2 and R6 are receiving each other's routes for that VRF but neither
address is pingable.

  =======================================
   R2#sh ip route vrf V26
   C 26.26.0.2 is directly connected, Loopback26
   B 26.26.0.6 [200/0] via 99.0.6.6, 01:08:05

 R6#sh ip route vrf V26
   B 26.26.0.2 [200/0] via 99.0.2.2, 01:06:34
   C 26.26.0.6 is directly connected, Loopback26
  =======================================

The cause of the problem is that R6 is advertising its LDP router ID with a
/24 mask instead of a /32. R3 is receiving the route via OSPF from R6 as a
/32 though. Why does R6 send the LDP advertisement as a /24?

 =======================================
R3#*sh ip route | i 99.0.6
*O 99.0.6.6/32 [110/2] via 99.2.36.6, 00:00:53, FastEthernet2/0.36
O E2 99.0.6.0/24 [110/20] via 99.2.23.2, 00:00:53, FastEthernet0/0
R3#*sh mpls forwarding-table 99.0.6.6
*Local Outgoing Prefix Bytes tag Outgoing Next Hop
tag tag or VC or Tunnel Id switched interface
17 *Untagged* 99.0.6.6/32 7095 Fa2/0.36 99.2.36.6
R3#sh mpls ldp bindings neighbor 99.0.6.6 *99.0.6.0 24
* tib entry: 99.0.6.0/24, rev 48
        remote binding: tsr: 99.0.6.6:0, tag: imp-null

R3#sh mpls ldp bindings neighbor 99.0.6.6 99.0.6.6 32
<nothing>
 =======================================

And the stated fix is to enter the "ip ospf net point-to-point" command
under the R6's Lo0 so that OSPF also advertises a /24.
After making that change i'm sitll unable to ping between loopbacks on R2
and R6

 =======================================
R3#*sh mpls forwarding-table 99.0.6.6
*Local Outgoing Prefix Bytes tag Outgoing Next Hop
tag tag or VC or Tunnel Id switched interface
34 *Pop tag* 99.0.6.0/24 0 Fa2/0.36 99.2.36.6

R6#*ping vrf V26 26.26.0.2*
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 26.26.0.2, timeout is 2 seconds:
.....
Success rate is 0 percent (0/5)
 =======================================

1st question - why did R6 advertise its loopback as a /24 via LDP?
2nd question - why didn't adding the "ip ospf net point-to-point" fix the
problem?

Thanks,
Matt

Blogs and organic groups at http://www.ccie.net
Received on Fri Dec 31 2010 - 14:01:30 ART