MPLS issue ( Route Reflector )

[heshamelsayed]

Hi have been struggling with a lab , to create an MPLS network using route reflector 

from PE R5 i can see from the trace that MPLS label is being used when pinging the loopback of the other PE R6

the routes for each customer is learned without issues on bgp

but when pinging from PE R5 the subnet learned by bgp 172.16.21.254 which is via the pingable loopback interface it timeout  

i have attached all the MPLS network routers and R5 include some show runs as well at the bottom 

if any suggestions please , that would be great 

[MHM Cisco World]

mpls ldp autoconfig

Instead of it use 

Mpls ip 

[heshamelsayed]

tried that , same result

[Harold Ritter · Accepted solution]

Hi @heshamelsayed ,

You need to configure "redistribute connected" under the vrf stanza if you want to be able to ping from the PE. Otherwise the destination does not have a route back to the source of the ping.

R5:

router bgp 65000
address-family ipv4 vrf SMART
redistribute connected

Regards,

[heshamelsayed]

i also tried ping using the loopback interface which is known by every other router but no success 

[MHM Cisco World]

CE-PE 

As @Harold Ritter you need to redistrubte not LO but the link connect CE to PE' and this done by redistrubte connect under vrf of CE.

The PE to PE use LO as next hop but it need connect link in vrf RIB.

[heshamelsayed]

i am actually trying to ping from the PE , but i also tried @Harold idea of redistributing the connected subnets , but still cant find any ping working 

R5#traceroute 1.1.6.1
Type escape sequence to abort.
Tracing the route to 1.1.6.1
VRF info: (vrf in name/id, vrf out name/id)
  1 10.16.16.2 [MPLS: Label 16 Exp 0] 12 msec 16 msec 8 msec
  2 10.50.50.1 12 msec 16 msec 16 msec

here i am trying to ping the other PE loopback and it is working fine and using MPLS label 

      10.0.0.0/8 is variably subnetted, 2 subnets, 2 masks
C        10.15.15.0/24 is directly connected, Ethernet1/0
L        10.15.15.1/32 is directly connected, Ethernet1/0
      172.16.0.0/24 is subnetted, 2 subnets
B        172.16.18.0 [20/0] via 10.15.15.2, 00:19:10
B        172.16.19.0 [200/0] via 1.1.6.1, 00:19:02
B     192.168.1.0/24 [20/0] via 10.15.15.2, 00:19:10
B     192.168.2.0/24 [200/0] via 1.1.6.1, 00:19:02

yet though on R5 i received the routes from R6 and i can ping its loopback ( which is the next hope for the bgp routes ) i get nothing

[heshamelsayed]

traceroute 1.1.6.1 ( this trace work fine using mpls )

B        172.16.19.0 [200/0] via 1.1.6.1, 00:19:02 ( but pinging this which is via 1.1.6.1 ) doesn't work

[MHM Cisco World]

1.1.6.1

This IP is for what ? CE subnet?

If yes then it appear in vrf of ce not in global.

And you need to redistrubte ce-pe link under vrf address family of bgp' and use this IP as source of ping.

[heshamelsayed]

This IP is the loopback of the other PE router , i am pinging from PE to PE

[heshamelsayed]

Yep at first i did it on one of the PE only and it didnt work , using wireshark i found that it true that the CE doesnt know how to return the packets , Thank you very much ,