BGP Nex hop self added to Border Router, still cannot PING

[AlFortner]

Hello, I'm having issues with my BGP configuration where I have used the next hop self command on my border routers which now display itself as the next hop to its IBGP neighbors in the internal routing table via BGP. Also, the destination router outside the AS as learned via BGP routes to Loopback networks on my IBGP router but they cannot PING. My topology have 6 routers in OSPF 1 Area 0 with complete reachability, and 3 routers (R-1,R-5, and R-6) configured in IBGP advertising it's loopback (1.1.1.1, 5.5.5.5, and 6.6.6.6 - AS 100) using the meshed OSPF as it's underlay. Routes to destinations are reachable via OSPF but not BGP thought both BGP in each AS has valid routes.

I have 3 routers in separate AS advertising routes to IBGP following this video: https://www.youtube.com/watch?v=c7kTbjSZwgs&t=764s 

[MHM Cisco World]

In bgp you need these points to ping 

1- config LO in router source of ping 

2- advertise this LO via bgp

3- ping <ip> source interface LO <you config in first point>

MHM

[David Ruess]

You need to make sure you have reachability (your networks in the routing table). BGP neighbors are one thing but advertising routes is another. Make sure the IPs you are trying to ping are advertised into BGP and use that as a source/destination in your ping.

-David

[AlFortner]

Yes sir, I have confirmed each loopback from R1 (1.1.1.1) to R6 (6.6.6.6) all have reachability. Loobacks for R1(1.1.1.1) R5 (5.5.5.5) and R6(6.6.6.6) (IBGP AS-500) have reachability using OSPF as IGP. Routers R2(2.2.2.) R3(3.3.3.3) and R4 (4.4.4.4) are not participating in BGP however, have reachability via internal route. 

[Martin L]

How do u ping? what is source and destination ? Ping BGP often will fail because ur ping takes outgoing interface as source; so, check if destination router has a route back to ur source ping outgoing interface IP OR use extended ping command where u specify the source of ping (not only detitanation).  Check routing tables; check BGP tables;  What others have mentioned is so called "black-hole" topology.  Solution for that is to redistribute OSPF (underlying protocol) into BGP. 

 Also, redistribute connected routes into BGP.  

Regards, ML
**Please Rate All Helpful Responses **

[AlFortner]

Hello,

I've sourced the ping with my loopback of R1 to destination 192.168.10.1. On R1, both route tables have the route to the destination with "sh ip route" displaying routes learned via BGP with next of of R-5 and R-6 as they have been configured with "next-hop-self as iBGP peer to R1 (1.1.1.1).  R9 (AS1234) advertises the 192.168.X.X networks to R5 and R6 via BGP. Again, applying the next-hop-self to R5 and R6 allowed advertisement to R1 to learn the External (192.168.X.X) networks

[MHM Cisco World]

Routers Between R1 and R5/R6 know about these LO? 

Instead ping use traceroute (also use source) and see in which router the traffic stop

MHM

[Martin L]

I've sourced the ping with my loopback of R1 to destination 192.168.10.1. On R1,  Source on R1 was 1.1.1.1?  if so, does R9 has route to 1.1.1.1 of R1 loopback ?   192.168.10.1. belongs to R9 , right?  do ping 192.168.10.1 source 1.1.1.1, what is result ? 

turn on ICMP debug on source and destination routers; also do traceroute source destination and check results. (or debug icmp on all routers)

Another TS trick is to check routes with show ip bgp x.x.x.x and check if there are any BGP-RIP failures

Regards, ML
**Please Rate All Helpful Responses **

[paul driver]

Hello
You need to make sure you have full ospf reachability throughout AS 500 and all the rtrs that ASN  have full ibgp mesh connectivity