Does BGP advertise rib-failure routes to its neighbor?

[gongya001]

In my understanding, RIB-Failure route is not advertised to its bgp neighbor, right?

1.1.1.1 is configured on R1 lo0. r1,r2,and r3 are configured with OSPF and iBGP, on R2, I redistributed ospf into BGP.

On R3, I see these

R3#show  ip route 1.1.1.1
Routing entry for 1.1.1.1/32
  Known via "ospf 1", distance 110, metric 11, type intra area
  Last update from 10.10.2.1 on Ethernet0/1, 2d02h ago
  Routing Descriptor Blocks:
  * 10.10.2.1, from 1.1.1.1, 2d02h ago, via Ethernet0/1
      Route metric is 11, traffic share count is 1
R3#show ip bgp | i 1.1.1.1
 r>i 1.1.1.1/32       2.2.2.2                 11    100      0 ?
R3#show ip bgp neighbors 20.10.2.4 advertised-routes | i 1.1.1.1
 r>i 1.1.1.1/32       2.2.2.2                 11    100      0 ?

On R4, I see these

R4#show ip bgp regexp ^65100$
BGP table version is 60, local router ID is 144.144.144.1
Status codes: s suppressed, d damped, h history, * valid, > best, i - internal,
              r RIB-failure, S Stale, m multipath, b backup-path, f RT-Filter,
              x best-external, a additional-path, c RIB-compressed,
Origin codes: i - IGP, e - EGP, ? - incomplete
RPKI validation codes: V valid, I invalid, N Not found

     Network          Next Hop            Metric LocPrf Weight Path
 *>  1.1.1.1/32       20.10.2.3                              0 65100 ?
 *                    20.10.1.2               11             0 65100 ?

What I do not understand is why R4 selects R3 instead of R2.

Should R3 advertise 1.1.1.1?

thanks for your help in advance.

[Martin L]

Redistribution occurs from routing table not from routing database.  Rule is routes are not in RIB cannot be redistributed. right?  but there is exception for BGP and OSPF;  Both OSPF and BGP can advertise routes that are not in RIB but are in database.  RIP Failure is how BGP warns us about potential looping may happen.

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

[Martin L]

there might be also restriction to that exception on OSFP side. something about route must be in database and in area from which your router is doing redistribution. 

can you show us show ip bgp 1.1.1.1/32 from R4;

[gongya001]

R4#show ip bgp 1.1.1.1/32
BGP routing table entry for 1.1.1.1/32, version 95
Paths: (4 available, best #4, table default)
  Advertised to update-groups:
     11
  Refresh Epoch 5
  65100, (received & used)
    20.10.1.2 from 20.10.1.2 (2.2.2.2)
      Origin incomplete, metric 11, localpref 100, valid, external
      rx pathid: 0, tx pathid: 0
  Refresh Epoch 3
  200 65100
    30.10.1.5 from 30.10.1.5 (5.5.5.5)
      Origin incomplete, localpref 100, valid, external
      rx pathid: 0, tx pathid: 0
  Refresh Epoch 3
  300 65100
    30.10.2.6 from 30.10.2.6 (6.6.6.6)
      Origin incomplete, localpref 100, valid, external
      rx pathid: 0, tx pathid: 0
  Refresh Epoch 4
  65100, (received & used)
    20.10.2.3 from 20.10.2.3 (3.3.3.3)
      Origin incomplete, localpref 100, valid, external, best
      rx pathid: 0, tx pathid: 0x0

thanks a lot !!

Can you show me where this exception is documented ?

[Martin L · Accepted solution]

not sure where is this documented;  I am sure it is somewhere in Doc but Cisco is moving pages often so it might be harder to find suff.  IF you going for CCIE, you should be reading Cisco Doc pages, so,  let me know if u find it.

I learnt it from my studies from INE course for CCIE RS v5 by Brian McGahan. 

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

[gongya001]

thanks !!

[Giuseppe Larosa]

Hello @gongya001 ,

>> What I do not understand is why R4 selects R3 instead of R2.

From R4 point of view it receives the eBGP update from R3 with a missing metric. And a missing MED is preferred (considered 0) over a MED.

Network          Next Hop            Metric LocPrf Weight Path
 *>  1.1.1.1/32       20.10.2.3                              0 65100 ?
 *                    20.10.1.2          11             0 65100 ?

As you noted the interesting part is why R3 advertises the 1.1.1.1/32 to R4

In this case the RIB faliure on R3 simply means that the BGP path is not installed in the routing table because iBGP AD 200 is worse then OSPF AD 110. A part from this the BGP advertisement is valid. However, the BGP advertisement exists and it is sent to the eBGP peer. it can be sent because the prefix exists also in IGP (OSPF).

If the above ouput taken from R4 is correct R3's advertisement to R4 has a missing MED attribute that makes it preferred over R2's advertisement.

Hope to help

Giuseppe

[gongya001]

thanks a million to make this so clear.