RE: EBGP peer is not advertising prefixes from other EBGP peer

From: Popgeorgiev Nikolay (nikolay.popgeorgiev@siemens.com)
Date: Wed Mar 08 2006 - 06:12:27 GMT-3


Dear Mushtaq,

I didn't go really through all your post but I can tell you how I did this thing and what problems I faced in a similar task:

I used RR in AS 100, R1 as a route reflector and R2 as its client.
Because the requirement says do't disable synchronization, in order for R1 for example to put network 20.20.20.20 into its routing table, it has to have this route known already via IGP or static or connected.
So my solution is to redistribute BGP into IGP on router R2. This is not a very clever thing to do if you have many routes coming from BGP but in this case helps :)

Try it and if I have missed something in your explanations sorry about that

Best,
Nick
                                               

 

-----Original Message-----
From: Mushtaq A. Khan [mailto:mak.ccie2b@gmail.com]
Sent: Wednesday, March 08, 2006 7:59 AM
To: ccielab@groupstudy.com
Subject: EBGP peer is not advertising prefixes from other EBGP peer while synchronization is enabled in Transit ASs

Hi Team,

I have a scenario as shown below.

     SW2(AS200) ----(RIP)-----R2(AS100)
                                                       \
                                                      (OSPF)
                                                        \
                                                         R1(AS100)
-----(OSPF)----R6(AS600)
                                                         /
                                                       (OSPF)
                                                        /
      R5(AS600)-----EIGRP-----R3(AS100)

The task is to advertise the following prefixes in BGP ONLY within their
respective ASs and these should be exchanged between all AS's.

     20.20.20.20(SW2)
      5.5.5.5 (R5)
      6.6.6.6 (R6)

AS100 should provide transit path between the above prefixes and MUST not
disable synchronization and use minimal number of BGP speakers.

One solution to above requirement is to use BGP confederation within AS100.
I tried to solve this using route reflector and facing difficulty: I noticed
R6 is able to see all 3 prefixes in its BGP table. However, R5 and SW2 can't
see each other prefixes. I tried different options but with no luck. Anyone
has any clue to resolve this issue here.

All relevent configuration are given below:

R2#sh run | beg router bgp
router bgp 100
 synchronization
 bgp log-neighbor-changes
 neighbor 172.16.20.10 remote-as 200
 neighbor 172.16.101.1 remote-as 100
 neighbor 172.16.101.1 update-source Loopback102
 no auto-summary
!
R2#sh run | beg router ospf|route-map
router ospf 1
 log-adjacency-changes
 redistribute rip subnets
 redistribute bgp 100 subnets route-map BGPtoOSPF
 network 172.16.102.1 0.0.0.0 area 20
 network 172.16.123.2 0.0.0.0 area 0
!
ip prefix-list BGP-Route seq 5 permit 20.20.20.0/24
!
route-map BGPtoOSPF permit 10
 match ip address prefix-list BGP-Route

R2#sh ip bgp | beg Network
   Network Next Hop Metric LocPrf Weight Path
* i5.5.5.0/24 172.16.35.5 0 100 0 500 i
r>i6.6.6.0/24 172.16.16.6 0 100 0 600 i
*> 20.20.20.0/24 172.16.20.10 0 0 200 i

sw2#sh ip bgp | beg Network
   Network Next Hop Metric LocPrf Weight Path
*> 6.6.6.0/24 172.16.20.2 0 100 600 i
*> 20.20.20.0/24 0.0.0.0 0 32768 i

R1#sh run | beg router ospf|route-map
router ospf 1
 router-id 172.16.101.1
  redistribute bgp 100 subnets route-map BGPtoOSPF
 !
router bgp 100
 synchronization
 bgp log-neighbor-changes
 neighbor AS100 peer-group
 neighbor AS100 remote-as 100
 neighbor AS100 update-source Loopback101
 neighbor AS100 route-reflector-client
 neighbor 172.16.16.6 remote-as 600
 neighbor 172.16.102.1 peer-group AS100
 neighbor 172.16.103.1 peer-group AS100
 no auto-summary
!
!
ip prefix-list BGP-Routes seq 5 permit 6.6.6.0/24
!
route-map BGPtoOSPF permit 10
 match ip address prefix-list BGP-Routes
!

R1#sh ip bgp | beg Network
   Network Next Hop Metric LocPrf Weight Path
r>i5.5.5.0/24 172.16.35.5 0 100 0 500 i
*> 6.6.6.0/24 172.16.16.6 0 0 600 i
r>i20.20.20.0/24 172.16.20.10 0 100 0 200 i

R6#sh ip bgp | beg Network
   Network Next Hop Metric LocPrf Weight Path
*> 5.5.5.0/24 172.16.16.1 0 100 500 i
*> 6.6.6.0/24 0.0.0.0 0 32768 i
*> 20.20.20.0/24 172.16.16.1 0 100 200 i

R3#sh run | beg router ospf|route-map
router ospf 1
 log-adjacency-changes
 redistribute bgp 100 subnets route-map BGPtoOSPF

router bgp 100
 synchronization
 bgp log-neighbor-changes
 neighbor 172.16.35.5 remote-as 500
 neighbor 172.16.101.1 remote-as 100
 neighbor 172.16.101.1 update-source Loopback103
 no auto-summary
!

!
ip prefix-list BGP-Routes seq 5 permit 5.5.5.0/24
!
route-map BGPtoOSPF permit 10
 match ip address prefix-list BGP-Routes

R3#sh ip bgp | beg Network
   Network Next Hop Metric LocPrf Weight Path
*> 5.5.5.0/24 172.16.35.5 0 0 500 i
r>i6.6.6.0/24 172.16.16.6 0 100 0 600 i
* i20.20.20.0/24 172.16.20.10 0 100 0 200 i
R3#
R5#sh ip bgp | beg Network
   Network Next Hop Metric LocPrf Weight Path
*> 5.5.5.0/24 0.0.0.0 0 32768 i
*> 6.6.6.0/24 172.16.35.3 0 100 600 i

TIA,
Mushtaq



This archive was generated by hypermail 2.1.4 : Sat Apr 01 2006 - 10:07:38 GMT-3