BGP Basic Labs

BGP Basic Labs
Hopefully you will not get bored
Emanuele Mazza
CCIE 11957
[email protected]
www.linkedin.com/in/emanuelemazza
Presentation_ID
2006 Cisco Systems, Inc. All rights reserved.
Cisco Confidential
Objective
These lab examples are not meant to demonstrate
proper or realistic BGP configurations, nor does it
include all the necessary commands to ensure proper
routing operation.
The objective of these examples are to show some of
the basic BGP configuration commands and their affect
on a network, including wrong configurations or
misconfigurations.
Hopefully this series of configurations and outputs will
help explain the how BGP is implemented and its
operation.
Our Scenario
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Lo0
AS200
RIP
version 2
192.168.1.32/27
Why IBGP?
.33
.194 SanJose3 .198
S0
S1
192.168.1.192/30
192.168.1.196/30
With very few

exceptions, interior BGP
(IBGP) BGP between
peers in the same AS
is used only in
multihomed scenarios.
Doyle
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
In order for the boundary routers SanJose1 and SanJose2 to exchange BGP routes that
they learn from ISP1 and ISP2 respectively, they must do one of the following:
Redistribution Not recommended!
Redistribute from BGP into the IGP (RIPv2) so those routes get propagated to the
other boundary router.
Redistribute from IGP (RIPv2) to BGP to learn routes from the other boundary
router or the use of BGP network statements.
Disadvantages: Can be a lot of routes being injected into the IGP; may also
require the use of route filters so not all routes are injected into the IGP or back4 out
to the other ISP.
Lo0
AS200
RIP
version 2
192.168.1.32/27
Why IBGP?
.33
.194 SanJose3 .198
S0
S1
192.168.1.192/30
192.168.1.196/30
With very few exceptions,

interior BGP (IBGP) BGP
between peers in the same AS
is used only in multihomed
scenarios. Doyle
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
In order for the boundary routers SanJose1 and SanJose2 to exchange BGP
routes that they learn from ISP1 and ISP2 respectively, they must do one of the
following:
IBGP
This allows the boundary routers to share EBGP routes without the
redistributing via an IGP.
Requires:
Synchronization: Cant advertise via EBGP if the boundary router
doesnt have it in its IGP routing table because the other IGP routers
will drop it. (Can turn this off if fully meshed IBGP or non-transit AS!)
5
Must know how to reach the next hop address of the route. (example
coming!)
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Setup
RIPv2 running within AS200
10.1.0.0/24 is between SanJose1 and ISP1s 10.2.0.0/24 networks,
both of which are part of 10.0.0.0/8 classful network.
RIPv2 was chosen because this will create a different situation than
with the two different Class B networks between SanJose2 and ISP2.
7
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Scenario A: Step #1 EBGP and IGP (RIPv2)

EBGP is configured between ISP routers and AS200 boundary routers.
RIPv2 is running within AS200
RIP network statements only include those networks in AS200
No IBGP has been configured.
There are no static routes.
8
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
Successful
pings
ISP1
.1
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Results:
AS 200 Boundary (SanJose1 and SanJose2) routers can ping peer ISP
networks
AS 200 Boundary (SanJose1 and SanJose2) routers can NOT ping
opposite ISP networks (SanJose1 can not ping 172.16.1.1)
ISP routers (ISP1 and ISP2) can ping remote peers attached networks
only, NOT SanJose3 network (we will see why in a moment).
9
ISP1
router bgp 100
network 10.2.0.0 mask 255.255.255.0
neighbor 10.1.0.2 remote-as 200
ISP2
router bgp 300
network 172.16.0.0
SanJose3
router rip
version 2
network 192.168.1.0
SanJose1
router rip
version 2
network 192.168.1.0
!
router bgp 200
network 192.168.1.0
SanJose2
router rip
version 2
network 192.168.1.0
!
router bgp 200
network 192.168.1.0
10
ISP1#show ip route
C
C
B
10.0.0.0/24 is
10.2.0.0 is
10.1.0.0 is
192.168.1.0/24
subnetted, 2 subnets
directly connected, Loopback0
directly connected, Serial0
[20/0] via 10.1.0.2, 00:25:16
ISPs only see the BGP

routes distributed by
themselves and their
peers in their BGP
tables.
ISP1#show ip bgp
Network
Next Hop
Metric LocPrf Weight Path
*> 10.2.0.0/24
0.0.0.0
0
32768 i
*> 192.168.1.0
10.1.0.2
0
0 200 i
-----------------------------------------------------------------------ISP2#show ip route
C
C
B
172.16.0.0/24 is subnetted, 1 subnets

172.16.1.0 is directly connected, Loopback0
172.30.1.0 is directly connected, Serial0/0
192.168.1.0/24 [20/0] via 172.30.1.2, 00:30:51
ISP2#show ip bgp
Network
*> 172.16.0.0
*> 192.168.1.0
Next Hop
0.0.0.0
172.30.1.2

0
32768 i
0
0 200 i
11
SanJose1#show ip route
B
C
R
C
R
C
R
SanJose1 only sees the BGP

routes distributed by itself and
its peer in their BGP tables.

10.2.0.0 [20/0] via 10.1.0.1, 00:32:26
192.168.1.0/24 is variably subnetted, 5 subnets, 2 masks
192.168.1.96/27 [120/2] via 192.168.1.194, 00:00:11, Serial0/0
192.168.1.64/27 is directly connected, Loopback0
192.168.1.32/27 [120/1] via 192.168.1.194, 00:00:11, Serial0/0
192.168.1.192/30 is directly connected, Serial0/0
192.168.1.196/30 [120/1] via 192.168.1.194, 00:00:11, Serial0/0
SanJose1#show ip bgp
Network
*> 10.2.0.0/24
*> 192.168.1.0
Next Hop
10.1.0.1
0.0.0.0

0
0 100 i
0
32768 i
Status codes - Status of the table entry. The status is displayed at the beginning of each line in the table. It can be one
of the following values:
s The table entry is suppressed.
* The table entry is valid.
> The table entry is the best entry to use for that network.
i The table entry was learned via an internal BGP (iBGP) session
12
B
C
C
R
R
R
C
SanJose2 only sees the

BGP routes distributed by
itself and its peer in their
BGP tables.
172.16.0.0/16 [20/0] via 172.30.1.1, 00:27:08

192.168.1.64/27 [120/2] via 192.168.1.198, 00:00:09, Serial0/0
192.168.1.32/27 [120/1] via 192.168.1.198, 00:00:09, Serial0/0
192.168.1.192/30 [120/1] via 192.168.1.198, 00:00:09, Serial0/0
Network
*> 172.16.0.0
*> 192.168.1.0
Next Hop
172.30.1.1
0.0.0.0

0
0 300 i
0
32768 i
13
R
R
C
C
C

192.168.1.96/27 [120/1] via 192.168.1.197, 00:00:18, Serial0/1
192.168.1.64/27 [120/1] via 192.168.1.193, 00:00:07, Serial0/0
SanJose3 does not have a route to the ISP networks or the networks
between the ISPs and the Boundary routers, so the ICMP Echo
Replies fail.
SanJose3 does not get any routes from the boundary routers except
IGP routes.
There are no BGP routes redistributed into the IGP and SanJose3 is
not IBGP peering with any BGP speakers.
14
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
IBGP
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Step #2 EBGP, IBGP and IGP (RIPv2)

IBGP is configured between SanJose1 and SanJose3.
15
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
IBGP
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Results:
AS 200 Boundary routers can ping peer ISP networks
AS 200 Boundary routers can NOT ping opposite ISP networks
ISP routers can ping remote peers attached networks only, NOT SanJose3
network
Synchronization is on, The BGP synchronization rule states that if an AS
provides transit service to another AS, BGP should not advertise a route until all
of the routers within the AS have learned about the route via an IGP. The
16
boundary routers do not forward the opposite ISP routes because the other
routers in the AS do not know about it.
ISP1
router bgp 100
network 10.2.0.0 mask 255.255.255.0
ISP2
router bgp 300
network 172.16.0.0
SanJose3
router rip
version 2
network 192.168.1.0
SanJose1
router rip
version 2
network 192.168.1.0
!
router bgp 200
network 192.168.1.0
SanJose2
router rip
version 2
network 192.168.1.0
!
router bgp 200
network 192.168.1.0
17
SanJose1 still only sees the BGP

routes distributed by itself and
its peer in their BGP tables.

B
10.2.0.0 [20/0] via 10.1.0.1, 00:09:13
C
R
192.168.1.96/27 [120/2] via 192.168.1.194, 00:00:27, Serial0/0
C
R
192.168.1.32/27 [120/1] via 192.168.1.194, 00:00:27, Serial0/0
C
R
192.168.1.196/30 [120/1] via 192.168.1.194, 00:00:27, Serial0/0
Network
*> 10.2.0.0/24
* i172.16.0.0
*> 192.168.1.0
* i
Next Hop
10.1.0.1
172.30.1.1
0.0.0.0
192.168.1.197

0
0 100 i
0
100
0 300 i
0
32768 i
0
100
0 i
Even though 172.16.0.0 network is in BGP table it is not in routing table

because there is no route to the next hop of 172.30.1.1
Also, the BGP Synchronization rule keeps the BGP route from being in the
18
routing table, since there is no IGP route.
B
C
C
R
R
R
C
172.16.0.0/16 [20/0] via 172.30.1.1, 00:11:27

192.168.1.64/27 [120/2] via 192.168.1.198, 00:00:19, Serial0/0
192.168.1.32/27 [120/1] via 192.168.1.198, 00:00:19, Serial0/0
192.168.1.192/30 [120/1] via 192.168.1.198, 00:00:19, Serial0/0
Network
Next Hop
* i10.2.0.0/24
10.1.0.1
*> 172.16.0.0
172.30.1.1
* i192.168.1.0
192.168.1.193
*>
0.0.0.0
SanJose2 still only sees

the BGP routes distributed
by itself and its peer in
their BGP tables.

0
100
0 100 i
0
0 300 i
0
100
0 i
0
32768 i
Even though 10.2.0.0 network is in BGP table it is not in routing table

because there is no route to the next hop of 10.1.0.1
Also, the BGP Synchronization rule keeps the BGP route from being in the
routing table, since there is no IGP route.
19
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
IBGP
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Step #3 Adding SanJose3 as a IBGP peer

IBGP is configured between SanJose1, SanJose2 and SanJose3, fullmesh.
No synchronization has been configured on all IBGP routers, because
we are fully meshed IBGP.
20
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
Synchronization
IBGP
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
When an IBGP router receives an update about a destination from an IBGP peer, it tries
to verify reachability to that destination via an IGP, such as RIP or OSPF.
If the IBGP router cant find the destination network in its IGP routing table, it will not
advertise the destination to other BGP peers.
If the route is not reachable through the IGP running within the AS, non-BGP routers
will not be able to route traffic passing through the AS towards this destination.
It is pointless to advertise destinations to external peers if traffic sent through this AS is
going to be dropped by some non-BGP router within the AS anyway.
Synchronization can be safely turned off on border routers:
When all transit routers inside the AS are running fully meshed IBGP.
When the AS is not a transit AS.
21
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
IBGP
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Results:
By itself, this will not change the reachability issues, because we have
not resolved the problem with the boundary routers being able to reach
the next-hop interface (SanJose1 still can not reach 172.30.1.1).
We will make another change in a moment.
Later, we might see that creating a full-mesh IBGP domain might not be
the best configuration choice.
22
ISP1
router bgp 100
network 10.2.0.0 mask 255.255.255.0
ISP2
router bgp 300
network 172.16.0.0
SanJose3
router rip
version 2
network 192.168.1.0
!
router bgp 200
no synchronization
SanJose1
router rip
version 2
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
!
SanJose2
router rip
version 2
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
23
Same Problem
SanJose1 still only sees the

BGP routes distributed by itself
and its peer in their BGP tables.

B
10.2.0.0 [20/0] via 10.1.0.1, 00:09:13
C
R
192.168.1.96/27 [120/2] via 192.168.1.194, 00:00:27, Serial0/0
C
R
192.168.1.32/27 [120/1] via 192.168.1.194, 00:00:27, Serial0/0
C
R
192.168.1.196/30 [120/1] via 192.168.1.194, 00:00:27, Serial0/0
Network
*> 10.2.0.0/24
* i172.16.0.0
*> 192.168.1.0
* i
Next Hop
10.1.0.1
172.30.1.1
0.0.0.0
192.168.1.197

0
0 100 i
0
100
0 300 i
0
32768 i
0
100
0 i
Even though 172.16.0.0 network is in BGP table it is not in routing

table because there is no route to the next hop of 172.30.1.1
24
B
C
C
R
R
R
C
Same Problem
SanJose2 still only sees

the BGP routes distributed
by itself and its peer in
their BGP tables.
172.16.0.0/16 [20/0] via 172.30.1.1, 00:11:27

192.168.1.64/27 [120/2] via 192.168.1.198, 00:00:19, Serial0/0
192.168.1.32/27 [120/1] via 192.168.1.198, 00:00:19, Serial0/0
192.168.1.192/30 [120/1] via 192.168.1.198, 00:00:19, Serial0/0
Network
Next Hop
* i10.2.0.0/24
10.1.0.1
*> 172.16.0.0
172.30.1.1
* i192.168.1.0
192.168.1.193
*>
0.0.0.0

0
100
0 100 i
0
0 300 i
0
100
0 i
0
32768 i
Even though 10.2.0.0 network is in BGP table it is not in routing

table because there is no route to the next hop of 10.1.0.1
25
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
Complete
reachability
IBGP
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Step #4 Adding the networks between the boundary routers and the ISP
routers to the RIP routing domain.
RIPv2 is running within AS200, including the 10.0.0.0 and 172.30.0.0 network
statements and passive interfaces.
IBGP is configured between SanJose1, SanJose2 and SanJose3, full-mesh.
No synchronization has been configured on all IBGP routers.
26
ISP1
router bgp 100
network 10.2.0.0 mask 255.255.255.0
ISP2
router bgp 300
network 172.16.0.0
SanJose3
router rip
version 2
network 192.168.1.0
!
router bgp 200
no synchronization
SanJose1
router rip
version 2
passive-interface Serial0/1
network 10.0.0.0
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
SanJose2
router rip
version 2
network 172.30.0.0
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
27 200
neighbor 192.168.1.198 remote-as
B
172.16.0.0/16 [200/0] via 172.30.1.1, 00:17:56
R
172.30.0.0/16 [120/2] via 192.168.1.194, 00:00:20, Serial0/0
B
10.2.0.0 [20/0] via 10.1.0.1, 00:17:50
C
R
192.168.1.96/27 [120/2] via 192.168.1.194, 00:00:20, Serial0/0
C
R
192.168.1.32/27 [120/1] via 192.168.1.194, 00:00:20, Serial0/0
C
R
192.168.1.196/30 [120/1] via 192.168.1.194, 00:00:20, Serial0/0
Network
Next Hop
*> 10.2.0.0/24
10.1.0.1
*>i172.16.0.0
172.30.1.1
*> 192.168.1.0
0.0.0.0
* i
192.168.1.197

0
0 100 i
0
100
0 300 i
0
32768 i
0
100
0 i
Now that the networks between the ISPs and the boundary routers are
being propagated by RIP, the boundary routers now have a route to the
BGP next hops.
Because of this the BGP routes to the opposite ISP network is
28
reachable and in the routing table.
B
172.16.0.0/16 [20/0] via 172.30.1.1, 00:21:51
C
B
10.2.0.0/24 [200/0] via 10.1.0.1, 00:21:40
R
10.0.0.0/8 [120/2] via 192.168.1.198, 00:00:03, Serial0/0
C
R
192.168.1.64/27 [120/2] via 192.168.1.198, 00:00:03, Serial0/0
R
192.168.1.32/27 [120/1] via 192.168.1.198, 00:00:03, Serial0/0
R
192.168.1.192/30 [120/1] via 192.168.1.198, 00:00:03, Serial0/0
C
Network
Next Hop
*>i10.2.0.0/24
10.1.0.1
*> 172.16.0.0
172.30.1.1
* i192.168.1.0
192.168.1.193
*>
0.0.0.0

0
100
0 100 i
0
0 300 i
0
100
0 i
0
32768 i
Not only is the 10.2.0.0 network reachable, but the 10.0.0.0/8 classful
network, which includes 10.1.0.0/24. This will come into play in a moment.
The BGP route to the opposite ISP network is reachable and in the
29
routing table.
B
R
B
R
R
R
C
B
C
C
172.16.0.0/16 [200/0] via 172.30.1.1, 00:20:06

172.30.0.0/16 [120/1] via 192.168.1.197, 00:00:12, Serial0/1
10.2.0.0/24 [200/0] via 10.1.0.1, 00:19:56
10.0.0.0/8 [120/1] via 192.168.1.193, 00:00:11, Serial0/0
192.168.1.96/27 [120/1] via 192.168.1.197, 00:00:12, Serial0/1
192.168.1.64/27 [120/1] via 192.168.1.193, 00:00:11, Serial0/0
192.168.1.0/24 [200/0] via 192.168.1.193, 00:19:53
Network
*>i10.2.0.0/24
*>i172.16.0.0
*>i192.168.1.0
* i
Next Hop
10.1.0.1
172.30.1.1
192.168.1.193
192.168.1.197

0
100
0 100 i
0
100
0 300 i
0
100
0 i
0
100
0 i
30
ISP1#show ip route
B
C
C
B
172.16.0.0/16 [20/0] via 10.1.0.2, 00:15:14

10.1.0.0 is directly connected, Serial0
192.168.1.0/24 [20/0] via 10.1.0.2, 00:15:14
ISP1#show ip bgp
Network
*> 10.2.0.0/24
*> 172.16.0.0
*> 192.168.1.0
Next Hop
0.0.0.0
10.1.0.2
10.1.0.2

0
32768 i
0 200 300 i
0
0 200 i
The 172.16.0.0 network is now propagated from SanJose1 to ISP1.
31
ISP2#show ip route
C
C
B
B

10.2.0.0 [20/0] via 172.30.1.2, 00:23:09
192.168.1.0/24 [20/0] via 172.30.1.2, 00:23:09
ISP2#show ip bgp
Network
*> 10.2.0.0/24
*> 172.16.0.0
*> 192.168.1.0
Next Hop
172.30.1.2
0.0.0.0
172.30.1.2
Metric LocPrf Weight

0
0
32768
0
0
Path
200 100 i
i
200 i
The 10.2.0.0 network (and the 10.0.0.0 network) is now propagated

from SanJose1 to ISP1.
32
SanJose1#ping 172.16.1.1
Sending 5, 100-byte ICMP Echos to 172.16.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 84/84/84
!!!!!
------------------------------------SanJose3#ping 172.16.1.1
!!!!!
!!!!!
--------------------------------------SanJose2#ping 172.16.1.1
!!!!!
!!!!!
ms
ms
ms
ms
ms
ms
33
ISP2#ping 10.2.0.1
Type escape sequence to abort.
.....
Success rate is 0 percent (0/5)
I thought you said all routers could reach all networks?

What happened???
ISP2 can reach 10.2.0.1
What do you think the problem is?
34
ISP2#show ip route
C
C
B
10.2.0.0 [20/0] via 172.30.1.2, 00:23:09
B
192.168.1.0/24 [20/0] via 172.30.1.2, 00:23:09
ISP2#show ip bgp
Network
*> 10.2.0.0/24
*> 172.16.0.0
*> 192.168.1.0
Next Hop
172.30.1.2
0.0.0.0
172.30.1.2

0
0
32768
0
0
Path
200 100 i
i
200 i
ISP1#show ip route
B
172.16.0.0/16 [20/0] via 10.1.0.2, 00:15:14
C
C
B
192.168.1.0/24 [20/0] via 10.1.0.2, 00:15:14
ISP1#show ip bgp
Network
*> 10.2.0.0/24
*> 172.16.0.0
*> 192.168.1.0
Next Hop
0.0.0.0
10.1.0.2
10.1.0.2

0
32768 i
0 200 300 i
0
0 200 I
ISP2 does have a route to 10.2.0.0 and ISP1 has a route to 172.16.0.0
The ping uses a source IP address of 172.30.1.1 and ISP1 does NOT have a route
35
back to this destination address. (This also applies to pings from ISP1.)
Extended pings with a reachable source IP address will work!
ISP2#ping
Protocol [ip]:
Target IP address: 10.2.0.1
Repeat count [5]:
Datagram size [100]:
Timeout in seconds [2]:
Extended commands [n]: y
Source address or interface: 172.16.1.1
Type of service [0]:
Set DF bit in IP header? [no]:
Validate reply data? [no]:
Data pattern [0xABCD]:
Loose, Strict, Record, Timestamp, Verbose[none]:
Sweep range of sizes [n]:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 108/112/116 ms
ISP2 does have a route to 10.2.0.0 and ISP1 has a route to 172.16.0.0
The ping uses a source IP address of 172.30.1.1 and ISP1 does NOT have
a route back to this destination address. (This also applies to pings from
ISP1.)
Extended pings with a reachable source IP address will work!
36
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
IBGP
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Scenario B: Step #1 IBGP between SanJose1 and SanJose2 only

RIPv2 is running within AS200, including the 10.0.0.0 and 172.30.0.0
network statements and passive interfaces.
IBGP is configured only between SanJose1, SanJose2.
37
ISP1
router bgp 100
network 10.2.0.0 mask 255.255.255.0
ISP2
router bgp 300
network 172.16.0.0
SanJose3
router rip
version 2
network 192.168.1.0
!
no router bgp 200
SanJose1
router rip
version 2
network 10.0.0.0
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
SanJose2
router rip
version 2
network 172.30.0.0
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
38
Results
SanJose3 does NOT have a route to 172.16.0.0 on ISP2.
SanJose3 does have a route to 10.0.0.0 on ISP1 only because
SanJose1 is propagating its 10.1.0.0/24 subnet, so it has a route to the
classful 10.0.0.0 network.
SanJose3 does NOT have the specific ISP networks in its routing table
so it may drop packets as it is passed through between SanJose1 and
SanJose2.
SanJose1 and SanJose2 can NOT reach the opposite ISP2s
172.16.1.1 because SanJose3 does not have a route to 172.16.0.0
on ISP1.
However, SanJose1, SanJose2 and SanJose3 can reach ISP1s
10.2.0.1 because RIP is sending the 10.0.0.0 classful route which
includes the serial link between SJ1 and ISP1.
ISPs They will NOT be able to ping each other, because SJ3 has a
route for the 10.0.0.0 network but NOT for the 172.16.0.0 network, thus
pings from ISP2 to ISP1 will get to ISP2 but the replies will get dropped
by SJ3 as it does not have a route for 172.16.0.0.
39
ISP1#show ip route
B
C
C
B
172.16.0.0/16 [20/0] via 10.1.0.2, 00:34:13

192.168.1.0/24 [20/0] via 10.1.0.2, 00:38:39
This looks fine

for ISP1, but
lets look at the
AS200 routers.
ISP1#show ip bgp
Network
*> 10.2.0.0/24
*> 172.16.0.0
*> 192.168.1.0
Next Hop
0.0.0.0
10.1.0.2
10.1.0.2

0
32768 i
0 200 300 i
0
0 200 i
ISP1#ping 192.168.1.33
!!!!!
ISP1#ping 172.16.1.1
MAKE THIS AN EXTENDED PING

U.U.U
40
B
172.16.0.0/16 [200/0] via 172.30.1.1, 00:35:40
R
172.30.0.0/16 [120/2] via 192.168.1.194, 00:00:04, Serial0/0
B
10.2.0.0 [20/0] via 10.1.0.1, 00:40:06
C
R
192.168.1.96/27 [120/2] via 192.168.1.194, 00:00:04, Serial0/0
C
R
192.168.1.32/27 [120/1] via 192.168.1.194, 00:00:04, Serial0/0
C
R
192.168.1.196/30 [120/1] via 192.168.1.194, 00:00:05, Serial0/0
This looks fine

for SJ1, but
lets look at
SJ3.
Network
Next Hop
*> 10.2.0.0/24
10.1.0.1
*>i172.16.0.0
172.30.1.1
* i192.168.1.0
192.168.1.197
*>
0.0.0.0

0
0 100 i
0
100
0 300 i
0
100
0 i
0
32768 i
!!!!!
U.U.U
41
R
172.30.0.0/16 [120/1] via 192.168.1.197, 00:00:25, Serial0/1
R
10.0.0.0/8 [120/1] via 192.168.1.193, 00:00:06, Serial0/0
R
192.168.1.96/27 [120/1] via 192.168.1.197, 00:00:25, Serial0/1
R
192.168.1.64/27 [120/1] via 192.168.1.193, 00:00:06, Serial0/0
C
Wheres
C
C
172.16.0.0?
!!!!!
.....
Notice that SanJose3 does NOT have a route for 172.16.0.0

172.16.0.0.
When a packet with a destination address for this network reaches
SanJose3 it drops it.
Remember, we removed SanJose3 as an IBGP peer.
42
B
172.16.0.0/16 [20/0] via 172.30.1.1, 00:43:13
C
B
10.2.0.0/24 [200/0] via 10.1.0.1, 00:38:57
R
10.0.0.0/8 [120/2] via 192.168.1.198, 00:00:04, Serial0/0
C
R
192.168.1.64/27 [120/2] via 192.168.1.198, 00:00:04, Serial0/0
R
192.168.1.32/27 [120/1] via 192.168.1.198, 00:00:05, Serial0/0
R
192.168.1.192/30 [120/1] via 192.168.1.198, 00:00:05, Serial0/0
C
Again,this
looks fine for
SJ2, but
172.16.0.0 does
NOT exist on
SJ3.
Network
Next Hop
*>i10.2.0.0/24
10.1.0.1
*> 172.16.0.0
172.30.1.1
* i192.168.1.0
192.168.1.193
*>
0.0.0.0

0
100
0 100 i
0
0 300 i
0
100
0 i
0
32768 i
!!!!!
!!!!!
43
ISP2#show ip route
C
C
B
10.2.0.0 [20/0] via 172.30.1.2, 00:44:58
B
192.168.1.0/24 [20/0] via 172.30.1.2, 00:49:15
ISP2#show ip bgp
Network
Next Hop
*> 10.2.0.0/24
172.30.1.2
*> 172.16.0.0
0.0.0.0
*> 192.168.1.0
172.30.1.2
This also looks

fine for ISP1, but
remember the
packets cant
return to ISP2
because SJ3 does
not have a route
for 172.16.0.0.

0 200 100 i
0
32768 i
0
0 200 i
ISP2#ping 10.2.0.1
.....
ISP2#ping 192.168.1.65
!!!!!
44
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
SanJose1
.2 S1
10.1.0.0/24
S0 .1
ISP1
.1
Complete
reachability
IBGP
S0
.197
.97
SanJose2
.2 S1
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Step 2 - The Solution, Redistributing BGP routes into RIP

RIPv2 is running within AS200, including the 10.0.0.0 and 172.30.0.0 network
statements and passive interfaces.
IBGP is configured only between SanJose1, SanJose2.
172.16.0.0 BGP routes are redistributed into AS200 via RIP
45
ISP1
router bgp 100
network 10.2.0.0 mask 255.255.255.0
ISP2
router bgp 300
network 172.16.0.0
SanJose3
router rip
version 2
network 192.168.1.0
SanJose1
router rip
version 2
network 10.0.0.0
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
SanJose2
router rip
version 2
redistribute bgp 200 metric 2
network 172.30.0.0
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
!
46
As before.
B
172.16.0.0/16 [20/0] via 172.30.1.1, 01:22:20
C
B
10.2.0.0/24 [200/0] via 10.1.0.1, 01:18:04
R
10.0.0.0/8 [120/2] via 192.168.1.198, 00:00:21, Serial0/0
C
R
192.168.1.64/27 [120/2] via 192.168.1.198, 00:00:21, Serial0/0
R
192.168.1.32/27 [120/1] via 192.168.1.198, 00:00:21, Serial0/0
R
192.168.1.192/30 [120/1] via 192.168.1.198, 00:00:21, Serial0/0
C
BGP table version is 4, local router ID is 192.168.1.97
Status codes: s suppressed, d damped, h history, * valid, > best, i - internal
Origin codes: i - IGP, e - EGP, ? - incomplete
Network
*>i10.2.0.0/24
*> 172.16.0.0
* i192.168.1.0
*>
SanJose2#
Next Hop
10.1.0.1
172.30.1.1
192.168.1.193
0.0.0.0

0
100
0 100 i
0
0 300 i
0
100
0 i
0
32768 i
47
ISP2#show ip route
C
C
B
10.2.0.0 [20/0] via 172.30.1.2, 01:19:44
B
192.168.1.0/24 [20/0] via 172.30.1.2, 01:24:01
ISP2#show ip bgp
Network
Next Hop
*> 10.2.0.0/24
172.30.1.2
*> 172.16.0.0
0.0.0.0
*> 192.168.1.0
172.30.1.2
As before.

0 200 100 i
0
32768 i
0
0 200 i
48
R
R
R
172.16.0.0/16 [120/2] via 192.168.1.197, 00:00:04, Serial0/1

172.30.0.0/16 [120/1] via 192.168.1.197, 00:00:04, Serial0/1
10.0.0.0/8 [120/1] via 192.168.1.193, 00:00:25, Serial0/0
R
192.168.1.96/27 [120/1] via 192.168.1.197, 00:00:04, Serial0/1
R
192.168.1.64/27 [120/1] via 192.168.1.193, 00:00:25, Serial0/0
C
C
C
SanJose3 now has a route for 172.16.0.0 which was redistributed into
RIP by SanJose2.
This gives us the rest of the connectivity we need.
10.0.0.0 network was already in SanJose2s routing table because
SanJose1s has a RIP network 10.0.0.0 statement.
Remember 10.2.0.0 and 10.1.0.0 belong to this classful network.
49
R
R
172.16.0.0/16 [120/3] via 192.168.1.194, 00:00:02, Serial0/0

172.30.0.0/16 [120/2] via 192.168.1.194, 00:00:02, Serial0/0
B
10.2.0.0 [20/0] via 10.1.0.1, 01:28:18
C
R
192.168.1.96/27 [120/2] via 192.168.1.194, 00:00:02, Serial0/0
C
R
192.168.1.32/27 [120/1] via 192.168.1.194, 00:00:02, Serial0/0
C
R
192.168.1.196/30 [120/1] via 192.168.1.194, 00:00:02, Serial0/0
Network
*> 10.2.0.0/24
*>i172.16.0.0
* i192.168.1.0
*>
Next Hop
10.1.0.1
172.30.1.1
192.168.1.197
0.0.0.0

0
0 100 i
0
100
0 300 i
0
100
0 i
0
32768 i
172.16.0.0 is now redistributed via RIP which has a lower administrative

distance (120) than IBGP (200); (EBGP is 20), which was how this
route was in SanJose1s routing table prior to redistribution by
SanJose2..
50
ISP1#show ip route
B
C
C
B
172.16.0.0/16 [20/0] via 10.1.0.2, 01:25:26

192.168.1.0/24 [20/0] via 10.1.0.2, 01:29:53
As before.
ISP1#show ip bgp
Network
*> 10.2.0.0/24
*> 172.16.0.0
*> 192.168.1.0
Next Hop
0.0.0.0
10.1.0.2
10.1.0.2

0
32768 i
0 200 300 i
0
0 200 i
51
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
IBGP
SanJose1
.2 S1
ISP1
.1
.197
.97
SanJose2
.2 S1
10.1.0.0/24
S0 .1
S0
172.30.1.0/24
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Scenario C: Another Way Step #1

What if we did not want to redistribute the serial networks between the boundary routers and the ISPs
into our rip network.
network What would be the consequences? How could we fix them?
Eventually, for ISP1 to reach ISP2 we will use next-hop attribute.
SanJose1 can ping 172.16.1.1. We will see that SanJose1 has an IGP route to 172.16.0.0 as this was
redistributed by SanJose2 from BGP into rip, so it can ping 172.16.1.1. However, the 172.16.0.0 route
in the BGP routing table shows a next hop of 172.30.1.1, in which SanJose1 does not have a route for.
Because BGP does not have a route for the next hop, it does not include this route in its update to its
EBGP peer, ISP1.
ISP2 does have a route to ISP1s 10.2.0.0/24 network because RIP is advertising the classful 10.0.0.0
network which includes the serial link between ISP1 and SanJose1. However, it can NOT ping it52
because the reply can NOT get back as explained in the previous paragraph.
ISP1
router bgp 100
network 10.2.0.0 mask 255.255.255.0
ISP2
router bgp 300
network 172.16.0.0
SanJose3
router rip
version 2
network 192.168.1.0
SanJose1
router rip
version 2
no network 10.0.0.0
network 192.168.1.0
router bgp 200
no synchronization
network 192.168.1.0
SanJose2
router rip
version 2
no network 172.30.0.0
network 192.168.1.0
router bgp 200
no synchronization
network 192.168.1.0
53
!
ISP1#show ip route
C
C
B
10.0.0.0/24 is
10.2.0.0 is
10.1.0.0 is
192.168.1.0/24
ISP1#show ip bgp
Network
*> 10.2.0.0/24
*> 192.168.1.0
subnetted, 2 subnets
directly connected, Loopback0
directly connected, Serial0
[20/0] via 10.1.0.2, 00:08:34
Notice: No route for 172.16.0.0/16 (well see why)

Next Hop
0.0.0.0
10.1.0.2

0
32768 i
0
0 200 i
ISP1#ping 192.168.1.97
!!!!!
ISP1#ping
Protocol [ip]:
.....
ISP1#ping 172.30.1.1
(Note: This is normal behavior and may not be necessary.)

.....
54
R
172.16.0.0/16 [120/3] via 192.168.1.194, 00:00:19, Serial0/0

B
10.2.0.0 [20/0] via 10.1.0.1, 01:20:45
Known via RIP
C
because is was
R
192.168.1.96/27 [120/2] via 192.168.1.194, 00:00:19, Serial0/0 redistributed from
C
BGP.
R
192.168.1.32/27 [120/1] via 192.168.1.194, 00:00:19, Serial0/0
C
NOT sent to ISP1
R
192.168.1.196/30 [120/1] via 192.168.1.194, 00:00:19, Serial0/0
via EBGP because
Network
Next Hop
*> 10.2.0.0/24
10.1.0.1
* i172.16.0.0
172.30.1.1
*> 192.168.1.0
0.0.0.0
* i
192.168.1.197

0
0 100 i
0
100
0 300 i
0
32768 i
0
100
0 i
Next hop is not

reachable.
EBGP routes carried into IBGP use the external

EBGP next hop.

!!!!!
!!!!!
55
R
R
172.16.0.0/16 [120/2] via 192.168.1.197, 00:00:09, Serial0/1

10.0.0.0/8 [120/1] via 192.168.1.193, 00:00:22, Serial0/0
R
192.168.1.96/27 [120/1] via 192.168.1.197, 00:00:09, Serial0/1
R
192.168.1.64/27 [120/1] via 192.168.1.193, 00:00:22, Serial0/0
C
C
C
SanJose3#
!!!!!
!!!!!
56
B
C
B
R
C
R
R
R
C
172.16.0.0/16 [20/0] via 172.30.1.1, 01:23:21

10.2.0.0/24 [200/0] via 10.1.0.1, 00:31:24
10.0.0.0/8 [120/2] via 192.168.1.198, 00:00:25, Serial0/0
192.168.1.64/27 [120/2] via 192.168.1.198, 00:00:25, Serial0/0
192.168.1.32/27 [120/1] via 192.168.1.198, 00:00:25, Serial0/0
192.168.1.192/30 [120/1] via 192.168.1.198, 00:00:25, Serial0/0
SanJose2
sends
10.2.0.0 to ISP2 via
EBGP because Next
hop of 10.1.0.1 is
reachable, as part of
the 10.0.0.0/8 net.
Network
Next Hop
*>i10.2.0.0/24
10.1.0.1
0
100
0 100 i
*> 172.16.0.0
172.30.1.1
0
0 300 i
* i192.168.1.0
192.168.1.193
0
100
0 i
*>
0.0.0.0
0
32768 i
RIP redistributed the
entire classful
!!!!!
10.0.0.0/8 network.
!!!!!
57
ISP2#show ip route
C
C
B
B

10.2.0.0 [20/0] via 172.30.1.2, 00:33:01
192.168.1.0/24 [20/0] via 172.30.1.2, 01:25:53
ISP2#show ip bgp
Network
Next Hop
*> 10.2.0.0/24
172.30.1.2
*> 172.16.0.0
0.0.0.0
*> 192.168.1.0
172.30.1.2
ISP2#ping
Protocol [ip]:
.....
Notice: There is a route for 10.2.0.0

0
0
32768
0
0
Path
200 100 i
i
200 i
58
Lo0
AS200
RIP
version 2
192.168.1.32/27
.33
.194 SanJose3 .198

S0
S1
192.168.1.192/30
192.168.1.196/30
192.168.1.64/27
Lo0
EBGP
.65
.193 S0
IBGP
SanJose1
.2 S1
10.1.0.0/24
.197
.97
SanJose2
.2 S1
172.30.1.0/24
S0 .1
ISP1
.1
Complete
reachability
S0
192.168.1.96/27
Lo0
EBGP
.1 S0
ISP2
.1
10.2.0.0/24
Lo0
172.16.1.0/24
Lo0
AS100
AS300
Step #2 - The Solution next

next--hop
hop--self

EBGP routes carried into IBGP use the external EBGP next hop. SanJose1s BGP
table has 172.30.1.1 as the next hop for 172.16.0.0/16, but does not have a route to
172.30.1.1, so it does not forward this route to ISP1.
BGP Path Selection: If the path specifies a next hop that is inaccessible, drop the
update. (coming)
Lets change the next hop to be the internal IBGP next hop.
IBGP peers SanJose1 and SanJose2 will use next-hop self in peering with each
other. This will change to next hops from the EBGP next hop to the internal IBGP
next hop.
59
All works!!!
ISP1
router bgp 100
network 10.2.0.0 mask 255.255.255.0
ISP2
router bgp 300
network 172.16.0.0
SanJose3
router rip
version 2
network 192.168.1.0
SanJose1
router rip
version 2
no network 10.0.0.0 (from before)
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
neighbor 192.168.1.197 next-hop-self
SanJose2
router rip
version 2
no network 172.30.0.0 (from before)
network 192.168.1.0
!
router bgp 200
no synchronization
network 192.168.1.0
neighbor 192.168.1.193 next-hop-self
!
60
ISP1#show ip route
B
C
C
B
172.16.0.0/16 [20/0] via 10.1.0.2, 00:03:49

192.168.1.0/24 [20/0] via 10.1.0.2, 00:03:22
Looks good!
ISP1#show ip bgp
Network
*> 10.2.0.0/24
*> 172.16.0.0
*> 192.168.1.0
Next Hop
0.0.0.0
10.1.0.2
10.1.0.2

0
32768 i
0 200 300 i
0
0 200 i
ISP1#ping 192.168.1.33
!!!!!
ISP1#ping
Protocol [ip]:
Extended commands [n]: y
Source address or interface: 10.2.0.1
!!!!!
ISP1#
61
SanJose2: neighbor 192.168.1.193 next-hop-self (->192.168.1.197)

R
B
C
R
C
R
R
R
172.16.0.0/16 [120/3] via 192.168.1.194, 00:00:24, Serial0/0

Looks good!
10.2.0.0 [20/0] via 10.1.0.1, 00:05:51
192.168.1.96/27 [120/2] via 192.168.1.194, 00:00:24, Serial0/0
192.168.1.32/27 [120/1] via 192.168.1.194, 00:00:24, Serial0/0
192.168.1.32/27 [120/1] via 192.168.1.194, 00:00:24, Serial0/0
192.168.1.196/30 [120/1] via 192.168.1.194, 00:00:25, Serial0/0
The route to the next hop

Network
*> 10.2.0.0/24
*>i172.16.0.0
* i192.168.1.0
*>
Effect of the next-hop-self (was 172.30.1.1)
Next Hop
10.1.0.1
192.168.1.197
192.168.1.197
0.0.0.0

0
0 100 i
0
100
0 300 i
0
100
0 i
0
32768 i
62
R
R
R
R
C
C
C
172.16.0.0/16 [120/2] via 192.168.1.197, 00:00:05, Serial0/1

10.0.0.0/8 [120/1] via 192.168.1.193, 00:00:26, Serial0/0
192.168.1.96/27 [120/1] via 192.168.1.197, 00:00:05, Serial0/1
192.168.1.64/27 [120/1] via 192.168.1.193, 00:00:27, Serial0/0
63
SanJose1: neighbor 192.168.1.197 next-hop-self (->192.168.1.193)

B
C
B
R
C
R
R
R
C
172.16.0.0/16 [20/0] via 172.30.1.1, 00:09:30

10.2.0.0/24 [200/0] via 192.168.1.193, 00:09:23
10.0.0.0/8 [120/2] via 192.168.1.198, 00:00:17, Serial0/0
192.168.1.64/27 [120/2] via 192.168.1.198, 00:00:17, Serial0/0
192.168.1.32/27 [120/1] via 192.168.1.198, 00:00:17, Serial0/0
192.168.1.192/30 [120/1] via 192.168.1.198, 00:00:17, Serial0/0
Network
*>i10.2.0.0/24
*> 172.16.0.0
* i192.168.1.0
*>
The route to the next hop

Affect of the next-hop-self (was 10.1.0.1)
Next Hop
192.168.1.193
172.30.1.1
192.168.1.193
0.0.0.0

0
100
0 100 i
0
0 300 i
0
100
0 i
0
32768 i
64
BGP Basic Labs

Hopefully you will not get bored
Emanuele Mazza
CCIE 11957
[email protected]
www.linkedin.com/in/emanuelemazza
Presentation_ID
2006 Cisco Systems, Inc. All rights reserved.
Cisco Confidential
65

BGP Basic Labs

Uploaded by

BGP Basic Labs

Uploaded by

BGP Basic Labs

Hopefully you will not get bored

2006 Cisco Systems, Inc. All rights reserved.

.194 SanJose3 .198

With very few

With very few exceptions,

.194 SanJose3 .198

.194 SanJose3 .198

.194 SanJose3 .198

Scenario A: Step #1 EBGP and IGP (RIPv2)

.194 SanJose3 .198

ISPs only see the BGP

172.16.0.0/24 is subnetted, 1 subnets

Metric LocPrf Weight Path

SanJose1 only sees the BGP

10.0.0.0/24 is subnetted, 2 subnets

Metric LocPrf Weight Path

s The table entry is suppressed.

* The table entry is valid.

SanJose2 only sees the

172.16.0.0/16 [20/0] via 172.30.1.1, 00:27:08

Metric LocPrf Weight Path

192.168.1.0/24 is variably subnetted, 5 subnets, 2 masks

.194 SanJose3 .198

Step #2 EBGP, IBGP and IGP (RIPv2)

.194 SanJose3 .198

SanJose1 still only sees the BGP

10.0.0.0/24 is subnetted, 2 subnets

Metric LocPrf Weight Path

Even though 172.16.0.0 network is in BGP table it is not in routing table

172.16.0.0/16 [20/0] via 172.30.1.1, 00:11:27

SanJose2 still only sees

Metric LocPrf Weight Path

Even though 10.2.0.0 network is in BGP table it is not in routing table

.194 SanJose3 .198

Step #3 Adding SanJose3 as a IBGP peer

.194 SanJose3 .198

.194 SanJose3 .198

SanJose1 still only sees the

10.0.0.0/24 is subnetted, 2 subnets

Metric LocPrf Weight Path

Even though 172.16.0.0 network is in BGP table it is not in routing

SanJose2 still only sees

172.16.0.0/16 [20/0] via 172.30.1.1, 00:11:27

Metric LocPrf Weight Path

Even though 10.2.0.0 network is in BGP table it is not in routing

.194 SanJose3 .198

Metric LocPrf Weight Path

Metric LocPrf Weight Path

172.16.0.0/16 [200/0] via 172.30.1.1, 00:20:06

Metric LocPrf Weight Path

172.16.0.0/16 [20/0] via 10.1.0.2, 00:15:14

Metric LocPrf Weight Path

The 172.16.0.0 network is now propagated from SanJose1 to ISP1.

172.16.0.0/24 is subnetted, 1 subnets

Metric LocPrf Weight

The 10.2.0.0 network (and the 10.0.0.0 network) is now propagated

I thought you said all routers could reach all networks?

Metric LocPrf Weight

Metric LocPrf Weight Path

.194 SanJose3 .198

Scenario B: Step #1 IBGP between SanJose1 and SanJose2 only

172.16.0.0/16 [20/0] via 10.1.0.2, 00:34:13

This looks fine

Metric LocPrf Weight Path

MAKE THIS AN EXTENDED PING

Sending 5, 100-byte ICMP Echos to 172.16.1.1, timeout is 2 seconds:

This looks fine

Metric LocPrf Weight Path

Notice that SanJose3 does NOT have a route for 172.16.0.0

Metric LocPrf Weight Path

This also looks

Metric LocPrf Weight Path

.194 SanJose3 .198

Step 2 - The Solution, Redistributing BGP routes into RIP

Metric LocPrf Weight Path