From: alberto vargas (avargasn@yahoo.com.br)
Date: Sun Oct 15 2006 - 18:46:55 ART
Hello all,
I'm facing a problem in a lab I'm not being able to
solve it.
This lab runs OSPF, EIGRP and BGP. R3 is the an ASBR
and is also the only way to get to EIGRP domain. So R3
is injecting a default route into OSPF and
redistributing OSPF into EIGRP.
R4 is directly connected to R3 through a frame relay
network and it runs both OSPF and BGP. R3 and R4 are
in differents AS, 100 and 400 respectively. The whole
network use subnets take from 164.1.0.0/16 block.
For clarifying 164.1.45.5 is the IP address of R5
router that receives the default route from R3 and
announces it to R4 via an OSPF virtual link. R4 also
receives that route from R3 but due to virtual link R4
chooses the default route receveid from R5.
Everything works fine until I configure R4 to send a
aggregate route, 164.1.0.0/18, to its EBGP peer BB3
whose ip address is 204.12.1.254.
Following is the routing table of R4 without the
discard route installed by the aggregate command:
C 204.12.1.0/24 is directly connected, Ethernet1/1
11.0.0.0/24 is subnetted, 1 subnets
B 11.11.11.0 [20/0] via 204.12.1.254, 00:51:51
164.1.0.0/16 is variably subnetted, 7 subnets, 2
masks
O 164.1.35.0/24 [110/128] via 164.1.45.5,
00:52:00, Serial0/1
C 164.1.34.0/24 is directly connected, Serial0/0
C 164.1.45.0/29 is directly connected, Serial0/1
C 164.1.47.0/24 is directly connected,
Ethernet1/0
O 164.1.55.0/24 [110/74] via 164.1.45.5,
00:52:00, Serial0/1
O 164.1.5.0/24 [110/74] via 164.1.45.5,
00:52:01, Serial0/1
O 164.1.7.0/24 [110/20] via 164.1.47.7,
00:51:27, Ethernet1/0
150.1.0.0/16 is variably subnetted, 3 subnets, 2
masks
O 150.1.5.0/24 [110/65] via 164.1.45.5,
00:52:11, Serial0/1
C 150.1.4.0/24 is directly connected, Loopback0
O 150.1.7.7/32 [110/11] via 164.1.47.7,
00:51:27, Ethernet1/0
O*E2 0.0.0.0/0 [110/1] via 164.1.45.5, 00:51:27,
Serial0/1
Following is the output of a debug ip packet when BGP
discard route wasn't installed:
Mar 1 00:53:15.815: IP: tableid=0, s=164.1.45.4
(local), d=164.1.26.6 (Serial0/1), routed via FIB
*Mar 1 00:53:15.815: IP: s=164.1.45.4 (local),
d=164.1.26.6 (Serial0/1), len 100, sending
*Mar 1 00:53:15.815: ICMP type=8, code=0
*Mar 1 00:53:16.071: IP: tableid=0, s=164.1.45.4
(local), d=164.1.26.6 (Serial0/1), routed via FIB
*Mar 1 00:53:16.075: IP: s=164.1.45.4 (local),
d=164.1.26.6 (Serial0/1), len 100, sending
*Mar 1 00:53:16.079: ICMP type=8, code=0
*Mar 1 00:53:16.319: IP: tableid=0, s=164.1.45.4
(local), d=164.1.26.6 (Serial0/1), routed via FIB
*Mar 1 00:53:16.323: IP: s=164.1.45.4 (local),
d=164.1.26.6 (Serial0/1), len 100, sending
*Mar 1 00:53:16.323: ICMP type=8, code=0
Now the discard is installed as you can see:
C 204.12.1.0/24 is directly connected, Ethernet1/1
11.0.0.0/24 is subnetted, 1 subnets
B 11.11.11.0 [20/0] via 204.12.1.254, 00:47:37
164.1.0.0/16 is variably subnetted, 10 subnets, 3
masks
O 164.1.35.0/24 [110/128] via 164.1.45.5,
00:47:47, Serial0/1
C 164.1.34.0/24 is directly connected, Serial0/0
C 164.1.45.0/29 is directly connected, Serial0/1
C 164.1.47.0/24 is directly connected,
Ethernet1/0
O 164.1.55.0/24 [110/74] via 164.1.45.5,
00:47:47, Serial0/1
O 164.1.5.0/24 [110/74] via 164.1.45.5,
00:47:47, Serial0/1
O 164.1.7.0/24 [110/20] via 164.1.47.7,
00:47:14, Ethernet1/0
B 164.1.0.0/18 [200/0] via 0.0.0.0, 00:47:10,
Null0
B 164.1.3.0/24 [20/0] via 150.1.3.3, 00:47:10
B 164.1.8.0/24 [20/0] via 150.1.3.3, 00:46:40
150.1.0.0/16 is variably subnetted, 3 subnets, 2
masks
O 150.1.5.0/24 [110/65] via 164.1.45.5,
00:47:57, Serial0/1
C 150.1.4.0/24 is directly connected, Loopback0
O 150.1.7.7/32 [110/11] via 164.1.47.7,
00:47:14, Ethernet1/0
O*E2 0.0.0.0/0 [110/1] via 164.1.45.5, 00:47:14,
Serial0/1
Follwoing is the output of a debug ip packet when the
discard route was installed:
*Mar 1 00:50:57.387: IP: tableid=0, s=164.1.34.4
(local), d=164.1.26.6 (Null0), routed via RIB
*Mar 1 00:50:57.387: IP: s=164.1.34.4 (local),
d=164.1.26.6 (Null0), len 100, sending
*Mar 1 00:50:57.387: ICMP type=8, code=0
*Mar 1 00:50:59.387: IP: tableid=0, s=164.1.34.4
(local), d=164.1.26.6 (Null0), routed via RIB
*Mar 1 00:50:59.391: IP: s=164.1.34.4 (local),
d=164.1.26.6 (Null0), len 100, sending
*Mar 1 00:50:59.395: ICMP type=8, code=0
*Mar 1 00:51:01.387: IP: tableid=0, s=164.1.34.4
(local), d=164.1.26.6 (Null0), routed via RIB
*Mar 1 00:51:01.391: IP: s=164.1.34.4 (local),
d=164.1.26.6 (Null0), len 100, sending
*Mar 1 00:51:01.395: ICMP type=8, code=0
In both debugs the ping command was issued on router
R4.
I can see that the 164.1.0.0/18 is a longest match to
network 164.1.26.0/24 than the default route. How can
I solve that? An possible answer is not to allow the
discard route installation, so I ask what would the
gurus do?
Regards,
Alberto
This archive was generated by hypermail 2.1.4 : Wed Nov 01 2006 - 07:29:05 ART