07-08-2011 01:25 AM - edited 03-04-2019 12:55 PM
Hi,
Please see the attached diagram.
We have SAP servers at DC from which invoice prints will to to Agri printers 172.16.11.0/24 (P1 P2 P3).
The router 10.30.1.2 is connected to Agri.
We have static route from 10.30.1.1 to Agri. (ip route 172.16.11.0 255.255.255.0 10.30.1.2)
We have MPLS and DMVPN connectivity between DC and US.
The Agri network (172.16.11.0) is advertised on both MPLS (BGP) and DMVPN (EIGRP) at US.
Now the issues is that the network 172.16.11.0 is not advertising via eigrp between 10.30.1.1 and 10.30.1.3
I found that the network 172.16.11.0 at 10.30.1.3 is learning via eigrp from DC (10.30.1.1-->BGP-->MPLS-->BGP-->10.246.6.1-->EIGRP-->192.168.199.12-->DIGRP-->10.30.1.3)
When I see sh ip route 172.16.11.0 in 10.30.1.3 found that it is learning from DC through eigrp 192.168.220.1 instead of 10.30.1.1
When the MPLS link is down the network 172.16.11.0 is not accessible from DC through EIGRP.
Could someone see on this issue and give me solution.
Please see the below routers config and output.
DATACENTER:
10.246.6.1
#sh ip route 172.16.11.0
Routing entry for 172.16.11.0/24
Known via "bgp 65100", distance 20, metric 0
Tag 1, type external
Redistributing via eigrp 100
Advertised by eigrp 100 metric 10000 1000 255 1 1500
Last update from 10.246.2.10 2d10h ago
Routing Descriptor Blocks:
* 10.246.2.10, from 10.246.2.10, 2d10h ago
Route metric is 0, traffic share count is 1
AS Hops 4
Route tag 1
#sh ip eig topology | in 172.16.11.0
P 172.16.11.0/24, 1 successors, FD is 512000, tag is 1
192.168.199.12
#sh ip route 172.16.11.0
Routing entry for 172.16.11.0/24
Known via "eigrp 100", distance 170, metric 514560
Tag 1, type external
Redistributing via nhrp, eigrp 100
Last update from 192.168.199.1 on GigabitEthernet0/1, 12:22:23 ago
Routing Descriptor Blocks:
* 192.168.199.1, from 192.168.199.1, 12:22:23 ago, via GigabitEthernet0/1
Route metric is 514560, traffic share count is 1
Total delay is 10100 microseconds, minimum bandwidth is 10000 Kbit
Reliability 255/255, minimum MTU 1500 bytes
Loading 1/255, Hops 1
Route tag 1
#sh ip eig top | in 172.16.11.0
P 172.16.11.0/24, 1 successors, FD is 514560, tag is 1
US
10.30.1.1
router eigrp 100
network 172.16.11.0 0.0.0.255
network 192.168.220.0
no auto-summary
router bgp 60600
no synchronization
bgp log-neighbor-changes
bgp bestpath as-path ignore
network 10.32.0.0 mask 255.255.0.0
network 172.16.11.0 mask 255.255.255.0
#sh ip route 172.16.11.0
Routing entry for 172.16.11.0/24
Known via "static", distance 1, metric 0
Advertised by bgp 60600
Routing Descriptor Blocks:
* 10.30.1.2
Route metric is 0, traffic share count is 1
#sh ip eig topo | in 172.16.11.0
P 172.16.11.0/24, 0 successors, FD is Inaccessible, tag is 1
10.30.1.3
router eigrp 100
network 10.30.0.0 0.0.255.255
network 10.32.0.0 0.0.255.255
network 172.16.11.0 0.0.0.255
network 192.168.220.0
#sh ip route 172.16.11.0
Routing entry for 172.16.11.0/24
Known via "eigrp 100", distance 170, metric 1794560
Tag 1, type external
Redistributing via eigrp 100
Last update from 192.168.220.1 on Tunnel10, 2d10h ago
Routing Descriptor Blocks:
* 192.168.220.1, from 192.168.220.1, 2d10h ago, via Tunnel10
Route metric is 1794560, traffic share count is 1
Total delay is 60100 microseconds, minimum bandwidth is 10000 Kbit
Reliability 255/255, minimum MTU 1400 bytes
Loading 8/255, Hops 2
Route tag 1
#sh ip eig topo | in 172.16.11.0
P 172.16.11.0/24, 1 successors, FD is 1794560, tag is 1
Jopeti.
07-08-2011 02:50 AM
Network 172.16.11.0/24 is not advertized by EIGRP because 10.30.1.1 has no interface in this subnet.
Therefore the network statement under eigrp config is ignored. As a result the net is only advertized by BGP.
You would need something like "redistribute static" to have it advertized by EIGRP.
When BGP loses the connection I think the route would/could end up in 10.30.1.3 via eigrp.
Note that this is not an ideal solution.
The best solution would be to have the route to 172.16.11.0/24 advertized by router 10.30.1.2 to both 10.30.1.1 and 10.30.1.3.
regards,
Leo
07-08-2011 03:49 AM
Leo,
Network 172.16.11.0/24 is not advertized by EIGRP because 10.30.1.1 has no interface in this subnet.
I dont think each network whcih advertised under EIGRP needs to be a interface on the router.
Then how it is advertising through BGP even though the router has no interface.
I would agree with you "the route to 172.16.11.0/24 advertized by router 10.30.1.2 to both 10.30.1.1 and 10.30.1.3"
Actually the 10.30.1.2 router is not managed by us. it is different vendor and managed by third party.
But I am sure they have reverse route by permitting all 10.0.0.0 network.
Coming to "redistribute static" I did it at my DC on router 192.168.199.12
Do you want me to do the same at 10.30.1.1? Still i dont think it is a good idea.
Jopeti.
07-08-2011 12:21 PM
This is a fundamental difference between BGP and EIGRP.
BGP will advertize a network when it has a network statement and a matching route, regardless of its source.
The network statement in EIGRP has a different meaning. It is intended to activate EIGRP on interfaces within the range covered by the network statement. Example: network 10.0.0.0 under router eigrp 1 means: Run EIGRP AS1 on all interfaces in network 10.0.0.0/8.
The command reference explicitly states the networks specified must be directly connected.
http://www.cisco.com/en/US/docs/ios/12_0/np1/command/reference/1reigrp.html#wp1021711
For your setup this implies you can just as well omit the network statement for 172.16.11.0 on router 10.3.1.1 because it will not cause the network to be advertized in EIGRP.
I already expected the explanation about you being not in control of the router which actually sources this subnet.
If this cannot be negotiated with the customer, you can advertize the subnet by using redistribute static under eigrp configuration like I already told you. As long as the two CE routers remain eigrp neighbors the subnet will be advertized in the AS.
It should not be necessary to redistribute a static on another point in the network i.e. on router 192.168.199.12 and in fact this static route may even lead to sending the traffic into a black hole.
regards,
Leo
07-10-2011 11:57 PM
Leo,
The reason why I did "redistribute static" on 192.168.199.12 is becuase this is hub router and other network like 172.16.11.0 is connected to it where I am not managing the rotuer.
Is it still recommended to do "redistribute static" on 10.30.1.1?
Jopeti.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide