BGP Redistributed into EIGRP - Manually Altering Metrics

[Gregor Blaj]

Hi,

I've set up a simulation to do some testing with BGP/EIGRP, see attached. In this network BGP and EIGRP are configured and BGP routes are redistributed into EIGRP.

What I want to know is how I can alter the routing configuration so the preferred path from host C2 over the WAN is always,

C2 > R4 > R5 > R3 > R1

Unless the R4 > R5 link becomes unavailable, in which case the path should be,

C2 > R2 > R1

I believe this is achievable with the metrics set on the the two routers which are redistributing BGP into EIGRP (R2 and R3) but I could be wrong?

R2: redistribute bgp 65100 metric 100 1 255 1 1500
R3: redistribute bgp 65100 metric 100 1 255 1 1500

Any help appreciated.

[ghostinthenet]

It's morning and I haven't had my coffee yet, so take this off-the-top-of-my-head thinking in that context, but have you tried increasing your redistribution bandwidth on R3? redistribute bgp 65100 metric 1000 1 255 1 1500 on R3 should cause those redistributed routes to be preferred over the ones from R2.

[Gregor Blaj]

Hi both,

It seems that whatever I change the redistribution metrics to on R3, both R4 and R5 always seem to take the R3 route (I know this is what I want, but I want to control it :) ). If I disable redistribution on R3 then the routes from R2 are learned and the same thing happens with this router.

Example for R4. Default setting on both R3 and R2, R3 route is accepted since it was configured first (I'm assuming).

R3(config-router)#redistribute bgp 65100 metric 100 1 255 1 1500

R4#sh ip ro eigrp
D EX    10.44.20.0 [170/25651456] via 192.168.1.2, 00:01:11, FastEthernet0/1

Then I disabled redistribution on R3, so it's only enabled on R2 with values 'metric 100 1 255 1 1500'. The route R4 takes from R2 now is better than the original route given by R3, but still the route from R3 was accepted until redistribution was disabled on R3.

R3(config-router)#no redistribute bgp 65100 metric 1000 1 255 1 1500

R4#sh ip ro eigrp
D EX    10.44.20.0 [170/25625856] via 10.44.0.254, 00:00:02, FastEthernet0/0

Also, if I execute 'R4#show ip eigrp topology 10.44.20.0 255.255.255.0', shouldn't I see routes from both R2 and R3 listed? I only see one.

192.168.1.0/30 > Network between R4 and R5

10.44.0.254 > R2

10.44.20.0/24 > LAN on R1

Thanks again.

[e.ciollaro]

May you upload router's config, I'd like to test on Gns3.

[Gregor Blaj]

Sure, all 5 router config's attached. I am doing this in GNS3 too :)

Let me know how you get on.

Cheers.

[Gregor Blaj]

I just started GNS3 up again and both routes showed up, then I changed one of the 'redistribution statements' and the route updated as expected. When I changed the redistribution statement on the 2nd router to make that route even better is when I started seeing problem.

It looks like after the second redistribution statement is changed EIGRP stops updating. Might be a GNS3 fault, wonder if I can test this in Packet Tracer.

[e.ciollaro · Accepted solution]

Ok, the problem is that you are using iBGP between R1 and R2.

Here the test I made:

I booted the network and checked available path on R4 and there are two path (preferred R2, as expected):

R4#sh ip ei topology all-links
IP-EIGRP Topology Table for AS(1)/ID(10.44.0.253)

Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply,
       r - reply Status, s - sia Status

P 10.44.0.0/24, 1 successors, FD is 281600, serno 1
        via Connected, FastEthernet0/0
P 10.44.20.0/24, 1 successors, FD is 25625856, serno 3
        via 10.44.0.254 (25625856/25600256), FastEthernet0/0
        via 192.168.1.2 (25651456/25625856), FastEthernet0/1
P 192.168.1.0/30, 1 successors, FD is 281600, serno 2
        via Connected, FastEthernet0/1
P 172.19.98.0/24, 1 successors, FD is 307200, serno 5
        via 192.168.1.2 (307200/281600), FastEthernet0/1
        via 10.44.0.254 (25625856/25600256), FastEthernet0/0

Then I changed the redistribution metric on R2:

R2#sh run | s eigrp
router eigrp 1
 redistribute bgp 65100 metric 10 100000 1 255 1500
 network 10.44.0.0 0.0.0.255
 no auto-summary

R4 suddenly use path through R5:

R4#sh ip rou 10.44.20.0
Routing entry for 10.44.20.0/24
  Known via "eigrp 1", distance 170, metric 25651456, type external
  Redistributing via eigrp 1
  Last update from 192.168.1.2 on FastEthernet0/1, 00:02:20 ago
  Routing Descriptor Blocks:
  * 192.168.1.2, from 192.168.1.2, 00:02:20 ago, via FastEthernet0/1
      Route metric is 25651456, traffic share count is 1
      Total delay is 2010 microseconds, minimum bandwidth is 100 Kbit
      Reliability 255/255, minimum MTU 1500 bytes
      Loading 1/255, Hops 2

R4#sh cdp neighbors
Capability Codes: R - Router, T - Trans Bridge, B - Source Route Bridge
                  S - Switch, H - Host, I - IGMP, r - Repeater

Device ID        Local Intrfce     Holdtme    Capability  Platform  Port ID
R2               Fas 0/0            137         R S I     3725      Fas 0/0
R5               Fas 0/1            139         R S I     3725      Fas 0/1

R4 has no path through R2 because R2 learn R1 network from R4. R2 receives advertisement for this network from two different protocol:

iBGP (from R1) with administrative distance 200

EIGRP (from R4) with administrative distance 170

Due to better administrative distance, R2 use R4 path and install that path in routing table. That's way R2 can't advertise that path back to R4. If you shut R3 interf,  R4 will suddenly use R2 path:

R4#sh ip rou 10.44.20.0
Routing entry for 10.44.20.0/24
  Known via "eigrp 1", distance 170, metric 281625600, type external
  Redistributing via eigrp 1
  Last update from 10.44.0.254 on FastEthernet0/0, 00:00:37 ago
  Routing Descriptor Blocks:
  * 10.44.0.254, from 10.44.0.254, 00:00:37 ago, via FastEthernet0/0
      Route metric is 281625600, traffic share count is 1
      Total delay is 1001000 microseconds, minimum bandwidth is 10 Kbit
      Reliability 1/255, minimum MTU 1500 bytes
      Loading 255/255, Hops 1

If if unshut R3's interf and  make iBGP administrative distance better then EIGRP administrative distance on R2 and R3, then both routers (R2 and R3) install in routing table the BGP route, redistribute it in EIGRP and advertise it to R4 and R5:

R2#sh  run | s bgp
 redistribute bgp 65100 metric 10 100000 1 255 1500
router bgp 65100
 no synchronization
 bgp log-neighbor-changes
 bgp redistribute-internal
 network 10.44.0.0 mask 255.255.255.0
 neighbor 172.20.20.1 remote-as 65100
 neighbor 172.20.20.3 remote-as 65100
 distance bgp 20 160 160
 no auto-summary

R2#sh ip route
Codes: C - connected, S - static, R - RIP, M - mobile, B - BGP
       D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
       N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
       E1 - OSPF external type 1, E2 - OSPF external type 2
       i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
       ia - IS-IS inter area, * - candidate default, U - per-user static route
       o - ODR, P - periodic downloaded static route

Gateway of last resort is not set

     172.19.0.0/24 is subnetted, 1 subnets
D       172.19.98.0 [90/332800] via 10.44.0.253, 00:16:50, FastEthernet0/0
     172.20.0.0/24 is subnetted, 1 subnets
C       172.20.20.0 is directly connected, FastEthernet0/1
     10.0.0.0/24 is subnetted, 2 subnets
C       10.44.0.0 is directly connected, FastEthernet0/0
B       10.44.20.0 [200/0] via 172.20.20.1, 00:02:31
     192.168.1.0/30 is subnetted, 1 subnets
D       192.168.1.0 [90/307200] via 10.44.0.253, 00:21:26, FastEthernet0/0
R2#

R3#sh run | s router bgp
router bgp 65100
 no synchronization
 bgp log-neighbor-changes
 bgp redistribute-internal
 network 172.19.98.0 mask 255.255.255.0
 neighbor 172.20.20.1 remote-as 65100
 neighbor 172.20.20.2 remote-as 65100
 distance bgp 20 160 160
 no auto-summary
R3#sh ip route
Codes: C - connected, S - static, R - RIP, M - mobile, B - BGP
       D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
       N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
       E1 - OSPF external type 1, E2 - OSPF external type 2
       i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
       ia - IS-IS inter area, * - candidate default, U - per-user static route
       o - ODR, P - periodic downloaded static route

Gateway of last resort is not set

     172.19.0.0/24 is subnetted, 1 subnets
C       172.19.98.0 is directly connected, FastEthernet0/0
     172.20.0.0/24 is subnetted, 1 subnets
C       172.20.20.0 is directly connected, FastEthernet0/1
     10.0.0.0/24 is subnetted, 2 subnets
D       10.44.0.0 [90/332800] via 172.19.98.253, 00:09:21, FastEthernet0/0
B       10.44.20.0 [160/0] via 172.20.20.1, 00:00:44
     192.168.1.0/30 is subnetted, 1 subnets
D       192.168.1.0 [90/307200] via 172.19.98.253, 00:09:21, FastEthernet0/0
R3#

Now R4 has R5 as next-hop:

R4#sh ip route
Codes: C - connected, S - static, R - RIP, M - mobile, B - BGP
       D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
       N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
       E1 - OSPF external type 1, E2 - OSPF external type 2
       i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
       ia - IS-IS inter area, * - candidate default, U - per-user static route
       o - ODR, P - periodic downloaded static route

Gateway of last resort is not set

     172.19.0.0/24 is subnetted, 1 subnets
D       172.19.98.0 [90/307200] via 192.168.1.2, 00:26:44, FastEthernet0/1
     10.0.0.0/24 is subnetted, 2 subnets
C       10.44.0.0 is directly connected, FastEthernet0/0
D EX    10.44.20.0 [170/25651456] via 192.168.1.2, 00:02:05, FastEthernet0/1
     192.168.1.0/30 is subnetted, 1 subnets
C       192.168.1.0 is directly connected, FastEthernet0/1

but is has also the path through R2:

R4#sh ip ei topology all-links
IP-EIGRP Topology Table for AS(1)/ID(10.44.0.253)

Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply,
       r - reply Status, s - sia Status

P 10.44.0.0/24, 1 successors, FD is 281600, serno 1
        via Connected, FastEthernet0/0
P 10.44.20.0/24, 1 successors, FD is 25651456, serno 12
        via 192.168.1.2 (25651456/25625856), FastEthernet0/1
        via 10.44.0.254 (281625600/281600000), FastEthernet0/0
P 192.168.1.0/30, 1 successors, FD is 281600, serno 2
        via Connected, FastEthernet0/1
P 172.19.98.0/24, 1 successors, FD is 307200, serno 5
        via 192.168.1.2 (307200/281600), FastEthernet0/1
        via 10.44.0.254 (25625856/25600256), FastEthernet0/0

Note that R2 is not a feasible successor (see metrics in blu in the previous show) so if you use sh EIGRP topology (without "all-link"), only path through R5 is showed:

R4#sh ip ei topology
IP-EIGRP Topology Table for AS(1)/ID(10.44.0.253)

Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply,
       r - reply Status, s - sia Status

P 10.44.0.0/24, 1 successors, FD is 281600
        via Connected, FastEthernet0/0
P 10.44.20.0/24, 1 successors, FD is 25651456
        via 192.168.1.2 (25651456/25625856), FastEthernet0/1
P 192.168.1.0/30, 1 successors, FD is 281600
        via Connected, FastEthernet0/1
P 172.19.98.0/24, 1 successors, FD is 307200
        via 192.168.1.2 (307200/281600), FastEthernet0/1

Let me know if everything is clear.

Bye,

e

PS please rate if useful

[Gregor Blaj]

Hi again,

I see, that makes sense. Suppose another way to get it to work is to change the external EIGRP route distance on R2 and R3?

R3#sh run | s eigrp
router eigrp 1
 redistribute bgp 65100 metric 100 1 255 1 1500
 network 172.19.98.0 0.0.0.255
 distance eigrp 90 210
 no auto-summary

One question though. In the original configuration why does R2 only get the EIGRP route in it's table after the redistribution metric is changed on R3? When the redistribution metric is the same on R2 and R3, both just have the BGP route to 10.44.20.0/24.

[e.ciollaro]

"Suppose another way to get it to work is to change the external EIGRP route distance on R2 and R3?"

Yes

 

"In the original configuration why does R2 only get the EIGRP route in it's table after the redistribution metric is changed on R3? When the redistribution metric is the same on R2 and R3, both just have the BGP route to 10.44.20.0/24."

In original configuration R2 has BGP route and inject EIGRP router in routing table just after changing the administrative distance not the redistribution metric. As far as EIGRP has a better administrative distance R2 will inject BGP router because that's the way IOS choose the subnet to inject: first based on administrative distance and, in  case of more path with the same admin distance, based on metric.

May be I didn't  understand, in that case please let me kno how you change R3 config after booting with initial config

Bye,

enrico

PS please rate if useful

[Gregor Blaj]

That makes sense! Thanks for all the help.

[Gregor Blaj]

Also, figured out another problem :)

If the preferred route for R4 to R1's LAN is via R3, then the packet flow would be.

R4 > R3 > R1 > R2 > R4

Since R1 only knows about R4's LAN from R2.

[cuharper]

Nice.........from your post the following worked for me....not exactly sure why......but now i can see D EX on the next router......

router eigrp 1
  redistribute bgp 65100 metric 10 100000 1 255 1500

router bgp 65100

  distance bgp 20 160 160

...................thanks

[e.ciollaro]

Supposing that R2 has a eBGP neighborship with R1, R2 has to path:

• one eBGP path with administrative distance 20 via R1
• one EIGRP path with administrative distance 170 (external) via R4

R4 has two path: both EIGRP with admin distance 170. If the default gateway is R2 you need to change the administrative distance so that eBGP is worst then EIGRP, no need to change the path cost. 

If R4 is the default gateway you can change the path cost so that R4-R5 is better than R4 - R2. This can be done both modifying the redistribute metric (R3) or f0/0 interf delay on R4's (EIGRP use delay in its metric).

Bye,

enrico

PS please grade if useful