时间：2006年2月12日
时间花费 总计约90分钟
实验设备：2501 5台
实验拓扑：
需求：在NBMA环境中配置EIGRP，研究如何使spoke之间可以互相更新路由。
参考资料：http://www.cisco.com/go/eigrp/
Deploying IGRP/EIGRP
http://www.cisco.com/application/pdf/en/us/guest/products/ps6630/c1161/cdccont_0900aecd80310f03.pdf
在完成配置之后（此时只配置了最基本的eigrp）可以看到下面的情况
R1#sh ip route （由于R1和R4是spoke由于水平分割的问题无法从R5处收到对方的路由，但是能收到R5本身通告的路由）
…
1.0.0.0/32 is subnetted, 1 subnets
C 1.1.1.1 is directly connected, Loopback0
5.0.0.0/32 is subnetted, 1 subnets
D 5.5.5.5 [90/2297856] via 10.1.1.5, 00:00:19, Serial1
10.0.0.0/24 is subnetted, 1 subnets
C 10.1.1.0 is directly connected, Serial1
…
R5#sh ip route
1.0.0.0/32 is subnetted, 1 subnets
D 1.1.1.1 [90/2297856] via 10.1.1.1, 00:00:12, Serial0
4.0.0.0/32 is subnetted, 1 subnets
D 4.4.4.4 [90/2297856] via 10.1.1.4, 00:00:12, Serial0
5.0.0.0/32 is subnetted, 1 subnets
C 5.5.5.5 is directly connected, Loopback0
10.0.0.0/24 is subnetted, 1 subnets
C 10.1.1.0 is directly connected, Serial0
R4#sh ip route
4.0.0.0/32 is subnetted, 1 subnets
C 4.4.4.4 is directly connected, Loopback0
5.0.0.0/32 is subnetted, 1 subnets
D 5.5.5.5 [90/2297856] via 10.1.1.5, 00:01:04, Serial0
10.0.0.0/24 is subnetted, 1 subnets
C 10.1.1.0 is directly connected, Serial0
R4#sh ip eigrp topology（在R4上可以看到eigrp拓扑中也没有R1的loopback口）
IP-EIGRP Topology Table for AS(1)/ID(4.4.4.4)
Codes: P – Passive, A – Active, U – Update, Q – Query, R – Reply,
r – reply Status, s – sia Status
P 5.5.5.5/32, 1 successors, FD is 2297856
via 10.1.1.5 (2297856/128256), Serial0
P 4.4.4.4/32, 1 successors, FD is 128256
via Connected, Loopback0
P 10.1.1.0/24, 1 successors, FD is 2169856
via Connected, Serial0
R4#ping 10.1.1.1 但是R4上可以ping到R1的s1接口的接口地址
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.1.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 132/134/144 ms
R4#ping 10.1.1.5当然也可以ping到R5的
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.1.1.5, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 68/69/72 ms
R5#sh ip eig nei此时可以在R5上看到R1和R4的接口地址是R5的neighbor，并且由于是NBMA的低速链路，hold timer默认情况下是180秒
IP-EIGRP neighbors for process 1
H Address Interface Hold Uptime SRTT RTO Q Seq
(sec) (ms) Cnt Num
1 10.1.1.4 Se0 170 00:03:56 60 360 0 4
0 10.1.1.1 Se0 156 00:04:05 60 360 0 4
如果在R5的s0口写上ip split-horizon eigrp as号，此时已经建立好的EIGRP邻接关系会中断，但是等待几秒后会重新建立好：
在R1(或R4)上打开debug eigrp packet和debug ip routing可以看到这个过程：
*Mar 1 00:48:35.223: %DUAL-5-NBRCHANGE: IP-EIGRP(0) 1: Neighbor 10.1.1.5 (Serial1) is up: new adjacency
*Mar 1 00:49:01.711: %DUAL-5-NBRCHANGE: IP-EIGRP(0) 1: Neighbor 10.1.1.5 (Serial1) is down: Interface Goodbye received
*Mar 1 00:49:59.255: %DUAL-5-NBRCHANGE: IP-EIGRP(0) 1: Neighbor 10.1.1.5 (Serial1) is up: new adjacency
*Mar 1 00:56:01.903: %DUAL-5-NBRCHANGE: IP-EIGRP(0) 1: Neighbor 10.1.1.5 (Serial1) is down: Interface Goodbye received 可以看出是R5主动要求中断邻接关系的
*Mar 1 00:56:01.907: RT: delete route to 5.5.5.5 via 10.1.1.5, eigrp metric [90/2297856]
*Mar 1 00:56:01.911: RT: no routes to 5.5.5.5
*Mar 1 00:56:01.911: RT: NET-RED 5.5.5.5/32
*Mar 1 00:56:01.915: RT: delete subnet route to 5.5.5.5/32
*Mar 1 00:56:01.919: RT: NET-RED 5.5.5.5/32
*Mar 1 00:56:01.923: RT: delete network route to 5.0.0.0
*Mar 1 00:56:01.923: RT: NET-RED 5.0.0.0/8 由于邻接中断，因此要删除相应学到的路由
*Mar 1 00:56:52.971: %DUAL-5-NBRCHANGE: IP-EIGRP(0) 1: Neighbor 10.1.1.5 (Serial1) is up: new adjacency 重新建立邻接
*Mar 1 00:56:56.335: RT: add 5.5.5.5/32 via 10.1.1.5, eigrp metric [90/2297856]
*Mar 1 00:56:56.339: RT: NET-RED 5.5.5.5/32
*Mar 1 00:56:56.407: RT: add 4.4.4.4/32 via 10.1.1.5, eigrp metric [90/2809856]
*Mar 1 00:56:56.411: RT: NET-RED 4.4.4.4/32 又把路由学到了，此时可以看到R4的路由通过R5更新过来了
R1#sh ip eigrp top
IP-EIGRP Topology Table for AS(1)/ID(1.1.1.1)
Codes: P – Passive, A – Active, U – Update, Q – Query, R – Reply,
r – reply Status, s – sia Status
P 4.4.4.4/32, 1 successors, FD is 2809856
via 10.1.1.5 (2809856/2297856), Serial1
P 5.5.5.5/32, 1 successors, FD is 2297856
via 10.1.1.5 (2297856/128256), Serial1
P 1.1.1.1/32, 1 successors, FD is 128256
via Connected, Loopback0
P 10.1.1.0/24, 1 successors, FD is 2169856
via Connected, Serial1
1.0.0.0/32 is subnetted, 1 subnets
C 1.1.1.1 is directly connected, Loopback0
4.0.0.0/32 is subnetted, 1 subnets
D 4.4.4.4 [90/2809856] via 10.1.1.5, 00:00:59, Serial1 送到R4的数据要通过R5转送
5.0.0.0/32 is subnetted, 1 subnets
D 5.5.5.5 [90/2297856] via 10.1.1.5, 00:00:59, Serial1
10.0.0.0/24 is subnetted, 1 subnets
C 10.1.1.0 is directly connected, Serial1
如果在R5上不指定no ip split-horizon eigrp as号的话，那么还可以用下面方法解决hub-spoke方式连接的问题，在各个spoke上手动互相指定neighbor，即形成逻辑上的全互连拓扑(full mesh)。此时路由器会采用单播(unicast)方式进行连接，从而建立邻接并传递路由。但是这种方法的扩展性不好，当spoke新增的时候要修改所有的spok
e的配置。
这样指定配置的时候可以得到下面的结果：
R1#sh ip route
…
1.0.0.0/8 is variably subnetted, 2 subnets, 2 masks
C 1.1.1.1/32 is directly connected, Loopback0
D 1.0.0.0/8 is a summary, 00:01:19, Null0
4.0.0.0/32 is subnetted, 1 subnets
D 4.4.4.4 [90/2297856] via 10.1.1.4, 00:00:13, Serial1此时送到R4的数据不用通过R5
5.0.0.0/32 is subnetted, 1 subnets
D 5.5.5.5 [90/2297856] via 10.1.1.5, 00:00:03, Serial1
10.0.0.0/24 is subnetted, 1 subnets
C 10.1.1.0 is directly connected, Serial1
…
R1#sh ip eigrp topology 看EIGRP拓扑可以得到相同的结论
P 5.5.5.5/32, 1 successors, FD is 2297856
via 10.1.1.5 (2297856/128256), Serial1
P 4.4.4.4/32, 1 successors, FD is 2297856
via 10.1.1.4 (2297856/128256), Serial1
P 1.1.1.1/32, 1 successors, FD is 128256
via Connected, Loopback0
P 1.0.0.0/8, 1 successors, FD is 128256
via Summary (128256/0), Null0
P 10.1.1.0/24, 1 successors, FD is 2169856
via Connected, Serial1