Configuration and non-equivalent load balancing experiments of EIGRP protocol

eigrp protocol: From the perspective of packet encapsulation, EIGRP is encapsulated in the IP header. The protocol is a three-layer protocol and is a CISCO private protocol that supports load balancing.
The eigrp protocol has three tables:
Neighbor Table: each other The neighbor table is created by hello.
Topology table: All the information notified by the neighbors are placed in this table. Algorithm: 107/minimum bandwidth (k) + total delay (10us a unit) 256=metric
Several parameters that affect the metric: bandwidth B, delay D, reliability A, load L, maximum MTU
In realizing load balancing, when the sub-optimal AD is greater than the optimal FD and the load balancing cannot be done, the bandwidth value is usually changed to achieve this.
Load balancing:
1, equal cost load balancing
2, non-equivalent load balancing
Only eigrp can do non-equivalent load balancing, which means that there are multiple paths to the same destination. When their metric values ​​are different, you can do non-equivalent load balancing. You can perform non-equivalent load balancing by setting the variance value, that is, divide two numbers and round up and add 1. Lines that meet this condition can be non-equivalent Data packets are sent in a cost-effective load balancing manner. The maximum support for non-equivalent cost is 16.
AD: advertised distance (the cost between the next hop node and the target network)
FD: feasible distance (calculated from the local router) Play the cost between the target networks)
The AD of the sub-optimal path must be less than the FD of the optimal path in order to do load balancing.
V value
V*Optimal path FD>Sub-optimal path FD
That is the FD of the sub-optimal path/optimal path FD rounded +1=V value
The network diagram is as follows :
IP address configuration: R1 13.1.1.1 12.1.1.1 11.1.1.1
SW1 11.1.1.2
R2 12.1.1.2 24.1.1.2
R3 13.1.1.3 34.1.1.3
R4 24.1 .1.4 34.1.1.4 44.1.1.1
SW2 44.1.1.2
The whole network realizes eigrp
eigrp protocol configuration and non-equivalent load balancing experiment
R1>enable
configure terminal
interface fa 1/0
no shutdown
ip address 11.1.1.1 255.255.255.0
interface se 2/0
no shutdown
ip address 12.1.1.1 255.255.255.0
interface fa 0/0
no shutdown
ip address 13.1.1.1 255.255.255.0
exit
router eigrp 100 Run the eigrp autonomous system number AS 100. This must be the same for everyone. Communication is not possible if you are not in the same AS.
no auto-summary Turn off auto-summary, which can prevent loops in the same network segment from forming loops and a large number of route entries added by auto-summary.
network 11.1.1.0 0.0.0.255 Use the mask form to declare the interface. If you do not enter the reverse mask, the main network number will be used by default (the main network number will cause too much range).
network 12.1.1.0 0.0.0.255
network 13.1.1.0 0.0.0.255
eigrp protocol configuration and non-equivalent load balancing experiment
Configure switch SW1 to test the network, configure a VLAN IP.
SW1>enable
configure terminal
interface fa 0/1
no shutdown
switchport mode trunk
exit
interface vlan 1
ip address 11.1.1.2 255.255. 255.0
no shutdown
ip default-gateway 11.1.1.1
eigrp protocol configuration and non-equivalent load balancing experiment
Configure router R2
R2>enable
configure terminal
interface se 2/0
no shutdown
ip address 12.1.1.2 255.255.255.0
interface se 3/0
no shutdown
ip address 24.1.1.2
exit
router eigrp 100
no auto-summary
network 12.1.1.0 0.0.0.255
network 24.1.1.0 0.0.0.255
eigrp protocol configuration and non-equivalent load balancing experiment
Other R3, R4 and SW2 configuration methods are the same.
Now look at the 3 tables
show ip eigrp topology
show ip eigrp neighbor
show ip route
eigrp protocol configuration and non-equivalent load balancing experiment

Now ping is connected to the whole network.
Look at the routing situation and topology table from R2 to R1 11.1.1.0 again. It turns out that the road to 24.1.1.0 is unworkable. Because the priority level is not as good as 12.1.1.0 this road
. Now change the port bandwidth of both R2 and R4 to 100000, and then set the V value to 38, and then look at the routing table and find that there are 2 roads.
via 24.1.1.4 (545280/33280), Serial3/0
via 12.1.1.1 (20514560/28160), Serial2/0
v value=20514560/545280 +1 =38 round to zero< br>Note that the AD of the sub-optimal path must be less than the FD of the optimal path for load balancing.
That is, the value of 28160 must be less than the value of 545280.
eigrp protocol configuration and non-equivalent load balancing experiment
Eigrp protocol configuration and non-equivalent load balancing experiment
eigrp protocol configuration and non- Equivalent load balancing experiment

Leave a Comment

Your email address will not be published.