Recursive routing of Advanced Static Routing (III)

Recursive routing of Advanced Static Routing (III)
Recursive query of routes is unnecessary for static routes. The purpose of recursive routing is to bypass frequently changing links and reduce the configuration workload of Static Routing. I. Experiment topology II. Experiment description 1. If R1 wants to communicate with R4, R1 must have a static route to R4. If it is a simple topology, several static routes can be used, if there are many or frequent changes to the link to R4, the administrator needs to write a large number of static routes to ensure communication. In this case, recursive routing can be used to reduce the workload. 2. Recursive routing is implemented by directly associating Static Routing with the IP address of the inbound interface of the target CIDR Block router. 3. Implement asymmetric routing. The icmp echo package of R1 and the icmp reply package of R4 use different paths. 3. experiment step 1. Configure static routes on R1 and R4. 2. When R1 receives data from 4.4.4.0/24, It queries the route table and sends the data from f0/0, but only knows the next hop address 1.1.4, then R1 sends an ARP broadcast to f0/0 to query the MAC address. What if R2 and R3 receive ARP broadcasts? By default, both R2 and R3 enable the proxy ARP. If the local ARP table contains the ARP record of this address, the ARP-replay packet is sent. But on R1, it was found that R1 did not record the mac address of 1.1.4 to the ARP table, but it was discarded. Why? Because R1 will first determine whether the network segment of the responder is reachable (Note: The PC receives proxy-ARP reply unconditionally and adds it to the ARP table ). Another problem is that both R2 and R3 receive ARP broadcasts. Will R1 choose R2 or R3? 3. configure a static route to route 1.1.0/24 on R1, and configure a static route to 10.1.1.0/24 on R4. Ping 4.4.4.4 again, and find that the arp table entry has a record of listen 1.1.4, and that it is the f0/0mac address of R2. After R1 sends an ARP broadcast, both R1 and R2 send an arp-reply packet. When the packet arrives, it writes its mac address to the record table. Therefore, the address of the next hop node to be queried in a recursive table does not affect the path selection, but proxy arp affects the path selection. 4. How to Implement asymmetric routing? By default, arp table items are stored for 4 hours, but after 4 hours, arp-echo packets are resold. In this case, when 4.4.4.0/24 is used, R2 or R3 is random. In this case, we may modify the arp expiration time and disable the proxy ARP function of the interface to implement asymmetric routing. R2 (config-if) # no ip proxy-arp R3 (config-if) # no ip proxy-arp5, although the proxy arp is disabled, when the data arrives at the specified vro and does not go to the route of the destination CIDR block, you need to add a static route in R2 and R3. The result is verified by capturing the packets on the link. R2 (config) # ip route 1.1.1.0 255.255.0 f0/0 10.1.1.1 R3 (config) # ip route 4.4.4.0 255.255.255.0 f0/1 255.1.1.4 R2 to R1 the packet capture results are all reply packages, all the links from R3 to R4 are request packets. Asymmetric routing is successfully implemented.