Comparison of common routing protocols

RIP (Routing Information Protocols) Routing Information Protocol OSPF (Open Shortest Path First) Open Path First (Enhanced Interior Gateway Routing Protocol) Enhanced Internal Gateway Routing Protocol Static Routing: static Routing is only applicable to small networks or small to medium-sized networks with only a small scale. Manual Input and manual management are required. Management overhead is a huge burden for dynamic routing. Advantages: Excellent bandwidth and good security. Dynamic Routing Protocol: vrouters in a network communicate with each other, transmit route information, and update and maintain route tables based on a specific routing protocol. Type: distance vector routing protocol and link status routing protocol. Features: reduces management tasks and occupies network bandwidth RIP: RIP is the most widely used distance vector routing protocol. RIP is designed for a small network environment, because the routing learning and route update of such protocols will generate a large amount of traffic, occupying excessive bandwidth. To avoid routing loops, RIP uses five mechanisms: horizontal segmentation, toxicity reversal, defining the maximum number of hops, transient update, and restraining timing. Horizontal segmentation is a rule used to prevent the generation of routing loops. The rule here refers to the routing information learned from an interface and will not be sent from this interface. The RIP Protocol is divided into version 1 and version 2. Both version 1 and Version 2 have the following features: 1. is the distance vector routing protocol; 2. use Hop Count as the measurement value. 3. the default route update cycle is 30 seconds. 4. the Management Distance (AD) is 120; 5. update triggering is supported; 6. the maximum number of hops is 15. supports equivalent paths. The default value is 4, and the maximum value is 6. 8. use the UDP520 port for Route updates. The differences between r1_1 and r1_2 are as follows: differences between r12001 and r00002 r00001 r00002 does not carry subnet information during route update, but does not provide authentication. plaintext and MD5 authentication do not support VLSM and CIDR support VLSM. and CIDR adopt broadcast ratio 255.255.255) the update uses multicast (224.0.0.9) to update Classful routing protocol Classless routing protocol. After a series of Route updates, each vro in the network has a complete route table, convergence. AS an internal Gateway Protocol (IGP), OSPF is used to publish route information between routers in the same autonomous domain (. Unlike the distance vector protocol (RIP), OSPF supports large networks, fast route convergence, and less network resources. It plays a very important role in the current application of routing protocols. OSPF version 2 is widely used. The latest standard is RFC2328. Based on the physical network connected by the router, OSPF divides the network into four types: Broadcast MultiAccess) non-Broadcast MultiAccess, NBMA, Point-to-Point, and Point-to-MultiPoint ). Broadcast Multi-Channel Access networks such as Ethernet, Token Ring, and FDDI. NBMA networks such as Frame Relay, X.25, and SMDS. Point-to-Point networks such as PPP and HDLC. The specific structure is shown in the following figure. Comparison between OSPF and RIP: Limitations of r00001 arising from the use of large networks: 1. 15 hop limitations of RIP, routes with more than 15 hops are considered inaccessible. 2. RIP cannot support a variable-length Subnet Mask (VLSM), resulting in inefficient IP Address Allocation. 3. periodically broadcast the entire route table, applications on low-speed links and wide area networks will cause major problems. 4. convergence speed is slower than OSPF. Convergence time in large networks takes several minutes. 5. RIP does not have the concept of network latency and link overhead, route selection is based on the number of hops. A route with a small number of hops is always selected as the best route, even if a long route has a low latency and overhead. 6. RIP has no region concept, route aggregation cannot be performed on any bit. Some enhanced functions are introduced into the new version r12002 of RIP. r00002 supports VLSM, authentication, and multicast update. However, the limit on the number of hops and slow convergence of r12002 make it not suitable for large networks. OSPF is more suitable for large networks than RIP: 1. There is no limit on the number of hops. 2. Support for variable-length Subnet Mask (VLSM). 3. Send link status updates using multicast. Trigger updates when link status changes, improves bandwidth utilization. 4. fast convergence. 5. uses the authentication function. The VPN protocol is a Cisco private route protocol. It combines the advantages of Distance Vector and link status :: fast Convergence, reduced bandwidth usage, MD5 authentication, route aggregation, load balancing, and simple configuration. VPN does not have a region. However, in the case of a large-scale network, OSPF can plan and limit the network size by dividing regions. Therefore, it is suitable for networks with relatively small network sizes. This is also the limitation of the vector-distance routing algorithm (the RIP Protocol uses this algorithm. Send HELLO messages at regular intervals. The routers running the network must send the HELLO message regularly to maintain the relationship between the neighbors. This relationship also needs to send the HELLO Message periodically even on the dial-up network, in this way, on the on-demand dial-up network, it is impossible to locate whether this is a useful service message or a scheduled inquiry message sent by the kernel, which may trigger a connection on the On-Demand dial-up network by mistake, especially on the backup network, causing unnecessary troubles. Therefore, generally, for a router running the VPN, you must configure the Dialer list and Dialer group on the dial-up backup port to filter unnecessary packets or run the TRIP protocol, in this way, the router operation overhead is increased. OSPF supports on-demand dial-up on the dial-up network. Only one routing protocol can meet the needs of various leased lines or dial-up network applications. Based on Distributed DUAL algorithms. The loop-free computing and convergence speed of the network are based on the distributed DUAL algorithm, which actually spreads uncertain routing information (send query Packets to neighbors ), the process of re-convergence after the confirmation of all neighbors is obtained. When the neighbors are not sure about the reliability of the route information, the spread will be repeated, therefore, in some cases, the route information may remain active (this route is called an active route stack). If the active route is in the DUAL computing process, when the measurement of the successor (successor) of the route changes, it enters multiple computations, which will affect the DUAL algorithm's convergence speed. The OSPF algorithm does not have this problem, so from the perspective of convergence speed, although the whole is similar, but in some special circumstances, there is still an unsatisfactory situation in the OSPF algorithm. This article is from Mark Sky"