Research and Implementation of RIP Routing Protocol in the IPv6 age

With the rapid development of the Internet, the network scale expands rapidly, the amount of information is increasing rapidly, and new applications are emerging one after another, the original Network Interconnection protocol version 4 that has been developed for more than 20 years -- IPv4 protocol has insufficient address space due to its inherent defects, low routing efficiency, poor security, low service quality, and lack valid mobile support ), it cannot fundamentally adapt to the needs of network development. In this context, the next generation network standard IPv6 Protocol emerged. This article discusses the RIP of Ripng for IPv6 networks. As the next-generation Internet IP layer technology has become a rule, IPv6 is one of the leading topics in the field of information technology to thoroughly study the simplicity and ease of use protocol Ripng, it has important economic and social significance.

I. basic working principle of RIP

The routing information protocol RIP (RoutingInformationProtocol) is a routing protocol specially designed by the Internal Gateway Protocol Workgroup for the Internet Engineering Task Group IETF, it is an internal gateway dynamic routing protocol based on distance vector algorithm.

Each router running RIP maintains a RIP route table. The content of this route table is 1.

The next hop nexthop indicates the address to which the next packet will arrive. The metric) indicates the cost required to deliver the packet from the current router to the destination station ). The maximum valid measurement supported by the RIP Protocol is 15. When the measurement of a route reaches 16, the route is considered invalid and the corresponding destination is considered inaccessible.

The flag indicates whether the route has changed recently for use when an update is triggered. The age is actually a timer used to maintain each route. If a route in RIP is still not refreshed after 180 seconds, the route is considered no longer valid and its measurement is set to 16.

The RIP Router periodically sends a copy of its route table to the neighbor in multicast mode, that is <目的，度量> Group. Each router that receives the message modifies the measurement of the route in the message, and adds the cost of receiving the router message interface to the measurement of each route. Then, judge the route quality based on the measurement size, and put the route with the smallest measurement into the route table. The judgment process is as follows:

1) check whether there are any routes for this purpose in the routing table; 2) Add the route if no route is found; 3) If yes, the route is updated only when the new metric is updated, otherwise, ignore the route.

Figure 2 shows the RIP workflow. As a result, we can clearly feel that RIP has such a simple workflow.

2. When a RIP is started on an interface of the routerrouter, the interface sends a route table information request to the neighbor in the form of multicast, requesting the neighbor to send the route table information to himself; the neighbor RouterB receives the route table information request and sends the entire route table information to respond to the request. After RouterB is started, the router is sent and updated periodically. When RouterB detects route changes, send a trigger update to the neighbor in multicast mode to notify the neighbor of route changes.

Ii. Evolution of RIP in the IPv6 age

As we all know, today's RIP has evolved from RIP-1 to RIP-2, and now there is a revolutionary IPv6-based Ripng! Studying the evolution of RIP and analyzing the similarities and differences of RIP in various stages of development is of great significance for optimizing and developing the protocol.

RIP-1 only supports subnets in a network with subnets. It is difficult to know the subnet mask outside the network, and thus the subnet and host items cannot be distinguished, in this way, you need to use a strictly hierarchical route: The External router transmits the group to the nearest vro in the network, regardless of which subnet the destination site belongs. Followed by a strict hierarchy of routes, this is the requirement for connectivity, that is, every vro in this network should know how to go to any subnet!

Based on the preceding disadvantages, RIP-2 defines an effective RIP improvement solution, and redefines some "must be zero" fields contained in the RIP-1 format.

A new AFI Item 0 xFFFF is added based on security considerations to transmit "verification data". "Route Selection domain" and "Next Hop domain" are added ", allows multiple RIP instances to run on a single vro to specify the destination IP address to which the packets are sent, the "Next Hop domain" attribute allows only one vro running other protocols to advertise all the routes, in addition, the router running RIP can find the router that runs the Protocol correctly as the next hop to achieve the destination address of all known protocols. The "subnet mask" is added ", to enhance the performance of subnet selection routing, and add a "routing label" to mark external routes.

RIP-2 retains domain names such as the command, address family identifier, IP address, and measurement defined in the RFC-1058.

No, it's better! With the continuous improvement of RIP-1 and RIP-2 based on IPv4, Although IPv4 has supported a mature Internet architecture for 20 years, it has long been a consensus, that is, IPv6 eventually replaces IPv4, which is the trend of the times: the number of IP addresses supported by the 128-bit address format can completely solve the problem of address depletion; IPv4 address space is scarce, routing efficiency is poor, and security is poor, poor mobility and poor service quality assurance. All these IPv4 problems will be solved without hesitation by the IPv6 protocol, so the Ripngv6 version will soon become a favorite.

Although Ripng belongs to the RIP family, Ripng has undergone a revolution. Different from RIP-1 and RIP-2 in various aspects: different ports are used; different grouping formats; different maximum packet lengths; different next hops; Different addressing; different special requests; security considerations are different.

Ripng and RIP-2 both have the next hop, but the next hop of RIP-2 is fixed in the package format, that is to say, no matter whether the next hop is added, it will leave 4 bytes of space. Ripng separates the next hop nexthop from the route table items and uses the value in the "measurement domain" to determine whether it is the next hop or the route table item. If the value is 0xff, it indicates the next hop! This method can be omitted without the need to add the next hop information, thus saving valuable space.

There is also a major change in Ripng and RIP face change! That is, in the RIP-2 packet format, there is a "subnet mask domain", and in Ripng, the "subnet mask domain" function is replaced by a byte prefix length, although this has an inseparable relationship with the characteristics of IPv6, the rationality, beauty, and simplification of its layout can also give us priority to use it.

Iii. Ripng implementation

Ripng is implemented in six functional modules: Ripng startup, Ripng shutdown, input processing, output processing, Timer processing, and routing operations, as shown in 3.

Ripng startup module: initializes the Ripng protocol and establishes a communication set interface. Ripng shutdown module: notifies the kernel to delete the routes imported by the RIPng process and releases the memory space occupied by the Ripng process; input Processing Module: Processing Ripng as a black box. The packets sent by neighbors are the information sent to the Black Box. This module processes the information accordingly. Output Processing Module: the output processing mainly includes periodic update and trigger update processing. The timer processing module is responsible for maintaining the three timers used in RIP to support seek operations, at the same time, it is also responsible for processing the timing mechanism introduced in trigger updates to prevent broadcast storms. The routing operation module is responsible for processing route entries and searching, adding, and deleting route tables.

Ripng is the first choice for solving future IPv6 network routing. Its simple internal gateway protocol has powerful advantages and unlimited potential. An excellent routing technology not only finds a high-speed channel for data transmission, but also needs to consider the transmission capacity and service quality of the selected path, that is, a routing algorithm with QoS capabilities, we also need to analyze the network-wide load to balance the data traffic of each channel in the network. Based on these factors, the study of the Ripng routing protocol based on IPv6 is the hot spot and focus of our future research.