IPv6 route list checking and repeating method

Abstract

The method comprises: a) according to statistical distribution, the IPv6 routing table is divided into two parts based on the length of prefix; b) said two parts are respectively saved in the RAM and TCAM; c) in IP route forwarding process, based on IP address, respectively looking up RAM and TCAM, and according to the rule of longest match, making forwarding. The Invention uses hardware and CPU to commonly manage routing table so as to reduce the load of CPU; and meanwhile, TCAM uses self-adaptive allocation algorithm so as to improve the utilization factor of TCAM.

Description

technical field [0001] The invention relates to the field of IP routing and forwarding of network equipment, in particular to a method for searching and forwarding an IPv6 routing table. Background technique [0002] With the rapid development of the Internet, IPv4 is facing more and more problems, the most prominent of which is the crisis of address space exhaustion. In addition, because the IPv4 address scheme cannot well support address aggregation, the existing Internet is facing the pressure of expanding routing tables; IPv4 also has its natural shortcomings in terms of service quality, mobility, and security. [0003] In order to solve the inherent problems of IPv4, IETF has proposed IPv6 Internet protocol with 128-bit addresses. Compared with IPv4, IPv6 has at least the following advantages: extended address space and CIDR-like address structure; support for automatic configuration; better mobility and security. At present, the main protocols of IPv6 have been perfe...

AI Technical Summary

Problems solved by technology

However, these two technologies cannot be simply applied to IPv6

First of all, the length of IPv6 address is 128bit. If RAM technology is used, the choice of step width and tree depth will be more difficult. If the step width is too large, more RAM will be occupied. If the tree depth is too large, RAM will be accessed when forwarding The number of times will increase, which will affect the forwarding performance; secondly, compared with RAM, TCAM per bit of TCAM is more expensive, and the storage capacity of TCAM chip is smaller

Therefore, for a large-capacity IPv6 routing table, it is completely costly to use TCAM

Again, for the management of the TCAM table, if the TCAM is fixedly allocated to routes with certain prefix lengths in advance, the utilization rate of the TCAM is not high

Moreover, the current algorithm usually requires the CPU to maintain the corresponding shadow table in memory, which increases memory overhead and CPU burden

Images