Vro monitoring is also very important in routing applications. At the same time, through the detection of vrouters, it can detect many problems in a timely manner. With the rapid development of computer networks, the network size is getting bigger and bigger, and the network is becoming more and more complex. Traditional network management methods cannot meet new requirements.
Therefore, the low efficiency of traditional network management has become an urgent problem. The core of the network is routing. By analyzing the routing, you can obtain a large amount of information about the network. Therefore, vro monitoring has become an important research area for network operation and management. Capture the route information of the entire network through a router monitoring agent. The dynamic nature of the entire network can be analyzed through routing information. This article mainly studies the dynamic topology reconstruction and network stability of the border gateway protocol BGP (border gateway protocol. Network topology changes intuitively reflect the dynamic nature of BGP. By changing the network topology, you can clearly obtain the fault information, so that you can promptly troubleshoot the fault. Through vro technology, real-time and dynamic network topology construction algorithms are implemented and displayed in a graphical manner.
BGP Inter-Domain router monitoring
From the management perspective, the Internet is composed of different Autonomous systems (. Autonomous systems refer to relatively independent networks with independent routing policies in an autonomous system. These AS may belong to different Internet Service providers (ISPs), and different ISPs may restructure into new ISPs to form a dynamic management organization structure. In AS, you can use the InteriorGateway Protocols (IGP) to manage scheduling routes, such as ospf, IS-IS, and RIP; exchange routing information between AS through the public Network Access Point (NAP) and the Border Gateway Protocol. The most obvious difference between BGP and IGP is that BGP allows each AS to select, advertise, and accept routes based on its own routing policies, providing a flexible routing mechanism. However, to implement this flexible routing mechanism, the BGP protocol configuration is also very flexible to adapt to a variety of complex tasks, but this flexibility is also easy to cause errors in the configuration. Some small local error configurations may cause network interruptions or even affect the entire Internet connection. Therefore, it is necessary to monitor and verify the AS question routing policy.
BGP-4 is a protocol that integrates Distance Vector Algorithm and link state algorithm. By analyzing route interaction information, especially the AS Path attribute in the Update frame, the AS sequence from the observation point to the target network can be obtained. Integrate these sequences to obtain the topology of the entire network. In addition, the BGP-4 is a dynamic routing protocol that, in addition to interacting with all route tables in the initial environment, the BGP-4 sends route change information incrementally. This information is sent to the adjacent BGP-4 Router through the Update frame. Through dynamic information analysis, You can dynamically analyze the entire BGP network in real time.
To study the dynamic behavior of BGP networks, we must first monitor BGP routes. The traditional route table download-based Router monitoring increases the load on routers and networks and has great defects. To solve the problem of router monitoring, a set of Inter-Domain router monitoring methods based on Route interaction are proposed. Based on this method, the BGP routing information is collected on the premise that the network and router load is extremely small, this information is used to analyze the dynamic behavior of BGP. A vro monitoring proxy is established through Zebra to connect to a VBR In the BGP network. In this way, the vro monitoring proxy can receive interactive routing packets, so as to monitor and analyze the dynamics of the BGP network. The BGP network topology is generated to monitor the overall observation of the BGP network and the routing changes between autonomous systems.
Introduction to Zebra
Zebra is a TCP/IP routing software that supports BGP-4, BGP-4 +, OSPFv2, OSPFv3, r1_1, r1_2, and RIPng. It complies with the gnu gpl standard and runs on Linux and other Unix variant systems. The Zebra kit was developed by Kunihiro Ishiguro and Yoshinari Yoshikawa in 1996. Currently, the Zebra suite is maintained by IP Infusion. Zebra uses the module method to manage various protocols. The protocols can be activated or stopped based on network requirements. Zebra has the following features:
(1) The modular Zebra can be easily upgraded and maintained using multiple threads. The upgrade of a protocol module does not affect the operation of other protocol modules.
(2) high-speed Zebra has a higher packet forwarding speed than traditional routing software.
(3) Reliability Zebra itself has a high reliability. Failure of a module does not affect the execution of other protocol modules.
Zebra System Structure: traditional routing software provides all router monitoring functions in a single thread mode. Zebra is completely different. It uses a group of daemon to jointly maintain the route table. For example, the ripd daemon processes the RIP Protocol, ospfd is used to support the ospf daemon, bgpd is used to support the BGP-4. At the same time, to manage the kernel route table, use the daemon zebra to manage the kernel route table.
This multi-process method provides scalability, modularity, and maintainability. It also provides multiple configuration files and interfaces. Each daemon has its own configuration file and terminal interface. When configuring a module, you must configure it in the Zebra configuration file. When configuring the BPG network, you need to configure the BGP module file, which brings problems to your use. To solve this problem, Zebra provides an integrated User Interface script called vryshell. The kernel mode of Zebra adopts multithreading.