Routing management method and device

Abstract

The present invention provides a routing management method and device. The method comprises: performing static route iteration according to an added dynamic host route in a routing table; determiningwhether a dynamic host route iterated by a static route derives from an ARP table of a virtual machine in an ARP table or not; if a determination result is yes, setting the static route correspondingto the virtual machine after iteration as valid, and sending the static route to a network; or else, setting the static route corresponding to the virtual machine after iteration as invalid. The routing management method and device can reduce load of a leaf device and improve detection and switching speed of virtual machine faults.

Description

technical field [0001] The present application relates to communication technology, in particular to a routing management method and device. Background technique [0002] At present, the network of the data center can adopt a spine-leaf two-layer structure (also known as a leaf-spine topology network), where the leaf can be a node connected to network devices such as servers or hosts (referred to as a leaf node), and the spine can be connected to a network switch. node (referred to as the ridge node). In addition, the network device connected to the leaf can be a virtual machine implemented using the virtualization technology NFV (Network Function Virtualization, Network Function Virtualization), so that the functions of the network device no longer depend on dedicated hardware, and resources can be shared flexibly. In order to ensure the reliability of the business, you can set the backup structure of the primary virtual machine and the backup virtual machine. [0003] Un...

AI Technical Summary

Problems solved by technology

As can be seen from the above methods, leaf needs to run link detection protocols such as the above-mentioned BFD to detect whether the connected virtual machine is faulty, and then decide whether to publish static routes, but this will make the leaf burden larger, especially when there are many leaf connections The burden is heavier when there are a large number of virtual machines, which affects the efficiency of fault detection and switching

Images