Implementation method for improving virtual machine network performance in openstack dvr mode

Abstract

The invention provides an implementation method for improving virtual machine network performance in an openstack dvr mode, which belongs to the technical field of cloud computing and computer networks and comprises two-layer traffic flooding repair on a br-int bridge and three-layer traffic flooding repair on the br-int bridge. The two-layer traffic flooding repair is responsible for repairing the problem of traffic flooding between cross-computing node virtual machines of the same subnet in the dvr mode; and the three-layer traffic flooding repair is responsible for repairing the traffic flooding problem of different subnet cross-computing node virtual machines in the dvr mode. According to the method, invalid traffic can be effectively avoided, and the virtual machine performance is greatly improved.

Description

technical field [0001] The invention relates to cloud computing and computer network technology, in particular to a method for realizing network performance improvement of a virtual machine in an openstack dvr mode. Background technique [0002] In the current dvr mode of openstack, when the virtual machines on the same subnet across computing nodes communicate with each other, the traffic sent from the virtual machine on the current node to the virtual machine on the remote node will be flooded on the local br-int bridge. The flooding of Layer 2 traffic on the br-int bridge seriously consumes system resources and greatly weakens the communication efficiency between virtual machines, so that when the number of virtual machines increases, the PPS performance between virtual machines on the same subnet Significant reduction: when different subnets communicate between virtual machines across computing nodes, the traffic will first be sent to the router on the node where the sen...

AI Technical Summary

Problems solved by technology

[0002] In the current dvr mode of openstack, when the virtual machines on the same subnet across computing nodes communicate with each other, the traffic sent from the virtual machine on the current node to the virtual machine on the remote node will be flooded on the local br-int bridge. The flooding of Layer 2 traffic on the br-int bridge seriously consumes system resources and greatly weakens the communication efficiency between virtual machines, so that when the number of virtual machines increases, the PPS performance between virtual machines on the same subnet Significant reduction: when different subnets communicate between virtual machines across computing nodes, the traffic will first be sent to the router on the node where the sender is located, and when it is sent from the gateway in the router to the br-int bridge, There is also the problem of Layer 2 communication, traffic will still be flooded on the br-int bridge, which will also consume system resources and reduce performance