Method and equipment for dynamic allocation of resources

Abstract

The embodiments of the invention disclose a method for dynamic allocation of resources. The method is applied to a distributed processing system, wherein the distributed processing system comprises at least two service CPUs (Central Processing Units). The method comprises the following steps that: a first service CPU acquires the resource usage information of a user on each service CPU; the first service CPU adjusts the resource allocation quantities allocated to the user by all the service CPUs according to the total resource quantity applied by the user and the resource usage information of the user on each service CPU, wherein the sum of the resource allocation quantities allocated to the user by all the service CPUs is not greater than the total resource quantity applied by the user; and the first service CPU instructs the corresponding service CPUs to allocate the resources to the user according to the adjusted resource allocation quantities. According to the method, the resource allocation for all users on all the service CPUs can be adjusted in real time, and then, the resource allocation situation can be dynamically adjusted according to the service conditions of all the service CPUs, so that the circumstances such as packet loss and the like caused by unreasonable resource allocation can be reduced.

Description

technical field [0001] The invention belongs to the communication field, and in particular relates to a resource dynamic allocation method and equipment. Background technique [0002] With the popularization of Internet video applications and cloud service applications, network data traffic continues to increase, and the requirements for data forwarding and service processing speed of network devices are getting higher and higher. High-speed network devices (routers, firewalls, etc.) on the Internet often integrate a large number of service CPUs internally, and provide high data throughput externally through load sharing and concurrent data processing among internal multi-service CPUs. [0003] In order to use the network bandwidth reasonably, the data flow of each user needs to be limited on the network device. Implementing user traffic limitation on high-speed network devices requires multi-service CPUs to cooperate to control the total incoming and outgoing data traffic ...

AI Technical Summary

Problems solved by technology

Using the method of the prior art, it often happens that one service CPU loses packets because the bandwidth value is used up, and the bandwidth value on another service CPU is idle, and the bandwidth value of a specified user cannot be dynamically shared between service CPUs. So that some CPUs start to lose packets when the total bandwidth value of the user is not used up

Images