Qos implementation method and system based on clustering design

A technology for implementing a system and an implementation method, which is applied in the field of congestion management of network switching chips, can solve the problems of complex linked list maintenance, insufficient control of different data flows, and limited network service quality, and achieves high network service quality and precise control of data flow. , the effect of rapid scheduling

Active Publication Date: 2020-02-18
武汉二进制半导体有限公司
View PDF6 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the above model is that the scheduling model between queues at the same layer is relatively fixed, the maintenance of the linked list is relatively complicated, the control of different data flows is not fine enough, and the quality of network service is limited.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Qos implementation method and system based on clustering design
  • Qos implementation method and system based on clustering design
  • Qos implementation method and system based on clustering design

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0059] see image 3 As shown, Embodiment 1 of the present invention provides a Qos implementation method based on clustering design, including the following steps:

[0060] S1. Generate multiple cluster layers, each cluster layer includes multiple ordered clusters, and each cluster in the upper cluster layer serves as a node, mapping multiple different continuous clusters in the lower cluster layer.

[0061] Taking four layers as an example, four cluster layers are generated, which are named L0 layer, L1 layer, L2 layer and L3 layer accordingly from bottom to top. There is 1 cluster in the L3 layer, and 1 cluster in the L3 layer is used as a node to map all the continuous nodes in the L2 layer. There are a total of E clusters in the L2 layer, and each cluster in the L2 layer is used as a node to map D consecutive clusters in the L2 layer. There are D*E clusters in the L2 layer; each cluster in the L2 layer is used as a node to map consecutive C clusters in the L1 layer, and t...

Embodiment 2

[0072] see figure 2 As shown, on the basis of Embodiment 1, the queues to be scheduled in the existing quality of service / layered quality of service Qos / Hqos structural framework are allocated to the underlying clusters in the manner that multiple queues correspond to one cluster layer, needs to confirm the relationship of the queue, and allocate according to the relationship of the queue, which includes:

[0073] According to the corresponding relationship of the first layer of queues in the existing Qos / Hqos structural framework, confirm the mapping relationship between the required corresponding clusters.

[0074] For example: there are 60 queues, and it is necessary to select from 60 queues in one scheduling; there are 3 cluster layers of bottom layer, middle layer, and top layer, and there are 12, 3, and 1 clusters from the lower layer to the upper layer.

[0075] Then its possible clustering methods are:

[0076] Correspond the 60 queues to the bottom cluster in the w...

Embodiment 3

[0098] On the basis of Embodiment 1, the priority level of the allocated cluster, whether it is qualified, the scheduling mode of the cluster, and the state in the scheduling mode are stored. The scheduling mode includes DWRR / RR scheduling and SP scheduling.

[0099] After the queue is allocated, before scheduling according to the priority level, it is preferred to store the priority level of the cluster allocated to the queue, whether it has qualifications, the scheduling method of the cluster, and the state in the scheduling mode. The priority level of the queue, so that the scheduling of the queue in the cluster layer corresponds to the scheduling of the cluster.

[0100] It should be noted that the priority level is the priority level of the cluster corresponding to its assigned queue, which depends on the cluster corresponding to all queues, and the priority of the queue is selected in the next scheduling.

[0101] For example, a L0-level cluster contains 16 queues, and a...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a Qos (Quality of Service) implementation method based on clustering design, which relates to the field of congestion management of network switching chips, and comprises the following steps: generating a plurality of cluster layers, wherein each cluster layer comprises a plurality of ordered clusters, each cluster in the previous cluster layer is used as a node, and a plurality of different continuous clusters in the next cluster layer are mapped; and allocating a to-be-scheduled layer of queue in the existing Qos/Hqos structure framework to a bottom cluster layer according to a mode that a plurality of queues correspond to one cluster, and modifying whether the cluster has qualification and priority according to whether the to-be-scheduled queue has qualificationand priority; scheduling the queue according to the priority: from the bottom layer to the top layer, performing dynamic weighted cycle (DWRR)/cycle (RR) scheduling on a plurality of clusters corresponding to the same node, and performing strict priority (SP) scheduling on the clusters with different priorities corresponding to the same node. According to the Qos implementation method based on clustering design, complex scheduling linked list management is avoided, the data flow can be controlled more finely, and the network service quality is higher.

Description

technical field [0001] The invention relates to the field of congestion management of network switching chips, in particular to a method for realizing Qos based on clustering design. Background technique [0002] With the ever-changing network and user requirements, application types and capacity, higher requirements are put forward for switching networks, so Qos (Quality of Service) should be born in response to the requirements. The QoS model of differential service is mainly adopted in the switching network. The QoS model mainly includes the following components: flow classification and marking, flow policing and shaping, congestion management and congestion avoidance. [0003] When data enters the network, according to the configured Qos mechanism, the data in the network is divided into different data flows and set to different priorities. This process is called flow classification and marking; For traffic management, various business traffic is limited to a specific ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): H04L12/801H04L12/851H04L12/863H04L12/869
CPCH04L47/24H04L47/10H04L47/2441H04L47/50H04L47/60H04L47/6215H04L47/6295
Inventor 李红军匡可
Owner 武汉二进制半导体有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap