Unlock instant, AI-driven research and patent intelligence for your innovation.

A Congestion Sensing and Marking Method for Microbursts in Multi-queue Data Center Environment

A technology of data center and marking method, applied in the field of computer network, can solve the problems of inability to perceive micro-burst flow, inability to adapt to dynamic network, etc., achieve the effect of high efficiency and easy deployment, reducing the generation of ECN false marking, and high throughput

Active Publication Date: 2021-07-23
SOUTHEAST UNIV
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The present invention is aimed at the problem that the traditional ECN cannot sense the micro-burst flow and cannot adapt to the dynamic network. It proposes a congestion sensing and marking method for the micro-burst flow in a multi-queue data center environment. Due to the change of the queue length of the switch, there are Regular oscillation, and the dynamic uncertain network transmission process can be modeled as a stable and controllable transmission process through steady-state analysis. This method intends to use steady-state analysis to analyze how to set the effective ECN threshold K so that the switch can Adapt to the emergence of micro-burst flows, avoid the generation of a large number of error congestion information, and improve network throughput

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
  • A Congestion Sensing and Marking Method for Microbursts in Multi-queue Data Center Environment
  • A Congestion Sensing and Marking Method for Microbursts in Multi-queue Data Center Environment
  • A Congestion Sensing and Marking Method for Microbursts in Multi-queue Data Center Environment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0027] Embodiment 1: as figure 1 As shown, a method of congestion perception and marking for micro-burst flows in a multi-queue data center environment, the method includes the following steps:

[0028]Step 1: Taking the data center as the background, model the generation and transmission process of micro-bursts in a multi-queue switch environment, use the steady-state analysis method to theoretically analyze the maximum value of micro-bursts, and derive ECN The lower bound of the threshold to ensure that the queue buffer can absorb the arrival of microbursts;

[0029] Step 2: In order to ensure that the exit rate of the switch queue is consistent with the entry rate in a multi-queue environment, and reduce network congestion caused by rate mismatch, based on the GPS model and the lower bound of the ECN threshold obtained in step 1 as the standard, solve the problem Find out the ECN threshold base value that is adapted to the multi-queue environment and can absorb the arrival...

specific Embodiment approach

[0049] Step 1: For the calculation of the lower bound of the ECN threshold based on the microburst behavior model, the present invention is based on Figure 4 For the topological structure shown, a microburst behavior model is established, and the model is derived through steady-state analysis. The specific implementation methods are as follows:

[0050] 11) Obtain the number N of network flows in the switch;

[0051] 12) For different end device congestion control algorithms, calculate the lower bound of the ECN threshold:

[0052] 1) When the end device deploys the DCTCP congestion control algorithm, the lower bound of the ECN threshold of each queue j, It is calculated as follows:

[0053] ·like

[0054]

[0055] ·like

[0056]

[0057] 2) When the end device deploys the ECN* congestion control algorithm, the lower bound of the ECN threshold of each queue j It is calculated as follows:

[0058] Step 2: Calculate for the ECN threshold base value, the spec...

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 marking method based on display congestion notification (ECN) in a multi-queue data center environment, which mainly solves the problems that the traditional ECN cannot perceive micro-burst flows and cannot adapt to dynamic networks. The present invention mainly includes three parts, which are the calculation of the lower bound of the ECN threshold, the setting of the base value of the threshold and the dynamic adjustment of the threshold. The present invention first conducts steady-state analysis on the generation and transmission process of micro-bursts in a multi-queue data center environment, and solves the lower bound of the ECN threshold for each switching queue to ensure that the queue buffer of the switch can effectively absorb the micro-bursts. Then based on the ideal Generic Packet Schedule (GPS) model, adjust the lower bound of the ECN threshold and set the threshold base value for each queue to ensure the balance between the entry rate and exit rate of the exchange queue. Finally, based on the goal of improving network throughput In this paper, an optimization model for maximizing the utilization of queue buffers is established for switch memory resources, and an algorithm for dynamically adjusting ECN thresholds is proposed.

Description

technical field [0001] The invention relates to a method, in particular to a congestion sensing and marking method for micro-burst flows in a multi-queue data center environment, and belongs to the technical field of computer networks. Background technique [0002] In recent years, the rapid development of cloud computing technology has attracted widespread attention from industry and academia, and Internet giants have launched cloud computing services. As the core infrastructure on which large-scale cloud computing services rely, data centers provide massive computing and storage resources while carrying large-scale network traffic generated and formed by cloud computing services. In particular, with the rapid expansion of data center business scale, data center network (Data Center Networking, DCN) traffic has changed from traditional "north-south traffic" to "east-west traffic" inside the data center. The competition of network bandwidth resources is inevitable, which ma...

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 Patents(China)
IPC IPC(8): H04L12/801
CPCH04L47/12H04L47/127H04L47/29
Inventor 张竞慧康克熙金嘉晖沈典罗军舟
Owner SOUTHEAST UNIV