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

Tail delay prediction method for online cloud service based on stochastic reward network

A prediction method and cloud service technology, applied in the transmission system, electrical components, etc., can solve the problems of queuing network without structural description, Markov chain state space explosion, etc., and achieve the effect of dynamic allocation of resources

Active Publication Date: 2022-07-22
TIANJIN UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The queuing network does not have a unified structure to describe the concurrency phenomenon, and the Markov chain has a state space explosion

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
  • Tail delay prediction method for online cloud service based on stochastic reward network
  • Tail delay prediction method for online cloud service based on stochastic reward network
  • Tail delay prediction method for online cloud service based on stochastic reward network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] In order to make the model more suitable for practical applications, the present invention proposes a random return network model based on a request processing process, and provides a method for predicting the response time distribution of an open-loop network. The invention is mainly composed of two modules: model construction and response time cumulative distribution calculation.

[0041] 1. Model building blocks

[0042] The model building module obtains a random return network model based on the open-loop queuing network by analyzing the processing process of the user request through the server. Specific steps are as follows:

[0043] 1) Determine the number of layers n of the model, which is the total number of different server types whose requests are processed by the server;

[0044] 2) The request processing process of each type of server, that is, one layer can be represented as a concurrent queue. Each tier can have one or more servers (assuming a tier has ...

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 relates to the field of online service performance evaluation, in order to predict the distribution and tail delay of the online cloud service request response time for system performance prediction, dynamic capacity configuration, and bottleneck identification. To this end, the technical solution adopted in the present invention is that the online cloud service tail delay prediction method based on random return network includes two parts: model construction and response time cumulative distribution calculation. Random return network model based on open-loop queuing network; cumulative distribution of response time is obtained by simulating the state of the system when any request arrives and the absorbing state of the request to find the response time of the request to obtain the cumulative distribution of the request response time. The present invention is mainly applied to online service performance evaluation.

Description

technical field [0001] The invention relates to the field of online service performance evaluation, in particular, to a method for predicting tail delay of online cloud services based on a random return network. Background technique [0002] The cloud computing market has grown rapidly in recent years. Numerous enterprise applications, such as social networking, e-commerce, video streaming, email, web search, MapReduce, Spark, etc., are migrating to cloud systems. In traditional environments, to ensure the quality of user experience, applications are deployed on an isolated server cluster in the data center. In order to improve resource utilization and save costs of cloud platforms, it is a new trend to deploy different applications on the same physical server. However, there will be performance disturbances when applications are deployed on the same server, resulting in unpredictable performance and high tail latency, seriously affecting user experience quality and even v...

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): H04L43/0852H04L43/045H04L41/14H04L41/142H04L67/10
CPCH04L41/142H04L41/145H04L43/045H04L43/0852H04L67/10
Inventor 张蓉琦赵来平周晓波蔡斌雷李克秋
Owner TIANJIN UNIV