CN116155835B - Cloud resource service quality assessment method and system based on queuing theory - Google Patents

Cloud resource service quality assessment method and system based on queuing theory Download PDF

Info

Publication number
CN116155835B
CN116155835B CN202310071364.8A CN202310071364A CN116155835B CN 116155835 B CN116155835 B CN 116155835B CN 202310071364 A CN202310071364 A CN 202310071364A CN 116155835 B CN116155835 B CN 116155835B
Authority
CN
China
Prior art keywords
service
node
quality
nodes
capacity
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202310071364.8A
Other languages
Chinese (zh)
Other versions
CN116155835A (en
Inventor
孟凡超
孟凡浩
朴学峰
初佃辉
卢阳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Harbin Institute of Technology Weihai
Original Assignee
Harbin Institute of Technology Weihai
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Harbin Institute of Technology Weihai filed Critical Harbin Institute of Technology Weihai
Priority to CN202310071364.8A priority Critical patent/CN116155835B/en
Publication of CN116155835A publication Critical patent/CN116155835A/en
Application granted granted Critical
Publication of CN116155835B publication Critical patent/CN116155835B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/70—Admission control; Resource allocation
    • H04L47/80—Actions related to the user profile or the type of traffic
    • H04L47/805—QOS or priority aware
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides a cloud resource service quality assessment method and a system based on queuing theory, which are used for determining the processing capacity and capacity of a service node relative to service; classifying the service nodes into basic service nodes and combined service nodes according to the service quantity contained in the service nodes; and aiming at different service nodes, adopting a corresponding queuing model to evaluate the service quality, and outputting an evaluation result. According to the cloud resource management method and the cloud resource management system, essential characteristics of cloud resource consumption can be described on a service level, and the service quality requirement of the terminal user is mapped into the consumption of the cloud resource, so that the renting cost of the cloud resource is reduced for a service provider on the premise that the service quality requirement of the user can be met.

Description

Cloud resource service quality assessment method and system based on queuing theory
Technical Field
The invention belongs to the field of cloud resource service quality evaluation, and particularly relates to a queuing theory-based cloud resource service quality evaluation method and a queuing theory-based cloud resource service quality evaluation system.
Background
With the rapid development of the internet and modern service industry, service systems have been widely deployed and applied in government, business, and educational fields. In order to ensure the performance of the system, the service provider needs to continuously increase the investment of the IT infrastructure, however, the problem of high operation cost, low resource utilization rate, difficult maintenance and the like exists in the expansion of the local data center of the enterprise. Report statistics show that the resource utilization of most enterprise data centers is below 15%, and still less than 5%. Because the load fluctuation of the service system is large, a large amount of residual resources exist in the data center in an idle period, and huge waste of energy consumption is caused.
The service quality, such as response time, reliability, etc., is the most concerned for the end user of the service system, while the main objective of the service provider is to meet the service quality requirement of the user with the lowest cost of renting cloud resources, so how to map the service quality requirement of the end user to the consumption of cloud resources, i.e. to build a quantitative relation model between the service quality requirement and the cloud resource consumption, is the premise and foundation for realizing the optimal supply of cloud resources.
Due to the characteristics of isomerism, dispersity, complexity, diversity and the like of cloud resources in the public cloud environment, the difficulty and the high cost of establishing a quantitative relation model between the service quality requirement and the cloud resource consumption are high, so that how to effectively solve the 'gap' between the service quality requirement and the cloud resource consumption is the problem to be solved by optimizing and supplying the cloud resources in a large-scale service system in the current public cloud environment.
Disclosure of Invention
The invention provides a cloud resource service quality assessment method based on queuing theory, which can measure the service capacity of cloud resources, establish a quantitative relation model between service quality requirements and the service capacity and user load of the cloud resources, thereby providing a theoretical basis for cloud resource optimization supply of a large-scale service system and reducing the lease cost of the cloud resources for service providers.
The cloud resource service quality assessment method comprises the following steps:
s1: calculating the processing capacity and capacity of the service node relative to the service;
s2: dividing the service nodes into basic service nodes and combined service nodes according to the service quantity contained in the service nodes;
s3: aiming at different service nodes and service node clusters, different queuing models are adopted for service quality evaluation;
it should be further noted that, a service node refers to a virtual unit for deploying a service. Let s be a service, vu be a virtual unit, letAnd->(t=1, 2, …) is the time spent by the service s on the virtual unit vu and the consumed memory, which are random variables, respectively, and the time spent by the service node on each execution and the memory data can be obtained by adopting a performance test method, and the average value of the time spent by the service node on each execution and the memory data can be measured by adopting a mathematical statistics method.
The processing power refers to the average number of service nodes per unit time to fulfill user service requests. Set random variableMean>Wherein (1)>The average number of times the service s is executed on the virtual unit vu per unit time, and thus the processing power of the service s on the virtual unit vu can be defined as: />Wherein ε vu The value of the processing capacity coefficient is related to the virtualization technology adopted by the virtual unit and the CPU architecture of the physical server.
Capacity refers to the maximum number of concurrent service requests that a service node can accept. The capacity of a service s on a virtual unit vu can be defined as:where D is vu Maximum available memory for virtual unit vu, < ->Is a random variable +.>Average value of θ vu Is the capacity coefficient, which takes on the valueIn relation to the virtualization technology employed by the virtual units and the storage shelves.
It should be further noted that the method further includes: the method comprises the steps of constructing a mapping relation between service system parameters and queuing system parameters, wherein the specific mapping relation among parameters in a service system is as follows:
{ load → average customer arrival rate, service node processing capability → average attendant processing rate, service node capacity → capacity of system (queue length+number of attendant), number of concurrent service requests → average queue length, response time → average stay rate, throughput → absolute throughput capability, reliability → relative throughput capability, error rate → loss rate }.
It should be further noted that, the classification method of service nodes in step S2 includes dividing service nodes into basic service nodes and combined service nodes according to the number of services included in the service nodes, where one basic service node includes only one service and one combined service node includes multiple services.
It should be further noted that step S3 further includes: aiming at basic service nodes, a single service window queuing model M/M/1/c with capacity constraint is established to evaluate service quality;
also for the combined service node, a multi-class customer single service window queuing model M/H with capacity constraint is established k 1/c to evaluate the quality of service;
a queuing network model is also built for the service node cluster to evaluate the quality of service.
It should be further noted that the manner of evaluating the quality of service based on M/1/c includes:
configuration ofA basic service node is represented, wherein s is a service, vu is a virtual unit, and the processing capacity and capacity of the service s on the virtual unit vu are respectively: mu and c;
basic service nodeThe response time of the service s deployment on the virtual unit vu under load λ is:
basic service nodeI.e. the throughput of the service s deployed on the virtual unit vu under load λ is:
basic service nodeI.e. the error rate of the service s deployment on the virtual unit vu under load λ is:
basic service nodeThe reliability of the service s deployment on the virtual unit vu under load λ is:
it is further noted that M/H based k The manner of evaluating the quality of service/1/c includes:
configuration ofRepresents a combined service node, wherein s= { S 1 ,s 2 ,…,s k -service set, vu is a virtual unit;
service s i The processing capacity, capacity and load of the e S on the virtual unit vu are respectively: mu (mu) i ,c i And lambda (lambda) i The method comprises the steps of carrying out a first treatment on the surface of the Order theThe execution time of the virtual unit vu to all services obeys a k-order hyper-exponential distribution, and the distribution function is as follows:
capacity constrained multi-class customer single service window queuing model M/H k 1/c, which emphasizes on analyzing each type of customer performance index in the model;
model pair is distributed in load delta= { lambda% 1 ,λ 2 ,…,λ k Combined service node under }Average quality of service of (a) and each service s i Evaluating the service quality of the E S;
wherein v= { s 0 ,s 1 ,…,s n ,s n+1 Is a collection of services for which i ∈V，λ i Representation s i Is a load of (2);
and->Representing service nodes +.>Average response time, throughput rate, error rate and reliability of all services in (a)>And->Representing services s respectively i At the service node->Response time, throughput, error rate, and reliability.
It should be further noted that, the way to build the queuing network model to evaluate the service quality includes:
configuration ofRepresents a service node cluster, wherein s= { S 1 ,s 2 ,…,s k -a service set; VU= { VU 0 ,vu 1 ,…,vu m Is a set of virtual units, vu 0 Vu, a virtual unit of a control node j (j=1, …, m) is a virtual unit of a service node, m is the number of service nodes, and virtual units of different service nodes may be the same;
let the service set S be in the virtual unit vu j The load distribution on this is: delta j ＝{λ 1j ,λ 2j ,…,λ kj }，λ ij (i=1, 2,., k) is service s i In virtual unit vu j Load on, service clusterIs a load distribution of delta = { lambda 1 ,λ 2 ,…,λ k }，/>
It should be further noted that the queuing network model is divided into two layers: a control layer and a service layer;
the control layer is a description of the service request process inside the control node, and the control node is used for controlling the service request processEach service s in (a) i (i=1, 2, …, k) is modeled as a single service node +.>The M/M/1/c model is adopted to evaluate the service quality;
the processing layer comprises m service nodes, if k=1, each service node is a basic service node, otherwise, each service node is a combined service node;
for basic service nodes, the M/M/1/c queuing model is adopted to evaluate the service quality;
for the combined service node, the M/H is adopted k A/1/c queuing model to evaluate its quality of service;
if consider a control nodeAnd (3) quality of service: service s i (i=1, 2, …, k) at the service node +.>The response time is as follows:
service s i (i=1, 2, …, k) at the serving nodeThe throughput rate is as follows:
service s i (i=1, 2, …, k) at the serving nodeError rate onThe method comprises the following steps:
service s i (i=1, 2, …, k) at the serving nodeThe reliability is as follows:
wherein,and->Respectively represent control nodes->Average response time, throughput, error rate, and reliability of processing service requests;
the service s is a service without considering the quality of service of the control node i At the service node->Response time, throughput, error rate, and reliability.
The invention also provides a cloud resource service quality evaluation system based on queuing theory, which comprises: the system comprises an evaluation determination module, a node classification module and a service quality evaluation module;
the evaluation determination module is used for determining the processing capacity and the capacity of the service node relative to the service;
the node classification module is used for classifying the service nodes into basic service nodes and combined service nodes according to the service quantity contained in the service nodes;
the service quality evaluation module is used for evaluating the service quality by adopting a corresponding queuing model aiming at different service nodes and outputting an evaluation result.
From the above technical scheme, the invention has the following advantages:
the cloud resource service quality assessment method based on queuing theory provided by the invention can be used for describing essential characteristics of cloud resource consumption on a service level, and providing concepts of processing capacity and capacity so as to realize high-level abstraction of cloud resource consumption.
In order to describe the problem of cloud resource quality assessment by adopting the queuing system, the invention establishes the mapping relation between the parameters in the service system and the parameters in the queuing system. Aiming at the characteristic of service node diversity, the service nodes are divided into basic service nodes and combined service nodes. Then according to the different types and scales of the service nodes, the invention provides an M/M/1/c queuing model and an M/H k And three service quality evaluation models, namely a 1/c service quality evaluation model and a queuing network model, are respectively used for evaluating the quality evaluation of the basic service node, the combined service node and the service node cluster, can meet the cloud resource service quality evaluation under different conditions, and have practical values.
Drawings
In order to more clearly illustrate the technical solutions of the present invention, the drawings that are needed in the description will be briefly introduced below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for evaluating the quality of service of cloud resources based on queuing theory;
FIG. 2 is a mapping relationship between service system parameters and queuing system parameters of the present invention;
fig. 3 is a schematic diagram of a queuing network model for service node cluster quality of service assessment of the present invention.
Detailed Description
The cloud resource service quality assessment method based on queuing theory provided by the invention measures the service capacity of cloud resources. For the cloud resources or cloud services evaluated by the method, the cloud resources or cloud services are the general names of network technology, information technology, integration technology, management platform technology, application technology and the like based on cloud computing business model application, can form a resource pool, and are used as required, flexible and convenient. Among them, the background service of the cloud resource service requires a large amount of computing and storage resources, such as video websites, picture websites and more portals. Along with the high development and application of the internet industry, each article possibly has an own identification mark in the future, the identification mark needs to be transmitted to a background system for logic processing, data with different levels can be processed separately, and various industry data needs strong system rear shield support and can be realized only through cloud computing. The cloud resource service is mainly applied to medical cloud, cloud security, artificial intelligent service, cloud Internet of things, cloud games, cloud education, cloud conferences and other directions; among them, the artificial intelligence cloud Service is also commonly called AIaaS (AIas a Service, chinese is "AI as Service"). The service mode of the artificial intelligent platform is the mainstream at present, and particularly, the AIaaS platform can split several common AI services and provide independent or packaged services at the cloud. This service mode is similar to an AI theme mall: all developers can access one or more artificial intelligence services provided by the use platform through an API interface, and partial deep developers can also use an AI framework and AI infrastructure provided by the platform to deploy and operate and maintain self-proprietary cloud artificial intelligence services.
The invention relates to a cloud resource service quality assessment method based on queuing theory, which utilizes the characteristic of service node diversity to divide service nodes into basic service nodes and combined service nodes, and establishes M/M/1/c queuing model and M/H k Three service quality evaluation models of/1/c and queuing network model, and the average response time, throughput rate, error rate and reliability of service are utilized to serve as a large-scale service systemProvides a theoretical basis for cloud resource optimization provisioning, and reduces the lease cost of cloud resources for service providers. The 'gap' between the service quality requirement and the cloud resource consumption is further effectively solved, and the problem to be solved is that cloud resources are optimally supplied in a large-scale service system under the current public cloud environment.
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, a flowchart of a method for evaluating quality of service of cloud resources based on queuing theory in an embodiment is shown, where the method includes:
s1: determining processing capacity and capacity of the service node relative to the service;
the specific steps of step S1 are as follows:
s1.1: the method of performance test is adopted to obtain the time and memory data consumed by the service node for executing the service each time, and the average value of the time and memory data is measured by adopting the method of mathematical statistics.
Let s be a service, vu be a virtual unit, letAnd->(t=1, 2, …) is the time spent and the memory consumed by the service s for each execution on the virtual unit vu, respectively. Then, the performance test can be passed to obtain multiple groups of data and measure the average value of the data.
S1.2: the processing power of the service node relative to the service is calculated. And calculating the evaluation times of the service executed on the service node in unit time by collecting the performance test information, thereby calculating the processing capacity of the service node relative to the service. Processing power refers to unit timeThe inner service node completes the average number of user service requests. Set random variableMean of (2)Wherein (1)>The average number of times the service s is executed on the virtual unit vu per unit time, and thus the processing power of the service s on the virtual unit vu can be defined as: />Wherein ε vu To process the capacity coefficient, the coefficient may be estimated by analyzing the virtualization technique employed by the virtual unit and the test results of the physical server CPU architecture.
S1.3: the capacity of the service node relative to the service is calculated.
And performing performance test by setting different concurrent service request numbers, recording and analyzing the maximum concurrent service request number which can be accepted by the service node in the acceptable response time, so as to calculate the capacity of the service node relative to the service.
The number here refers to the maximum number of concurrent service requests that can be accepted by the service node. The capacity of a service s on a virtual unit vu can be defined as:where D is vu Maximum available memory for virtual unit vu, < ->Is a random variable +.>Average value of θ vu For capacity coefficients, their values and virtualization techniques employed by the virtual units and their storageStorage racks are defined by performance testing of virtualization technology and storage architecture.
S2: classifying the service nodes into basic service nodes and combined service nodes according to the service quantity contained in the service nodes;
wherein the number of services contained in the service node is determined. If the service node only contains one service, the service node is a basic service node. If the service node only contains a plurality of services, the service node is a combined service node. If a plurality of basic service nodes and combined service nodes work together, and the load distribution of each service node can be different, and a cluster is formed, the cluster is called a service node cluster.
S3: and aiming at different service nodes, adopting a corresponding queuing model to evaluate the service quality, and outputting an evaluation result.
In this embodiment, an applicable qos evaluation model is established according to the classification result of S2. For basic service nodes, a single service window queuing model M/M/1/c with capacity constraint is established to evaluate the service quality. For the combined service node, a multi-class customer single service window queuing model M/H with capacity constraint is established k 1/c to evaluate its quality of service; for a service node cluster, a queuing network model is built to evaluate its quality of service.
The method comprises the following steps of:
according to the processing capacity mu and capacity c of the service S on the virtual unit vu obtained in the step S2, respectively solving the basic service nodeResponse time, throughput, error rate, and reliability of (c).
The specific calculation formula is as follows: basic service nodeThe response time of the service s deployment on the virtual unit vu under load λ is:
basic service nodeI.e. the throughput of the service s deployed on the virtual unit vu under load λ is:
basic service nodeI.e. the error rate of the service s deployment on the virtual unit vu under load λ is:
basic service nodeThe reliability of the service s deployment on the virtual unit vu under load λ is:
M/H based k The 1/c queuing model construction steps are as follows:
according to the processing capacity, capacity and load of the service si e S on the virtual unit vu acquired in step S2, these are denoted as μi, ci and λi, respectively. Wherein,representing a combined service node, s= { S1, S2, …, sk } is a service set, and vu is a virtual unit. Let->The execution time of the virtual unit vu to all services obeys a k-order hyper-exponential distribution, and the distribution function is as follows:
each type of customer performance index in the model is then analyzed with emphasis. Model pair combined service node under load distribution delta= { λ1, λ2, …, λk })And the quality of service for each service si e S. Wherein v= { s 0 ,s 1 ,…,s n ,s n+1 Is a collection of services for which i ∈V，λ i Representation s i Is a load of (a).
And->Representing service nodes +.>Average response time, throughput rate, error rate and reliability of all services in (a)> And->Representing services s respectively i At the service node->Response time, throughput, error rate, and reliability.
As shown in fig. 2 and 3, the queuing-based network model construction steps are as follows:
step 3.3.1: queuing into a network model is divided into two layers: a control layer and a service layer.
Step 3.3.2: each service in the control node is modeled as a single service node, and the M/1/c model is used to evaluate its quality of service.
Step 3.3.3: the service nodes of the processing layer are evaluated.
Wherein,represents a service node cluster, wherein s= { S 1 ,s 2 ,…,s k -a service set; VU= { VU 0 ,vu 1 ,…,vu m Is a set of virtual units, vu 0 Vu, a virtual unit of a control node j (j=1, …, m) is a virtual unit of a service node, m is the number of service nodes, and virtual units of different service nodes may be the same. Let the service set S be in the virtual unit vu j The load distribution on this is: Δj= { λ1j, λ2j, …, λkj }, λij (i=1, 2,., k) is service S i In virtual unit vu j Load on, service cluster->The load profile of (a) = { λ1, λ2, …, λk }, +.>
The control layer is a description of the service request process inside the control node, and we will control the nodeEach service s in (a) i (i=1, 2, …, k) is modeled as a single service node +.>The M/M/1/c model is used to evaluate its quality of service.
The processing layer contains m service nodes, each of which is a basic service node if k=1, and otherwise a combined service node. The control layer is a description of the service request process inside the control node, and we will control the nodeEach service s in (a) i (i=1, 2, …, k) is modeled as a single service node +.>The M/M/1/c model is used to evaluate its quality of service. The processing layer contains m service nodes, each of which is a basic service node if k=1, and otherwise a combined service node. For basic service nodes, the M/M/1/c queuing model can be adopted to evaluate the service quality; the method comprises the steps of carrying out a first treatment on the surface of the For the combined service node, the M/H is adopted k The/1/c queuing model evaluates its quality of service.
Consider a control nodeThe specific calculation formula is as follows:
service s i (i=1, 2, …, k) at the serving nodeThe response time is as follows:
service s i (i=1, 2, …, k) at the serving nodeThe throughput rate is as follows:
service s i (i=1, 2, …, k) at the serving nodeThe error rate is as follows:
service s i (i=1, 2, …, k) at the serving nodeThe reliability is as follows:
wherein,and->Respectively represent control nodes->Average response time, throughput, error rate, and reliability of processing service requests; /> The service s is a service without considering the quality of service of the control node i At the service node->Response time, throughput rate,Error rate and reliability.
Thus, the invention is based on the M/M/1/c queuing model, M/H k The three service quality evaluation models, namely the 1/c service quality evaluation model and the queuing network model, are respectively used for evaluating the quality evaluation of the basic service node, the combined service node and the service node cluster, can meet the cloud resource service quality evaluation under different conditions, and improve the cloud resource service management level and efficiency, thereby realizing timeliness and scientificity of the overall process supervision, management and control of the cloud resource service.
The following is an embodiment of a queuing theory-based cloud resource quality of service evaluation system provided by the embodiments of the present disclosure, where the system and the queuing theory-based cloud resource quality of service evaluation method of the foregoing embodiments belong to the same inventive concept, and details that are not described in detail in the queuing theory-based cloud resource quality of service evaluation system embodiment may refer to the foregoing queuing theory-based cloud resource quality of service evaluation method embodiment.
The system comprises: the system comprises an evaluation determination module, a node classification module and a service quality evaluation module;
the evaluation determination module is used for determining the processing capacity and the capacity of the service node relative to the service;
the node classification module is used for classifying the service nodes into basic service nodes and combined service nodes according to the service quantity contained in the service nodes;
the service quality evaluation module is used for evaluating the service quality by adopting a corresponding queuing model aiming at different service nodes and outputting an evaluation result.
The units and algorithm steps of each example described in the embodiments disclosed in the cloud resource service quality assessment method and system based on queuing theory can be implemented in electronic hardware, computer software or a combination of the two, and in order to clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described in terms of functions in the above description. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The present invention provides a flow chart of a queuing theory-based cloud resource quality of service assessment method, illustrating the architecture, functionality, and operation of possible implementations of devices, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners.
Furthermore, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the invention. One skilled in the relevant art will recognize, however, that the invention may be practiced without one or more of the specific details, or with other methods, components, devices, steps, etc. In other instances, well-known methods, devices, implementations, or operations are not shown or described in detail to avoid obscuring aspects of the invention.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (3)

1. A cloud resource service quality assessment method based on queuing theory is characterized by comprising the following steps:
s1: determining processing capacity and capacity of the service node relative to the service;
the service node is a virtual unit for deploying services;
let s be a service, vu be a virtual unit, letAnd->The time spent and the memory consumed by the service s on each execution of the virtual unit vu are respectively;
acquiring time and memory data consumed by each execution of a service node by adopting a performance test method, and measuring the average value of the service node by adopting a mathematical statistics method;
the processing capacity of the service node is the average number of user service requests completed by the service node in unit time;
set random variableMean>
Wherein,average number of times the service s is executed on the virtual unit vu per unit time;
the processing power of the service s on the virtual unit vu is defined as:
wherein ε vu The value of the processing capacity coefficient is related to the virtualization technology adopted by the virtual unit and the CPU architecture of the physical server;
the capacity of the service node is the maximum concurrent service request number which can be accepted by the service node;
the capacity of the service s on the virtual unit vu is defined as:D vu maximum available memory for virtual unit vu, < ->Is a random variable +.>Average value of θ vu The value of the capacity coefficient is related to the virtualization technology and the storage architecture adopted by the virtual unit;
s2: classifying the service nodes into basic service nodes and combined service nodes according to the service quantity contained in the service nodes;
the classification method of the service nodes comprises the steps of dividing the service nodes into basic service nodes and combined service nodes according to the number of the services contained in the service nodes, wherein one basic service node contains only one service, and one combined service node contains a plurality of services;
s3: aiming at different service nodes, adopting corresponding queuing models to evaluate service quality, and outputting an evaluation result;
aiming at basic service nodes, a single service window queuing model M/M/1/c with capacity constraint is established to evaluate service quality;
also for the combined service node, a multi-class customer single service window queuing model M/H with capacity constraint is established k 1/c to evaluate the quality of service; a queuing network model is established for the service node cluster to evaluate the service quality;
the manner of evaluating quality of service based on M/M/1/c includes:
configuration ofA basic service node is represented as such,wherein s is a service, vu is a virtual unit, and the processing capacity and capacity of the service s on the virtual unit vu are respectively: mu and c;
basic service nodeThe response time of the service s deployment on the virtual unit vu under load λ is:
basic service nodeI.e. the throughput of the service s deployed on the virtual unit vu under load λ is:
basic service nodeI.e. the error rate of the service s deployment on the virtual unit vu under load λ is:
basic service nodeThe reliability of the service s deployment on the virtual unit vu under load λ is:
M/H based k The manner of evaluating the quality of service/1/c further comprises:
configuration ofRepresents a combined service node, wherein s= { S 1 ,s 2 ,…,s k -service set, vu is a virtual unit;
service s i The processing capacity, capacity and load of the e S on the virtual unit vu are respectively: mu (mu) i ,c i And lambda (lambda) i The method comprises the steps of carrying out a first treatment on the surface of the Order theThe execution time of the virtual unit vu to all services obeys a k-order hyper-exponential distribution, and the distribution function is as follows:
capacity constrained multi-class customer single service window queuing model M/H k 1/c, which emphasizes on analyzing each type of customer performance index in the model;
model pair is distributed in load delta= { lambda% 1 ,λ 2 ,…,λ k Combined service node under }Average quality of service of (a) and each service s i Evaluating the service quality of the E S;
and->Representing service nodes +.>Average response time, throughput rate, error rate and reliability of all services in (a)> And->Representing services s respectively i At the service node->Response time, throughput, error rate and reliability;
the way to build a queuing network model to evaluate quality of service includes:
configuration ofRepresents a service node cluster, wherein s= { S 1 ,s 2 ,…,s k -a service set; VU= { VU 0 ,vu 1 ,…,vu m Is a set of virtual units, vu 0 Vu, a virtual unit of a control node j (j=1, …, m) is a virtual unit of a service node, m is the number of service nodes, and virtual units of different service nodes may be the same;
let the service set S be in the virtual unit vu j The load distribution on this is: delta j ＝{λ 1j ,λ 2j ,…,λ kj }，λ ij (i=1, 2,., k) is service s i In virtual unit vu j Load on, service clusterIs a load distribution of delta = { lambda 1 ,λ 2 ,…,λ k }，
The queuing network model is divided into two layers: a control layer and a service layer;
the control layer is a description of the service request process inside the control node, and the control node is used for controlling the service request processEach service s in (a) i (i=1, 2, …, k) is modeled as a single service node +.>The M/M/1/c model is adopted to evaluate the service quality;
the processing layer comprises m service nodes, if k=1, each service node is a basic service node, otherwise, each service node is a combined service node;
for basic service nodes, the M/M/1/c queuing model is adopted to evaluate the service quality;
for the combined service node, the M/H is adopted k A/1/c queuing model to evaluate its quality of service;
if consider a control nodeAnd (3) quality of service: service s i (i=1, 2, …, k) at service node +.>The response time is as follows:
service s i (i=1, 2, …, k) at the serving nodeThe throughput rate is as follows:
service s i (i=1, 2, …, k) at the serving nodeThe error rate is as follows:
service s i (i=1, 2, …, k) at the serving nodeThe reliability is as follows:
wherein,and->Respectively represent control nodes->Average response time, throughput, error rate, and reliability of processing service requests;
the service s is a service without considering the quality of service of the control node i At the service node->Response time, throughput, error rate, and reliability.
2. The queuing theory-based cloud resource quality of service assessment method of claim 1, wherein the method further comprises: the mapping relation between the service system parameters and the queuing system parameters is constructed, and the specific mapping relation is as follows: the load maps out the average arrival rate of customers, the processing capacity of the service node maps out the average processing rate of the service staff, the capacity of the service node maps out the capacity of the system, the number of concurrent service requests maps out the average queue length, the response time maps out the average stay rate, the throughput rate maps out the absolute throughput capacity, the reliability maps out the relative throughput capacity, and the error rate maps out the loss rate; the capacity of the system includes, among other things, the queue length and the number of attendant.
3. A cloud resource service quality assessment system based on queuing theory, which is characterized in that the system adopts the cloud resource service quality assessment method based on queuing theory according to any one of claims 1 to 2;
the system comprises: the system comprises an evaluation determination module, a node classification module and a service quality evaluation module;
the evaluation determination module is used for determining the processing capacity and the capacity of the service node relative to the service;
the node classification module is used for classifying the service nodes into basic service nodes and combined service nodes according to the service quantity contained in the service nodes;
the service quality evaluation module is used for evaluating the service quality by adopting a corresponding queuing model aiming at different service nodes and outputting an evaluation result.
CN202310071364.8A 2023-01-17 2023-01-17 Cloud resource service quality assessment method and system based on queuing theory Active CN116155835B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310071364.8A CN116155835B (en) 2023-01-17 2023-01-17 Cloud resource service quality assessment method and system based on queuing theory

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310071364.8A CN116155835B (en) 2023-01-17 2023-01-17 Cloud resource service quality assessment method and system based on queuing theory

Publications (2)

Publication Number Publication Date
CN116155835A CN116155835A (en) 2023-05-23
CN116155835B true CN116155835B (en) 2024-04-16

Family

ID=86359560

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310071364.8A Active CN116155835B (en) 2023-01-17 2023-01-17 Cloud resource service quality assessment method and system based on queuing theory

Country Status (1)

Country Link
CN (1) CN116155835B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104038392A (en) * 2014-07-04 2014-09-10 云南电网公司 Method for evaluating service quality of cloud computing resources
CN104123189A (en) * 2014-06-30 2014-10-29 复旦大学 Web multilayer application dynamic resource adjustment method based on IaaS layer application perception
CN105630575A (en) * 2015-12-23 2016-06-01 一兰云联科技股份有限公司 Performance evaluation method aiming at KVM virtualization server
CN107172710A (en) * 2017-04-23 2017-09-15 西安电子科技大学 A kind of resource allocation and service access control method based on virtual subnet
CN110753083A (en) * 2019-09-06 2020-02-04 江苏中云科技有限公司 Cloud service resource uniform distribution system for multiple service providers
CN113515351A (en) * 2021-09-07 2021-10-19 华南理工大学 Resource scheduling implementation method based on energy consumption and QoS (quality of service) cooperative optimization

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9363190B2 (en) * 2013-07-31 2016-06-07 Manjrasoft Pty. Ltd. System, method and computer program product for energy-efficient and service level agreement (SLA)-based management of data centers for cloud computing
WO2015048737A1 (en) * 2013-09-30 2015-04-02 Do Rosario Jackseario Antonio Dionisio Power quality of service optimization for microgrids

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104123189A (en) * 2014-06-30 2014-10-29 复旦大学 Web multilayer application dynamic resource adjustment method based on IaaS layer application perception
CN104038392A (en) * 2014-07-04 2014-09-10 云南电网公司 Method for evaluating service quality of cloud computing resources
CN105630575A (en) * 2015-12-23 2016-06-01 一兰云联科技股份有限公司 Performance evaluation method aiming at KVM virtualization server
CN107172710A (en) * 2017-04-23 2017-09-15 西安电子科技大学 A kind of resource allocation and service access control method based on virtual subnet
CN110753083A (en) * 2019-09-06 2020-02-04 江苏中云科技有限公司 Cloud service resource uniform distribution system for multiple service providers
CN113515351A (en) * 2021-09-07 2021-10-19 华南理工大学 Resource scheduling implementation method based on energy consumption and QoS (quality of service) cooperative optimization

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
An Evaluation Methodology of Cloud Service Quality Based on Queuing Model;G. Yan, F. Meng , D. Chu;2017 IEEE International Symposium on Parallel and Distributed Processing with Applications and 2017 IEEE International Conference on Ubiquitous Computing and Communications (ISPA/IUCC);20171215;925-930 *

Also Published As

Publication number Publication date
CN116155835A (en) 2023-05-23

Similar Documents

Publication Publication Date Title
Guo et al. Cloud resource scheduling with deep reinforcement learning and imitation learning
CN104951425B (en) A kind of cloud service performance self-adapting type of action system of selection based on deep learning
Tuli et al. HUNTER: AI based holistic resource management for sustainable cloud computing
WO2021088207A1 (en) Mixed deployment-based job scheduling method and apparatus for cloud computing cluster, server and storage device
US20120101968A1 (en) Server consolidation system
CN104391879B (en) The method and device of hierarchical clustering
Li et al. Research on QoS service composition based on coevolutionary genetic algorithm
Lockhart et al. Scission: Performance-driven and context-aware cloud-edge distribution of deep neural networks
CN115543626A (en) Power defect image simulation method adopting heterogeneous computing resource load balancing scheduling
CN114936708A (en) Fault diagnosis optimization method based on edge cloud collaborative task unloading and electronic equipment
CN105786626B (en) The Storm platform thread distribution method divided based on K
CN114297041A (en) Network heterogeneous computing platform testing method and device and computer equipment
Zhang Storage optimization algorithm design of cloud computing edge node based on artificial intelligence technology
Shimonishi et al. Energy optimization of distributed video processing system using genetic algorithm with bayesian attractor model
CN116155835B (en) Cloud resource service quality assessment method and system based on queuing theory
Zhu et al. A multi-resource scheduling scheme of Kubernetes for IIoT
Lockhart et al. Scission: Context-aware and performance-driven edge-based distributed deep neural networks
Mays et al. Decentralized data allocation via local benchmarking for parallelized mobile edge learning
JP2019106120A (en) Process scheduling device, process execution device, and process scheduling method
CN111598390B (en) Method, device, equipment and readable storage medium for evaluating high availability of server
AlQerm et al. BEHAVE: Behavior-aware, intelligent and fair resource management for heterogeneous edge-IoT systems
CN111784029A (en) Fog node resource allocation method
Li et al. Fairness-aware task offloading and load balancing with delay constraints for Power Internet of Things
Chen et al. Deep Reinforcement Learning Based Container Cluster Placement Strategy in Edge Computing Environment
Preetham et al. Resource provisioning in cloud using arima and lstm technique

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant