CN103685541A - Device, system and method for dynamically controlling running speed of IaaS (infrastructure as a service) cloud system - Google Patents

Abstract

The invention discloses a device, a system and a method for dynamically controlling the running speed of an IaaS (infrastructure as a service) cloud system, belongs to the field of cloud computing, and aims to change the running speed of a cloud host, instantly respond to the change of a load and accordingly optimize system performances and energy consumption, so that the system takes execution efficiency under a high-load condition and energy conservation under a low-load condition into account. Dynamic fluctuation of the load of the cloud system is taken fully into account, the requirement of the cloud system for future running speed is predicted by tracking the trend of the cloud system, and acceleration or deceleration is controlled in advance. As the influence of abnormal points in historical load data is removed, trend prediction accuracy is ensured. Future expected load gaps are estimated, variable control interval time is set, and the extremes of 'too dense control' and 'too sparse control' are avoided.

Description

IaaS cloud system operating rate device for controlling dynamically, system and method

Technical field

The invention belongs to cloud computing field, particularly relate to a kind of IaaS cloud system operating rate device for controlling dynamically, system and method.

Background technology

Cloud computing is a kind of account form based on the Internet, and in this way, shared software and hardware resources and information can offer computer and other equipment by demand.With respect to traditional software with calculate form, the advantage significantly such as cloud computing has loose couplings, on-demand, cost is controlled, resource is virtual, isomery is collaborative, makes it more adapt to the application such as ecommerce now, flexible manufacturing, mobile Internet.The implication that cloud computing comprises two aspects a: aspect is the cloud computing platform infrastructure that bottom builds, and is for building the basis of upper level applications; Implication is on the other hand the cloud computing application program being structured on this basic platform.Cloud computing can be divided three classes according to COS: using infrastructure as service (IaaS, Infrastructure as a Service), using platform as service (PaaS, Platform as a Service) and using software as service (SaaS, Software as a Service).IaaS cloud is virtual by system virtualization, multiprocessor, internal memory virtualization, I/O are virtual etc., and Intel Virtualization Technology is virtualized into resource pool by physical resource, and these resources are carried out unified management and dispatching by cloud computing platform again.

Main computing unit in IaaS cloud--PM(Physical Machine, physical machine), when operation, can produce power consumption, because cloud computing system now applies to extensive science more, calculate, the application of the high capacity such as finance, online transaction, Streaming Media multicast in real time and high complexity, its energy consumption problem is more outstanding with respect to traditional Distributed Calculation and service compute.In addition, because the Internet that the many frameworks of cloud computing distribute in wide area is upper, the when and where of task requests distributes and embodies erratic behavior and artificial contingency, and therefore the real-time load of cloud system also has more dynamic fluctuation.Owing to there is being difficult to opportunity prediction in the crest of system load and trough, and the managing power consumption strategy of system must be considered " guaranteeing enough processing speeds and task throughput during peak load " and " guaranteeing energy saving during trough load " two requirements simultaneously, there is very large difficulty in the management of the dynamic rate of cloud system main frame.Traditional technology is often difficult to obtain good balance between high-performance and low energy consumption.

In recent years, the importance of dynamic rate expansion technique (DVS, dynamic voltage scaling) in computing system managing power consumption has obtained the extensive understanding of academia and industrial quarters.DVS technology mainly applies in the central processing unit (CPU, central processing unit) of computer originally, and this technology is come real-time adjust operation voltage and frequency by the indexs such as temperature of real-time measurement processor, to guarantee the safety of system.In recent years, this technology was successfully applied in fields such as cluster calculating, large-scale data centers, and the main frame running load that its basic principle is exactly dynamic monitoring, raises running frequency during Task Congestion, when system is idle, lowered.Yet the research of carrying out DVS control in cloud computing and cloud application is also very rare.In addition, with respect to cluster, calculate and large-scale data center, the business of moving in cloud computing is looser flexibly, it is the science calculation task of high complexity both, also the microminiature application program based on mobile terminal, because the variability of task load is larger, thereby can not indiscriminately imitate existing management strategy and algorithm.

Existing technical scheme, the deficiency below main existence: (1) adopts the means of fixed cycle control.Existing method presets a fixing interval time more and periodically controls.Yet due to the dynamically changeable of system load, the control strategy of fixed interval is often difficult to the instant load in short time is increased sharply and makes rapid response.(2) lack the mechanism that quantizes trend prediction.Existing technology, not fully to system historic load analyze, modeling and trend prediction, and be mostly that the historical average or nearest data of mechanical employing are as according to carrying out control decision.(3) do not consider the difference of different main frames.Existing technology is made no exception each main frame in IaaS cloud, takes same control strategy and control time interval, and does not consider the difference of different main frames in load, execution efficiency.

Summary of the invention

Because the above-mentioned defect of prior art, technical problem to be solved by this invention is to provide a kind of controls rationally, can take into account performance under different access loading condition and the IaaS cloud system operating rate dynamic control method of energy-conservation demand.

For achieving the above object, the invention provides a kind of IaaS cloud system operating rate dynamic control method, comprise the following steps:

Step 1, storage information task time of collecting IaaS cloud system main frame:

Collect the time of implementation sequence ct of k having completed of task on IaaS cloud system main frame ₁, ct ₂... ct _kstill carrying out on IaaS cloud system main frame of task is the time of implementation sequence et of use ₁, et ₂... .et _m, wherein m is still in the quantity of carrying out of task, k and m are positive integer;

The time of implementation that step 2, prediction are finished the work:

Because the operation of actual cloud system is subject to the impact of many system factors (message exception deferral, connect bandwidth variation, calculation resources conflict etc.) and nonsystematic factor (the accidental inefficacy of system, software and hardware, information drop-out etc.), in sequence, exist part record value obviously to depart from the situation of overall variation rule, namely so-called abnormity point.These projecting points can not be regarded as general routine data and carry out assessment and analysis, and need to reject.

Set the task of having completed successively in the time of implementation sequence on IaaS cloud system main frame, the logarithm step-length average increment of having rejected after exceptional value is disturbed is loinc;

Calculate loinc=mean{inc _{i, j}| 0 < i < j≤k, the task that yc (i, j)=0} obtains having completed successively the rejecting in the time of implementation sequence on IaaS cloud system main frame the logarithm step-length average increment that disturbs of exceptional value; Described yc (i, j) is judgement inc _{i, j}whether is the function of abnormity point, mean is for gathering the operation of asking arithmetic mean value, inc _{i, j}the logarithm step-length equivalence increment forming between i and j record value in expression sequence, described i and j are positive integer; $\mathrm{in}{c}_{i,j}=\frac{{\mathrm{ct}}_j-{\mathrm{ct}}_i}{\log (j-i)};$

$\mathrm{yc}(i,j)=\left\{\begin{array}{c}1\mathrm{if}\frac{\left|\mathrm{in}{c}_{i,j}\right|}{\mathrm{mn}}>a+1,\left|\mathrm{in}{c}_{i,j}\right|>\max \left\{\right|\mathrm{in}{c}_{t,t+1}\left|\right|i\&le;t\&le;j\}\\ 0\mathrm{else}\end{array}\right.;$

Wherein a is given parameter, 0.1≤a≤0.5;

Mn value is: mn=mean{|inc _{u, v}|| 0 < u < v≤k};

The time of implementation that setting is finished the work is nxt: calculate

$\mathrm{nxt}=\left\{\begin{array}{c}\mathrm{mean}\{c{t}_i+\mathrm{loinc}\&times;e^{k-i}|0<i\&le;k\}\mathrm{if}(\mathrm{loinc}+{\mathrm{ct}}_i)>\max \{{\mathrm{et}}_j|0<j\&le;m\}\\ \min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}+\underset{0<i\&le;k,({\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1})<\min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}}{\mathrm{\&Sigma;}}{\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1}\end{array}\right.$

Obtained the time of implementation of next task.

Step 3, calculating dynamic rate expansion controlled quentity controlled variable and control stand-by period:

Described calculating dynamic rate expansion controlled quentity controlled variable is carried out according to the following steps:

Setting following load breach value is qkz, calculates $\mathrm{qkz}=\underset{0<j\&le;m,{\mathrm{et}}_j>\mathrm{nxt}}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})$ Obtain following load breach value;

Read the current residing speed class L of processor and the equivalent implementation rate exe (L) corresponding to this speed class of IaaS cloud system main frame, it should be noted that, the speed class of processor and implementation rate, determined by its hardware specification, conventionally by equipment supplier, provided.

If IaaS cloud system main frame has n processor, its highest speed class is g level, and L, n and g are positive integer;

Setting dynamic rate expansion controlled quentity controlled variable is kzl,

Calculate $\mathrm{kzl}=\left\{\begin{array}{c}-1\mathrm{ifqkz}=0\\ \min \left\{i\right|n\&GreaterEqual;\frac{\mathrm{qkz}}{\mathrm{exe}(l+i)-\mathrm{exe}{\left(i\right)}^{\&prime;}}0\&le;i\&le;g-l\}\mathrm{else}\end{array}\right.$ Obtain dynamic rate expansion controlled quentity controlled variable; If the following load breach value calculating is 0, represent without promoting the current execution speed of main frame; Otherwise, find the minimum-rate lifting capacity that can fill up load breach.

Described calculating is controlled the stand-by period and is carried out according to the following steps:

According to the time of implementation sequence of completing the time of implementation of next task and still carrying out on IaaS cloud system main frame of task use, determine the stand-by period of once controlling, the stand-by period of setup control is dsj;

Calculate $\mathrm{dsj}=\left\{\begin{array}{c}\mathrm{wcif}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})<0\mathrm{or}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})>0\\ \mathrm{\&beta;}\&times;\mathrm{wcelse}\end{array}\right.,$ Obtain next time

The stand-by period of controlling; Wc is the time in control interval that system initially provides, wc>0,1 < β < 10;

When following load breach value is plus or minus, the current operating rate of expression system, below or above the needs that change of system load trend, needs the instant control of carrying out operating rate, and therefore controlling the stand-by period is wc default value; When following load breach value is 0, expression system present rate is adapted to the needs that system load trend changes just, need not be urgent carry out speed adjustment, therefore controlling the stand-by period is that wc default value is multiplied by a factor beta, in the present embodiment, β is any real number between 2 to 5, and wc value is between 10000 to 100000 milliseconds.

Step 4, IaaS cloud system main frame, according to the stand-by period dsj of described control and dynamic rate expansion controlled quentity controlled variable kzl, are controlled the operating rate of IaaS cloud system main frame in the stand-by period section of controlling.

In the described stand-by period section controlling, the operating rate of IaaS cloud system main frame is controlled according to the following steps and is carried out:

If dynamic rate expansion controlled quentity controlled variable kzl is 0, in the time period that is dsj' in length, keep the original operating rate of main frame; If kzl value is-1, and main frame is current does not operate in minimum speed grade, operating rate is lowered to one-level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is-1, and main frame is currently operating in minimum speed grade, keeps existing speed in the time period that is dsj in length; If kzl value is greater than zero, and main frame current speed grade L adds and is less than or equal to the speed class value g of high permission of main frame after kzl, at once operating rate raised to kzl level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is greater than zero, and main frame current speed grade L adds and is greater than the speed class value of high permission of main frame after kzl, will on operating rate, be transferred to maximum speed grade at once, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value equals zero, in the time period that is dsj in length, keep existing speed;

Then return to execution step 1.

Another technical problem to be solved by this invention is to provide a kind of device that can realize the dynamic control of IaaS cloud system operating rate.

For achieving the above object, the invention provides a kind of IaaS cloud system operating rate device for controlling dynamically, comprise load monitoring module, control decision module and main frame speed control module; The output of described load monitoring module connects the input of described control decision module, and the output of described control decision module connects the input of described main frame speed control module;

Described load monitoring module is for obtaining storage information task time of IaaS cloud system main frame;

Described control decision module is used for having predicted the time of implementation of next task, and calculates dynamic rate expansion controlled quentity controlled variable and carry out the control stand-by period of task next time;

Described main frame speed control module is for controlling the operating rate of IaaS cloud system main frame.

Described load monitoring module is collected the time of implementation sequence C t of k having completed of task on IaaS cloud system main frame ₁, Ct ₂... Ct _kstill carrying out on IaaS cloud system main frame of task is the time of implementation sequence et of use ₁, et ₂... .et _m, wherein m is still in the quantity of carrying out of task, k and m are positive integer.

Described control decision module comprises data pretreatment unit, controlled quentity controlled variable computing unit and control decision package on opportunity; The first output of described load monitoring module connects the input of described data pretreatment unit, the second output of described load monitoring module connects the first input end of described controlled quentity controlled variable computing unit, and the 3rd output of described load monitoring module connects the first input end of described control decision package on opportunity; The first output of described data pretreatment unit connects the second input of described controlled quentity controlled variable computing unit, and the second output of described data pretreatment unit connects the second input of described control decision package on opportunity; The output of described controlled quentity controlled variable computing unit connects the first input end of described main frame speed control module, and the output of described control decision package on opportunity connects the second input of described main frame speed control module;

Described data pretreatment unit is for having predicted the time of implementation of next task;

Described controlled quentity controlled variable computing unit is used for calculating dynamic rate expansion controlled quentity controlled variable;

Described control decision package on opportunity is carried out the control stand-by period of task next time for calculating;

Described data pretreatment unit is set the task of having completed successively in the time of implementation sequence on IaaS cloud system main frame, and the logarithm step-length average increment of having rejected after exceptional value is disturbed is loinc;

Calculate loinc=mean{inc _{i, j}| 0 < i < j≤k, the task that yc (i, j)=0} obtains having completed successively the rejecting in the time of implementation sequence on IaaS cloud system main frame the logarithm step-length average increment that disturbs of exceptional value; Described yc (i, j) is judgement inc _{i, j}whether is the function of abnormity point, mean is for gathering the operation of asking arithmetic mean value, inc _i,jthe logarithm step-length equivalence increment forming between i and j record value in expression sequence, described i and j are positive integer; $\mathrm{in}{c}_{i,j}=\frac{{\mathrm{ct}}_j-{\mathrm{ct}}_i}{\log (j-i)};$

$\mathrm{yc}(i,j)=\left\{\begin{array}{c}1\mathrm{if}\frac{\left|\mathrm{in}{c}_{i,j}\right|}{\mathrm{mn}}>a+1,\left|\mathrm{in}{c}_{i,j}\right|>\max \left\{\right|\mathrm{in}{c}_{t,t+1}\left|\right|i\&le;t\&le;j\}\\ 0\mathrm{else}\end{array}\right.;$

Wherein a is given parameter, 0.1≤a≤0.5; Mn value is: mn=mean{|inc _{u, v}|| 0 < u < v≤k}; The time of implementation that setting is finished the work is nxt: described data pretreatment unit calculates

$\mathrm{nxt}=\left\{\begin{array}{c}\mathrm{mean}\{c{t}_i+\mathrm{loinc}\&times;e^{k-i}|0<i\&le;k\}\mathrm{if}(\mathrm{loinc}+{\mathrm{ct}}_i)>\max \{{\mathrm{et}}_j|0<j\&le;m\}\\ \min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}+\underset{0<i\&le;k,({\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1})<\min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}}{\mathrm{\&Sigma;}}{\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1}\end{array}\right.$ The time of implementation nxt that obtains finishing the work; Described data pretreatment unit is issued controlled quentity controlled variable computing unit and the control decision package on opportunity in control decision module by nxt value; It is qkz that described controlled quentity controlled variable computing unit is set following load breach value, and described controlled quentity controlled variable computing unit calculates $\mathrm{qkz}=\underset{0<j\&le;m,{\mathrm{et}}_j>\mathrm{nxt}}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})$ Obtain following load breach value;

Described controlled quentity controlled variable computing unit reads the current residing speed class L of processor and the equivalent implementation rate exe (L) corresponding to this speed class of IaaS cloud system main frame, if IaaS cloud system main frame has n processor, its the highest speed class is g level, and L, n and g are positive integer; Setting dynamic rate expansion controlled quentity controlled variable is kzl,

By calculating $\mathrm{kzl}=\left\{\begin{array}{c}-1\mathrm{ifqkz}=0\\ \min \left\{i\right|n\&GreaterEqual;\frac{\mathrm{qkz}}{\mathrm{exe}(l+i)-\mathrm{exe}{\left(i\right)}^{\&prime;}}0\&le;i\&le;g-l\}\mathrm{else}\end{array}\right.$ Described controlled quentity controlled variable computing unit obtains dynamic rate expansion controlled quentity controlled variable; Described controlled quentity controlled variable computing unit sends to described main frame speed control module by kzl value; Described control decision package on opportunity determines the stand-by period of once controlling according to the time of implementation sequence of completing the time of implementation of next task and still carrying out on IaaS cloud system main frame of task use, the stand-by period of setup control is dsj;

By calculating $\mathrm{dsj}=\left\{\begin{array}{c}\mathrm{wcif}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})<0\mathrm{or}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})>0\\ \mathrm{\&beta;}\&times;\mathrm{wcelse}\end{array}\right.,$ The described control stand-by period that opportunity, decision package obtained controlling next time; Wc is the time in control interval that system initially provides, wc>0,1 < β < 10; Described control decision package on opportunity is issued described main frame speed control module by the stand-by period dsj controlling next time.

Described main frame speed control module, according to the stand-by period dsj of described control and dynamic rate expansion controlled quentity controlled variable kzl, is controlled the operating rate of IaaS cloud system main frame in the stand-by period section of controlling:

If dynamic rate expansion controlled quentity controlled variable kzl is 0, in the time period that is dsj in length, keep the original operating rate of main frame; If kzl value is-1, and main frame is current does not operate in minimum speed grade, operating rate is lowered to one-level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is-1, and main frame is currently operating in minimum speed grade, keeps existing speed in the time period that is dsj in length; If kzl value is greater than zero, and main frame current speed grade L adds and is less than or equal to the speed class value g of high permission of main frame after kzl, at once operating rate raised to kzl level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is greater than zero, and main frame current speed grade L adds and is greater than the speed class value of high permission of main frame after kzl, will on operating rate, be transferred to maximum speed grade at once, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value equals zero, in the time period that is dsj in length, keep existing speed.

The technical problem that the present invention also will solve is to provide a kind of system that can realize the dynamic control of IaaS cloud system operating rate.

For achieving the above object, the invention provides a kind of IaaS cloud system operating rate kinetic-control system, comprise the server of IaaS cloud system, in the server of described IaaS cloud system, be provided with IaaS cloud system operating rate device for controlling dynamically; Described IaaS cloud system operating rate device for controlling dynamically comprises load monitoring module, control decision module and main frame speed control module; The output of described load monitoring module connects the input of described control decision module, and the output of described control decision module connects the input of described main frame speed control module;

Described load monitoring module is for obtaining storage information task time of IaaS cloud system main frame;

Described control decision module is used for having predicted the time of implementation of next task, and calculates dynamic rate expansion controlled quentity controlled variable and carry out the control stand-by period of task next time;

Described main frame speed control module is for controlling the operating rate of IaaS cloud system main frame;

Described load monitoring module is collected the time of implementation sequence ct of k having completed of task on IaaS cloud system main frame ₁, ct ₂... ct _kstill carrying out on IaaS cloud system main frame of task is the time of implementation sequence et of use ₁, et ₂... .et _m, wherein m is still in the quantity of carrying out of task, k and m are positive integer;

Described control decision module comprises data pretreatment unit, controlled quentity controlled variable computing unit and control decision package on opportunity; The first output of described load monitoring module connects the input of described data pretreatment unit, the second output of described load monitoring module connects the first input end of described controlled quentity controlled variable computing unit, and the 3rd output of described load monitoring module connects the first input end of described control decision package on opportunity; The first output of described data pretreatment unit connects the second input of described controlled quentity controlled variable computing unit, and the second output of described data pretreatment unit connects the second input of described control decision package on opportunity; The output of described controlled quentity controlled variable computing unit connects the first input end of described main frame speed control module, and the output of described control decision package on opportunity connects the second input of described main frame speed control module;

Described data pretreatment unit is for having predicted the time of implementation of next task;

Described controlled quentity controlled variable computing unit is used for calculating dynamic rate expansion controlled quentity controlled variable;

Described control decision package on opportunity is carried out the control stand-by period of task next time for calculating;

Described data pretreatment unit is set the task of having completed successively in the time of implementation sequence on IaaS cloud system main frame, and the logarithm step-length average increment of having rejected after exceptional value is disturbed is loinc;

Calculate loinc=mean{inc _{i, j}| 0 < i < j≤k, the task that yc (i, j)=0} obtains having completed successively the rejecting in the time of implementation sequence on IaaS cloud system main frame the logarithm step-length average increment that disturbs of exceptional value; Described yc (i, j) is judgement inc _{i, j}whether is the function of abnormity point, mean is for gathering the operation of asking arithmetic mean value, inc _{i, j}represent the 1st opening in sequence] the logarithm step-length equivalence increment that forms between individual record value, described i and j are positive integer; $\mathrm{in}{c}_{i,j}=\frac{{\mathrm{ct}}_j-{\mathrm{ct}}_i}{\log (j-i)};$

$\mathrm{yc}(i,j)=\left\{\begin{array}{c}1\mathrm{if}\frac{\left|\mathrm{in}{c}_{i,j}\right|}{\mathrm{mn}}>a+1,\left|\mathrm{in}{c}_{i,j}\right|>\max \left\{\right|\mathrm{in}{c}_{t,t+1}\left|\right|i\&le;t\&le;j\}\\ 0\mathrm{else}\end{array}\right.;$

Wherein a is given parameter, 0.1≤a≤0.5;

Mn value is: mn=mean{|inc _{u, v}|| 0 < u < v≤k};

The time of implementation that setting is finished the work is nxt: described data pretreatment unit calculates

$\mathrm{nxt}=\left\{\begin{array}{c}\mathrm{mean}\{c{t}_i+\mathrm{loinc}\&times;e^{k-i}|0<i\&le;k\}\mathrm{if}(\mathrm{loinc}+{\mathrm{ct}}_i)>\max \{{\mathrm{et}}_j|0<j\&le;m\}\\ \min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}+\underset{0<i\&le;k,({\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1})<\min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}}{\mathrm{\&Sigma;}}{\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-\mathrm{<figure-callout\; id="1"\; label="i\; +"\; filenames="HDA0000444347320000012.png,HDA0000444347320000021.png"\; state="\{\{state\}\}">i</figure-callout>}+1}\end{array}\right.$

The time of implementation nxt that obtains finishing the work; Described data pretreatment unit is issued controlled quentity controlled variable computing unit and the control decision package on opportunity in control decision module by nxt value;

It is qkz that described controlled quentity controlled variable computing unit is set following load breach value,

Described controlled quentity controlled variable computing unit calculates $\mathrm{qkz}=\underset{0<j\&le;m,{\mathrm{et}}_j>\mathrm{nxt}}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})$ Obtain following load breach value;

Described controlled quentity controlled variable computing unit reads the current residing speed class L of processor and the equivalent implementation rate exe (L) corresponding to this speed class of IaaS cloud system main frame, if IaaS cloud system main frame has n processor, its the highest speed class is g level, and L, n and g are positive integer;

Setting dynamic rate expansion controlled quentity controlled variable is kzl,

By calculating $\mathrm{kzl}=\left\{\begin{array}{c}-1\mathrm{ifqkz}=0\\ \min \left\{i\right|n\&GreaterEqual;\frac{\mathrm{qkz}}{\mathrm{exe}(l+i)-\mathrm{exe}{\left(i\right)}^{\&prime;}}0\&le;i\&le;g-l\}\mathrm{else}\end{array}\right.$ Described controlled quentity controlled variable computing unit obtains dynamic rate expansion controlled quentity controlled variable; Described controlled quentity controlled variable computing unit sends to described main frame speed control module by kzl value;

Described control decision package on opportunity determines the stand-by period of once controlling according to the time of implementation sequence of completing the time of implementation of next task and still carrying out on IaaS cloud system main frame of task use, the stand-by period of setup control is dsj;

By calculating $\mathrm{dsj}=\left\{\begin{array}{c}\mathrm{wcif}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})<0\mathrm{or}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})>0\\ \mathrm{\&beta;}\&times;\mathrm{wcelse}\end{array}\right.,$ The described control stand-by period that opportunity, decision package obtained controlling next time; Wc is the time in control interval that system initially provides, wc>0,1 < β < 10; Described control decision package on opportunity is issued described main frame speed control module by the stand-by period dsj controlling next time.

Described main frame speed control module sending controling instruction, to IaaS cloud system main frame, to expand controlled quentity controlled variable kzl according to the stand-by period dsj of described control and dynamic rate, is controlled the operating rate of IaaS cloud system main frame in the stand-by period section of controlling:

If dynamic rate expansion controlled quentity controlled variable kzl is 0, in the time period that is dsj in length, keep the original operating rate of main frame; If kzl value is-1, and main frame is current does not operate in minimum speed grade, operating rate is lowered to one-level, and the speed after keeping in the time period that is dsj in length adjusting; If kal value is-1, and main frame is currently operating in minimum speed grade, keeps existing speed in the time period that is dsj in length; If kzl value is greater than zero, and main frame current speed grade L adds and is less than or equal to the speed class value g of high permission of main frame after kzl, at once operating rate raised to kzl level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is greater than zero, and main frame current speed grade L adds and is greater than the speed class value of high permission of main frame after kzl, will on operating rate, be transferred to maximum speed grade at once, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value equals zero, in the time period that is dsj in length, keep existing speed.

The invention has the beneficial effects as follows: the present invention has taken into full account the dynamic fluctuation of cloud system load, by following the tracks of its trend predicting system speed of service needs in future, and carry out in advance the control of speedup or deceleration; Owing to having rejected the abnormity point impact in historic load, guaranteed the accuracy of trend prediction; The present invention can be according to cloud system load variations trend, the load breach of the following expection of estimation, and formulate the variable time in control interval, avoided " controlling overstocked " and " controlling thin " two extreme, play the effect that reduces energy consumption.

Accompanying drawing explanation

Fig. 1 is the schematic flow sheet of IaaS cloud system operating rate dynamic control method in the present invention.

Fig. 2 is the circuit theory schematic diagram of IaaS cloud system operating rate device for controlling dynamically in the present invention.

Fig. 3 is the circuit theory schematic diagram of IaaS cloud system operating rate kinetic-control system in the present invention.

Embodiment

Below in conjunction with drawings and Examples, the invention will be further described:

As shown in Figure 1, a kind of IaaS cloud system operating rate dynamic control method, comprises the following steps:

Step 1, storage information task time of collecting IaaS cloud system main frame:

In the present embodiment, collect the time of implementation sequence ct of nearest k having completed of task on IaaS cloud system main frame _l, ct ₂... ct _kstill carrying out on IaaS cloud system main frame of task is the time of implementation sequence et of use ₁, et ₂... .et _m, wherein m is still in the quantity of carrying out of task, k and m are positive integer, 5≤k≤10, described time series ct ₁, ct ₂... ct _kfor according to completing the sequencing time of implementation sequence on IaaS cloud system main frame successively.

The time of implementation that step 2, prediction are finished the work:

Set the task of having completed successively in the time of implementation sequence on IaaS cloud system main frame, the logarithm step-length average increment of having rejected after exceptional value is disturbed is loinc;

Calculate loinc=mean{inc _{i, j}| 0 < i < j≤k, the task that yc (i, j)=0} obtains having completed successively the rejecting in the time of implementation sequence on IaaS cloud system main frame the logarithm step-length average increment that disturbs of exceptional value; Described yc (i, j) is judgement inc _{i, j}whether is the function of abnormity point, mea is for gathering the operation of asking arithmetic mean value, inc _{i, j}the logarithm step-length equivalence increment forming between i and j record value in expression sequence, described i and j are positive integer; $\mathrm{in}{c}_{i,j}=\frac{{\mathrm{ct}}_j-{\mathrm{ct}}_i}{\log (j-i)};$

$\mathrm{yc}(i,j)=\left\{\begin{array}{c}1\mathrm{if}\frac{\left|\mathrm{in}{c}_{i,j}\right|}{\mathrm{mn}}>a+1,\left|\mathrm{in}{c}_{i,j}\right|>\max \left\{\right|\mathrm{in}{c}_{t,t+1}\left|\right|i\&le;t\&le;j\}\\ 0\mathrm{else}\end{array}\right.;$

Wherein a is given parameter, 0.1≤a≤0.5;

Mn value is: mn=mean{|inc _{u, v}|| 0 < u < v≤k};

The time of implementation that setting is finished the work is nxt: calculate

$\mathrm{nxt}=\left\{\begin{array}{c}\mathrm{mean}\{c{t}_i+\mathrm{loinc}\&times;e^{k-i}|0<i\&le;k\}\mathrm{if}(\mathrm{loinc}+{\mathrm{ct}}_i)>\max \{{\mathrm{et}}_j|0<j\&le;m\}\\ \min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}+\underset{0<i\&le;k,({\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1})<\min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}}{\mathrm{\&Sigma;}}{\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-\mathrm{<figure-callout\; id="1"\; label="i\; +"\; filenames="HDA0000444347320000012.png,HDA0000444347320000021.png"\; state="\{\{state\}\}">i</figure-callout>}+1}\end{array}\right.$

Obtained the time of implementation of next task.

Step 3, calculating dynamic rate expansion controlled quentity controlled variable and control stand-by period:

Described calculating dynamic rate expansion controlled quentity controlled variable is carried out according to the following steps:

Setting following load breach value is qkz, calculates $\mathrm{qkz}=\underset{0<j\&le;m,{\mathrm{et}}_j>\mathrm{nxt}}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})$ Obtain following load breach value;

Read the current residing speed class L of processor and the equivalent implementation rate exe (L) corresponding to this speed class of IaaS cloud system main frame,

If IaaS cloud system main frame has n processor, its highest speed class is g level, and L, n and g are positive integer;

Setting dynamic rate expansion controlled quentity controlled variable is kzl,

Calculate $\mathrm{kzl}=\left\{\begin{array}{c}-1\mathrm{ifqkz}=0\\ \min \left\{i\right|n\&GreaterEqual;\frac{\mathrm{qkz}}{\mathrm{exe}(l+i)-\mathrm{exe}{\left(i\right)}^{\&prime;}}0\&le;i\&le;g-l\}\mathrm{else}\end{array}\right.$ Obtain dynamic rate expansion controlled quentity controlled variable;

Described calculating is controlled the stand-by period and is carried out according to the following steps:

According to the time of implementation sequence of completing the time of implementation of next task and still carrying out on IaaS cloud system main frame of task use, determine the stand-by period of once controlling, the stand-by period of setup control is dsj;

Calculate $\mathrm{dsj}=\left\{\begin{array}{c}\mathrm{wcif}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})<0\mathrm{or}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})>0\\ \mathrm{\&beta;}\&times;\mathrm{wcelse}\end{array}\right.,$ The stand-by period that obtains next time controlling; Wc is the time in control interval that system initially provides, wc>0,1 < β < 10

Step 4, IaaS cloud system main frame, according to the stand-by period dsj of described control and dynamic rate expansion controlled quentity controlled variable kzl, are controlled the operating rate of IaaS cloud system main frame in the stand-by period section of controlling.

In the described stand-by period section controlling, the operating rate of IaaS cloud system main frame is controlled according to the following steps and is carried out:

If dynamic rate expansion controlled quentity controlled variable kzl is 0, in the time period that is dsj in length, keep the original operating rate of main frame; If kzl value is-1, and main frame is current does not operate in minimum speed grade, operating rate is lowered to one-level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is-1, and main frame is currently operating in minimum speed grade, keeps existing speed in the time period that is dsj in length; If kzl value is greater than zero, and main frame current speed grade L adds and is less than or equal to the speed class value g of high permission of main frame after kzl, at once operating rate raised to kzl level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is greater than zero, and main frame current speed grade L adds and is greater than the speed class value of high permission of main frame after kzl, will on operating rate, be transferred to maximum speed grade at once, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value equals zero, in the time period that is dsj in length, keep existing speed.After completing, control returns to execution step 1.

As shown in Figure 2, a kind of IaaS cloud system operating rate device for controlling dynamically, comprises load monitoring module 1, control decision module 2 and main frame speed control module 3; The output of described load monitoring module 1 connects the input of described control decision module 2, and the output of described control decision module 2 connects the input of described main frame speed control module 3;

Described load monitoring module 1 is for obtaining storage information task time of IaaS cloud system main frame;

Described control decision module 2 is for having predicted the time of implementation of next task, and calculates dynamic rate expansion controlled quentity controlled variable and carry out the control stand-by period of task next time;

Described main frame speed control module 3 is for controlling the operating rate of IaaS cloud system main frame.

Described load monitoring module 1 is collected the time of implementation sequence ct of k having completed of task on IaaS cloud system main frame ₁, ct ₂... ct _kstill carrying out on IaaS cloud system main frame of task is the time of implementation sequence et of use ₁, et ₂... .et _m, wherein m is still in the quantity of carrying out of task, k and m are positive integer.

Described control decision module 2 comprises data pretreatment unit 201, controlled quentity controlled variable computing unit 202 and control decision package on opportunity 203; The first output of described load monitoring module 1 connects the input of described data pretreatment unit 201, the second output of described load monitoring module 1 connects the first input end of described controlled quentity controlled variable computing unit 202, and the 3rd output of described load monitoring module 1 connects the first input end of described control decision package on opportunity 203; The first output of described data pretreatment unit 201 connects the second input of described controlled quentity controlled variable computing unit 202, and the second output of described data pretreatment unit 201 connects the second input of described control decision package on opportunity 203; The output of described controlled quentity controlled variable computing unit 202 connects the first input end of described main frame speed control module 3, and the output of described control decision package on opportunity 203 connects the second input of described main frame speed control module 3.

Described data pretreatment unit 201 is for having predicted the time of implementation of next task;

Described controlled quentity controlled variable computing unit 202 is for calculating dynamic rate expansion controlled quentity controlled variable;

Described control decision package on opportunity 203 is carried out the control stand-by period of task next time for calculating.

Described data pretreatment unit 201 is set the task of having completed successively in the time of implementation sequence on IaaS cloud system main frame, and the logarithm step-length average increment of having rejected after exceptional value is disturbed is loinc;

Calculate loinc=mean{inc _{i, j}| 0 < i < j≤k, the task that yc (i, j)=0} obtains having completed successively the rejecting in the time of implementation sequence on IaaS cloud system main frame the logarithm step-length average increment that disturbs of exceptional value; Described yc (i, j) is judgement inc _{i, j}whether is the function of abnormity point, mean is for gathering the operation of asking arithmetic mean value, inc _{i, j}the logarithm step-length equivalence increment forming between i and j record value in expression sequence, described i and j are positive integer; $\mathrm{in}{c}_{i,j}=\frac{{\mathrm{ct}}_j-{\mathrm{ct}}_i}{\log (j-i)};$

$\mathrm{yc}(i,j)=\left\{\begin{array}{c}1\mathrm{if}\frac{\left|\mathrm{in}{c}_{i,j}\right|}{\mathrm{mn}}>a+1,\left|\mathrm{in}{c}_{i,j}\right|>\max \left\{\right|\mathrm{in}{c}_{t,t+1}\left|\right|i\&le;t\&le;j\}\\ 0\mathrm{else}\end{array}\right.;$

Wherein a is given parameter, 0.1≤a≤0.5;

Mn value is: mn=mean{|inc _{u, v}|| 0 < u < v≤k};

The time of implementation that setting is finished the work is nxt: described data pretreatment unit 201 calculates

$\mathrm{nxt}=\left\{\begin{array}{c}\mathrm{mean}\{c{t}_i+\mathrm{loinc}\&times;e^{k-i}|0<i\&le;k\}\mathrm{if}(\mathrm{loinc}+{\mathrm{ct}}_i)>\max \{{\mathrm{et}}_j|0<j\&le;m\}\\ \min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}+\underset{0<i\&le;k,({\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1})<\min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}}{\mathrm{\&Sigma;}}{\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-\mathrm{<figure-callout\; id="1"\; label="i\; +"\; filenames="HDA0000444347320000012.png,HDA0000444347320000021.png"\; state="\{\{state\}\}">i</figure-callout>}+1}\end{array}\right.$

The time of implementation nxt that obtains finishing the work; Described data pretreatment unit 201 is issued controlled quentity controlled variable computing unit 202 and the control decision package on opportunity 203 in control decision module 2 by nxt value;

It is qkz that described controlled quentity controlled variable computing unit 202 is set following load breach value,

Described controlled quentity controlled variable computing unit 202 calculates $\mathrm{qkz}=\underset{0<j\&le;m,{\mathrm{et}}_j>\mathrm{nxt}}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})$ Obtain following load breach value;

Described controlled quentity controlled variable computing unit 202 reads the current residing speed class L of processor and the equivalent implementation rate exe (L) corresponding to this speed class of IaaS cloud system main frame, if IaaS cloud system main frame has n processor, its the highest speed class is g level, and L, n and g are positive integer;

Setting dynamic rate expansion controlled quentity controlled variable is kzl,

By calculating $\mathrm{kzl}=\left\{\begin{array}{c}-1\mathrm{ifqkz}=0\\ \min \left\{i\right|n\&GreaterEqual;\frac{\mathrm{qkz}}{\mathrm{exe}(l+i)-\mathrm{exe}{\left(i\right)}^{\&prime;}}0\&le;i\&le;g-l\}\mathrm{else}\end{array}\right.$ Described controlled quentity controlled variable computing unit 202 obtains dynamic rate expansion controlled quentity controlled variable; Described controlled quentity controlled variable computing unit 202 sends to described main frame speed control module 3 by kzl value;

Described control decision package on opportunity 203 determines the stand-by period of once controlling according to the time of implementation sequence of completing the time of implementation of next task and still carrying out on IaaS cloud system main frame of task use, the stand-by period of setup control is dsj;

By calculating $\mathrm{dsj}=\left\{\begin{array}{c}\mathrm{wcif}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})<0\mathrm{or}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})>0\\ \mathrm{\&beta;}\&times;\mathrm{wcelse}\end{array}\right.,$ The stand-by period that described control decision package on opportunity 203 obtains controlling next time; Wc is the time in control interval that system initially provides, wc>0,1 < β < 10; Described control decision package on opportunity 203 is issued described main frame speed control module 3 by the stand-by period dsj controlling next time.

Described main frame speed control module 3, according to the stand-by period dsj of described control and dynamic rate expansion controlled quentity controlled variable kzl, is controlled the operating rate of IaaS cloud system main frame in the stand-by period section of controlling:

If dynamic rate expansion controlled quentity controlled variable kzl is 0, in the time period that is dsj in length, keep the original operating rate of main frame; If kzl value is-1, and main frame is current does not operate in minimum speed grade, operating rate is lowered to one-level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is-1, and main frame is currently operating in minimum speed grade, keeps existing speed in the time period that is dsj in length; If kzl value is greater than zero, and main frame current speed grade L adds and is less than or equal to the speed class value g of high permission of main frame after kzl, at once operating rate raised to kzl level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is greater than zero, and main frame current speed grade L adds and is greater than the speed class value of high permission of main frame after kzl, will on operating rate, be transferred to maximum speed grade at once, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value equals zero, in the time period that is ds in length, keep existing speed.

As shown in Figure 3, a kind of IaaS cloud system operating rate kinetic-control system, comprises and the server 4 of IaaS cloud system in the server 4 of described IaaS cloud system, is provided with IaaS cloud system operating rate device for controlling dynamically 5; Described IaaS cloud system operating rate device for controlling dynamically 5 comprises load monitoring module 1, control decision module 2 and main frame speed control module 3; Comprise data pretreatment unit 201, controlled quentity controlled variable computing unit 202 and control decision package on opportunity 203; The first output of described load monitoring module 1 connects the input of described data pretreatment unit 201, the second output of described load monitoring module 1 connects the first input end of described controlled quentity controlled variable computing unit 202, and the 3rd output of described load monitoring module 1 connects the first input end of described control decision package on opportunity 203; The first output of described data pretreatment unit 201 connects the second input of described controlled quentity controlled variable computing unit 202, and the second output of described data pretreatment unit 201 connects the second input of described control decision package on opportunity 203; The output of described controlled quentity controlled variable computing unit 202 connects the first input end of described main frame speed control module 3, and the output of described control decision package on opportunity 203 connects the second input of described main frame speed control module 3.

Described load monitoring module 1 is for obtaining storage information task time of IaaS cloud system main frame;

Described data pretreatment unit 201 is for having predicted the time of implementation of next task;

Described controlled quentity controlled variable computing unit 202 is for calculating dynamic rate expansion controlled quentity controlled variable;

Described control decision package on opportunity 203 is carried out the control stand-by period of task next time for calculating;

Described main frame speed control module 3 is for controlling the operating rate of IaaS cloud system main frame;

Described load monitoring module 1 is collected the time of implementation sequence ct of k having completed of task on IaaS cloud system main frame ₁, ct ₂... ct _kstill carrying out on IaaS cloud system main frame of task is the time of implementation sequence et of use ₁, et ₂... .et _m, wherein m is still in the quantity of carrying out of task, k and m are positive integer;

Described data pretreatment unit 201 is set the task of having completed successively in the time of implementation sequence on IaaS cloud system main frame, and the logarithm step-length average increment of having rejected after exceptional value is disturbed is loinc;

Calculate loinc=mean{inc _{i, j}| 0 < i < j≤k, the task that yc (i, j)=0} obtains having completed successively the rejecting in the time of implementation sequence on IaaS cloud system main frame the logarithm step-length average increment that disturbs of exceptional value; Described yc (i, j) is judgement inc _{i, j}whether is the function of abnormity point, mean is for gathering the operation of asking arithmetic mean value, inc _{i, j}the logarithm step-length equivalence increment forming between i and j record value in expression sequence, described i and j are positive integer; $\mathrm{in}{c}_{i,j}=\frac{{\mathrm{ct}}_j-{\mathrm{ct}}_i}{\log (j-i)};$

$\mathrm{yc}(i,j)=\left\{\begin{array}{c}1\mathrm{if}\frac{\left|\mathrm{in}{c}_{i,j}\right|}{\mathrm{mn}}>a+1,\left|\mathrm{in}{c}_{i,j}\right|>\max \left\{\right|\mathrm{in}{c}_{t,t+1}\left|\right|i\&le;t\&le;j\}\\ 0\mathrm{else}\end{array}\right.;$

Wherein a is given parameter, 0.1≤a≤0.5;

Mn value is: mn=mean{|inc _{u, v}|| 0 < u < v≤: k};

The time of implementation that setting is finished the work is nxt: described data pretreatment unit 201 calculates

$\mathrm{nxt}=\left\{\begin{array}{c}\mathrm{mean}\{c{t}_i+\mathrm{loinc}\&times;e^{k-i}|0<i\&le;k\}\mathrm{if}(\mathrm{loinc}+{\mathrm{ct}}_i)>\max \{{\mathrm{et}}_j|0<j\&le;m\}\\ \min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}+\underset{0<i\&le;k,({\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1})<\min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}}{\mathrm{\&Sigma;}}{\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-\mathrm{<figure-callout\; id="1"\; label="i\; +"\; filenames="HDA0000444347320000012.png,HDA0000444347320000021.png"\; state="\{\{state\}\}">i</figure-callout>}+1}\end{array}\right.$

The time of implementation nxt that obtains finishing the work; Described data pretreatment unit 201 is issued controlled quentity controlled variable computing unit 202 and the control decision package on opportunity 203 in control decision module 2 by nxt value;

It is qkz that described controlled quentity controlled variable computing unit 202 is set following load breach value,

Described controlled quentity controlled variable computing unit 202 calculates $\mathrm{qkz}=\underset{0<j\&le;m,{\mathrm{et}}_j>\mathrm{nxt}}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})$ Obtain following load breach value;

Described controlled quentity controlled variable computing unit 202 reads the current residing speed class L of processor and the equivalent implementation rate exe (L) corresponding to this speed class of IaaS cloud system main frame, if IaaS cloud system main frame has n processor, its the highest speed class is g level, and L, n and g are positive integer;

Setting dynamic rate expansion controlled quentity controlled variable is kzl,

By calculating $\mathrm{kzl}=\left\{\begin{array}{c}-1\mathrm{ifqkz}=0\\ \min \left\{i\right|n\&GreaterEqual;\frac{\mathrm{qkz}}{\mathrm{exe}(l+i)-\mathrm{exe}{\left(i\right)}^{\&prime;}}0\&le;i\&le;g-l\}\mathrm{else}\end{array}\right.$ Described controlled quentity controlled variable computing unit 202 obtains dynamic rate expansion controlled quentity controlled variable; Described controlled quentity controlled variable computing unit 202 sends to described main frame speed control module 3 by kzl value;

Described control decision package on opportunity 203 determines the stand-by period of once controlling according to the time of implementation sequence of completing the time of implementation of next task and still carrying out on IaaS cloud system main frame of task use, the stand-by period of setup control is dsj;

By calculating $\mathrm{dsj}=\left\{\begin{array}{c}\mathrm{wcif}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})<0\mathrm{or}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})>0\\ \mathrm{\&beta;}\&times;\mathrm{wcelse}\end{array}\right.,$ The stand-by period that described control decision package on opportunity 203 obtains controlling next time; Wc is the time in control interval that system initially provides, wc>0,1 < β < 10; Described control decision package on opportunity 203 is issued described main frame speed control module 3 by the stand-by period dsj controlling next time.

Described main frame speed control module 3 sending controling instructions, to IaaS cloud system main frame, to expand controlled quentity controlled variable kzl according to the stand-by period dsj of described control and dynamic rate, are controlled the operating rate of IaaS cloud system main frame in the stand-by period section of controlling:

If dynamic rate expansion controlled quentity controlled variable kzl is 0, in the time period that is dsj in length, keep the original operating rate of main frame; If kzl value is-1, and main frame is current does not operate in minimum speed grade, operating rate is lowered to one-level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is-1, and main frame is currently operating in minimum speed grade, keeps existing speed in the time period that is dsj in length; If kzl value is greater than zero, and main frame current speed grade L adds and is less than or equal to the speed class value g of high permission of main frame after kzl, at once operating rate raised to kzl level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is greater than zero, and main frame current speed grade L adds and is greater than the speed class value of high permission of main frame after kzl, will on operating rate, be transferred to maximum speed grade at once, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value equals zero, in the time period that is dsj in length, keep existing speed.

More than describe preferred embodiment of the present invention in detail.In the description of this specification, the description of reference term " embodiment ", " some embodiment ", " example ", " concrete example " or " some examples " etc. means to be contained at least one embodiment of the present invention or example in conjunction with specific features, structure, material or the feature of this embodiment or example description.In this manual, the schematic statement of above-mentioned term is not necessarily referred to identical embodiment or example.And the specific features of description, structure, material or feature can be with suitable mode combinations in any one or more embodiment or example.

Although illustrated and described embodiments of the invention, those having ordinary skill in the art will appreciate that: in the situation that not departing from principle of the present invention and aim, can carry out multiple variation, modification, replacement and modification to these embodiment, scope of the present invention is limited by claim and equivalent thereof.

Claims (7)

1. an IaaS cloud system operating rate dynamic control method, is characterized in that comprising the following steps:

Step 1, storage information task time of collecting IaaS cloud system main frame:

Collect the time of implementation sequence ct of k having completed of task on IaaS cloud system main frame ₁, dt ₂... dt _kstill carrying out on IaaS cloud system main frame of task is the time of implementation sequence et of use ₁, et ₂... .et _m, wherein m is still in the quantity of carrying out of task, k and m are positive integer;

The time of implementation that step 2, prediction are finished the work:

Set the task of having completed successively in the time of implementation sequence on IaaS cloud system main frame, the logarithm step-length average increment of having rejected after exceptional value is disturbed is loinc;

Calculate 1olnc=mean{inc _{i, j}| 0 < i < j≤k, the task that yc (i, j)=0} obtains having completed successively the rejecting in the time of implementation sequence on IaaS cloud system main frame the logarithm step-length average increment that disturbs of exceptional value; Described yc (i, j) is judgement inc _i,jwhether is the function of abnormity point, mean is for gathering the operation of asking arithmetic mean value, inc _i,jrepresent in sequence the and individual record value between the logarithm step-length equivalence increment that forms, described i and j are positive integer; $\mathrm{in}{c}_{i,j}=\frac{{\mathrm{ct}}_j-{\mathrm{ct}}_i}{\log (j-i)};$

$\mathrm{yc}(i,j)=\left\{\begin{array}{c}1\mathrm{if}\frac{\left|\mathrm{in}{c}_{i,j}\right|}{\mathrm{mn}}>a+1,\left|\mathrm{in}{c}_{i,j}\right|>\max \left\{\right|\mathrm{in}{c}_{t,t+1}\left|\right|i\&le;t\&le;j\}\\ 0\mathrm{else}\end{array}\right.;$

Wherein a is given parameter, 0.1≤a≤0.5;

Mn value is: mn=mean{|inc _u,v|| 0<u<v≤k};

The time of implementation that setting is finished the work is nxt: calculate

$\mathrm{nxt}=\left\{\begin{array}{c}\mathrm{mean}\{c{t}_i+\mathrm{loinc}\&times;e^{k-i}|0<i\&le;k\}\mathrm{if}(\mathrm{loinc}+{\mathrm{ct}}_i)>\max \{{\mathrm{et}}_j|0<j\&le;m\}\\ \min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}+\underset{0<i\&le;k,({\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1})<\min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}}{\mathrm{\&Sigma;}}{\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1}\end{array}\right.$

Obtained the time of implementation of next task;

Step 3, calculating dynamic rate expansion controlled quentity controlled variable and control stand-by period:

Described calculating dynamic rate expansion controlled quentity controlled variable is carried out according to the following steps:

Setting following load breach value is qkz, calculates $\mathrm{qkz}=\underset{0<j\&le;m,{\mathrm{et}}_j>\mathrm{nxt}}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})$ Obtain following load breach value;

Read the current residing speed class L of processor and the equivalent implementation rate exe (L) corresponding to this speed class of IaaS cloud system main frame,

If IaaS cloud system main frame has n processor, its highest speed class is g level, and L, n and g are positive integer;

Setting dynamic rate expansion controlled quentity controlled variable is kzl,

Calculate $\mathrm{kzl}=\left\{\begin{array}{c}-1\mathrm{ifqkz}=0\\ \min \left\{i\right|n\&GreaterEqual;\frac{\mathrm{qkz}}{\mathrm{exe}(l+i)-\mathrm{exe}{\left(i\right)}^{\&prime;}}0\&le;i\&le;g-l\}\mathrm{else}\end{array}\right.$ Obtain dynamic rate expansion controlled quentity controlled variable;

Described calculating is controlled the stand-by period and is carried out according to the following steps:

According to the time of implementation sequence of completing the time of implementation of next task and still carrying out on IaaS cloud system main frame of task use, determine the stand-by period of once controlling, the stand-by period of setup control is dsj;

Calculate $\mathrm{dsj}=\left\{\begin{array}{c}\mathrm{wcif}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})<0\mathrm{or}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})>0\\ \mathrm{\&beta;}\&times;\mathrm{wcelse}\end{array}\right.,$ The stand-by period that obtains next time controlling; Wc is the time in control interval that system initially provides, wc>0,1 < β < 10;

Step 4, IaaS cloud system main frame, according to the stand-by period dsj of described control and dynamic rate expansion controlled quentity controlled variable kzl, are controlled the operating rate of IaaS cloud system main frame in the stand-by period section of controlling.

2. IaaS cloud system operating rate dynamic control method as claimed in claim 1, is characterized in that: in the described stand-by period section controlling, the operating rate of IaaS cloud system main frame is controlled according to the following steps and carried out:

If dynamic rate expansion controlled quentity controlled variable kzl is 0, in the time period that is dsj in length, keep the original operating rate of main frame; If kzl value is-1, and main frame is current does not operate in minimum speed grade, operating rate is lowered to one-level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is-1, and main frame is currently operating in minimum speed grade, keeps existing speed in the time period that is dsj in length; If kzl value is greater than zero, and main frame current speed grade L adds and is less than or equal to the speed class value g of high permission of main frame after kzl, at once operating rate raised to kzl level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is greater than zero, and main frame current speed grade L adds and is greater than the speed class value of high permission of main frame after kzl, will on operating rate, be transferred to maximum speed grade at once, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value equals zero, in the time period that is dsj in length, keep existing speed;

Then return to execution step 1.

3. an IaaS cloud system operating rate device for controlling dynamically, is characterized in that: comprise load monitoring module (1), control decision module (2) and main frame speed control module (3); The output of described load monitoring module (1) connects the input of described control decision module (2), and the output of described control decision module (2) connects the input of described main frame speed control module (3);

Described load monitoring module (1) is for obtaining storage information task time of IaaS cloud system main frame;

Described control decision module (2) is for having predicted the time of implementation of next task, and calculates dynamic rate expansion controlled quentity controlled variable and carry out the control stand-by period of task next time;

Described main frame speed control module (3) is for controlling the operating rate of IaaS cloud system main frame.

4. IaaS cloud system operating rate device for controlling dynamically as claimed in claim 3, is characterized in that: described load monitoring module (1) is collected the time of implementation sequence ct of k having completed of task on IaaS cloud system main frame ₁, ct ₂... ct _kstill carrying out on IaaS cloud system main frame of task is the time of implementation sequence et of use ₁, et ₂... et _m, wherein m is still in the quantity of carrying out of task, k and m are positive integer.

5. the IaaS cloud system operating rate device for controlling dynamically as described in claim 3 or 4, is characterized in that: described control decision module (2) comprises data pretreatment unit (201), controlled quentity controlled variable computing unit (202) and controls decision package on opportunity (203); The first output of described load monitoring module (1) connects the input of described data pretreatment unit (201), the second output of described load monitoring module (1) connects the first input end of described controlled quentity controlled variable computing unit (202), and the 3rd output of described load monitoring module (1) connects the first input end of described control decision package on opportunity (203); The first output of described data pretreatment unit (201) connects the second input of described controlled quentity controlled variable computing unit (202), and the second output of described data pretreatment unit (201) connects the second input of described control decision package on opportunity (203); The output of described controlled quentity controlled variable computing unit (202) connects the first input end of described main frame speed control module (3), and the output of described control decision package on opportunity (203) connects the second input of described main frame speed control module (3);

Described data pretreatment unit (201) is for having predicted the time of implementation of next task;

Described controlled quentity controlled variable computing unit (202) is for calculating dynamic rate expansion controlled quentity controlled variable;

Described control decision package on opportunity (203) is carried out the control stand-by period of task next time for calculating;

Described data pretreatment unit (201) is set the task of having completed successively in the time of implementation sequence on IaaS cloud system main frame, and the logarithm step-length average increment of having rejected after exceptional value is disturbed is loinc;

Calculate loinc=mean{inc _{i, j}| 0 < i < j≤k, the task that yc (i, j)=0} obtains having completed successively the rejecting in the time of implementation sequence on IaaS cloud system main frame the logarithm step-length average increment that disturbs of exceptional value; Described yc (i, j) is judgement inc _{i, j}whether is the function of abnormity point, mean is for gathering the operation of asking arithmetic mean value, inc _{i, j}represent in sequence the and individual record value between the logarithm step-length equivalence increment that forms, described i and j are positive integer; $\mathrm{in}{c}_{i,j}=\frac{{\mathrm{ct}}_j-{\mathrm{ct}}_i}{\log (j-i)};$

$\mathrm{yc}(i,j)=\left\{\begin{array}{c}1\mathrm{if}\frac{\left|\mathrm{in}{c}_{i,j}\right|}{\mathrm{mn}}>a+1,\left|\mathrm{in}{c}_{i,j}\right|>\max \left\{\right|\mathrm{in}{c}_{t,t+1}\left|\right|i\&le;t\&le;j\}\\ 0\mathrm{else}\end{array}\right.;$

Wherein a is given parameter, 0.1≤a≤0.5;

Mn value is: mn=mean{|inc _{u, v}|| 0 < u < v≤k};

The time of implementation that setting is finished the work is nxt: described data pretreatment unit (201) calculates

$\mathrm{nxt}=\left\{\begin{array}{c}\mathrm{mean}\{c{t}_i+\mathrm{loinc}\&times;e^{k-i}|0<i\&le;k\}\mathrm{if}(\mathrm{loinc}+{\mathrm{ct}}_i)>\max \{{\mathrm{et}}_j|0<j\&le;m\}\\ \min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}+\underset{0<i\&le;k,({\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1})<\min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}}{\mathrm{\&Sigma;}}{\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1}\end{array}\right.$

The time of implementation nxt that obtains finishing the work; Described data pretreatment unit (201) is issued nxt value the controlled quentity controlled variable computing unit (202) in control decision module (2) and controls decision package on opportunity (203);

It is qkz that described controlled quentity controlled variable computing unit (202) is set following load breach value,

Described controlled quentity controlled variable computing unit (202) calculates $\mathrm{qkz}=\underset{0<j\&le;m,{\mathrm{et}}_j>\mathrm{nxt}}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})$ Obtain following load breach value;

Described controlled quentity controlled variable computing unit (202) reads the current residing speed class L of processor and the equivalent implementation rate exe (L) corresponding to this speed class of IaaS cloud system main frame, if IaaS cloud system main frame has n processor, its the highest speed class is g level, and L, n and g are positive integer;

Setting dynamic rate expansion controlled quentity controlled variable is kzl,

By calculating $\mathrm{kzl}=\left\{\begin{array}{c}-1\mathrm{ifqkz}=0\\ \min \left\{i\right|n\&GreaterEqual;\frac{\mathrm{qkz}}{\mathrm{exe}(l+i)-\mathrm{exe}{\left(i\right)}^{\&prime;}}0\&le;i\&le;g-l\}\mathrm{else}\end{array}\right.$ Described controlled quentity controlled variable computing unit (202) obtains dynamic rate expansion controlled quentity controlled variable; Described controlled quentity controlled variable computing unit (202) sends to described main frame speed control module (3) by kzl value;

Described control decision package on opportunity (203) determines the stand-by period of once controlling according to the time of implementation sequence of completing the time of implementation of next task and still carrying out on IaaS cloud system main frame of task use, the stand-by period of setup control is dsj;

By calculating $\mathrm{dsj}=\left\{\begin{array}{c}\mathrm{wcif}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})<0\mathrm{or}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})>0\\ \mathrm{\&beta;}\&times;\mathrm{wcelse}\end{array}\right.,$ The stand-by period that described control decision package on opportunity (203) obtains controlling next time; Wc is the time in control interval that system initially provides, wc>0,1 < β < 10; Described control decision package on opportunity (203) is issued described main frame speed control module (3) by the stand-by period dsj controlling next time.

6. IaaS cloud system operating rate device for controlling dynamically as claimed in claim 5, it is characterized in that: described main frame speed control module (3), according to the stand-by period dsj of described control and dynamic rate expansion controlled quentity controlled variable kzl, is controlled the operating rate of IaaS cloud system main frame in the stand-by period section of controlling:

If dynamic rate expansion controlled quentity controlled variable kzl is 0, in the time period that is dsj in length, keep the original operating rate of main frame; If kzl value is-1, and main frame is current does not operate in minimum speed grade, operating rate is lowered to one-level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is-1, and main frame is currently operating in minimum speed grade, keeps existing speed in the time period that is dsj in length; If kzl value is greater than zero, and main frame current speed grade L adds and is less than or equal to the speed class value g of high permission of main frame after kzl, at once operating rate raised to kal level, and the speed after keeping in the time period that is dsj in length adjusting; If kal value is greater than zero, and main frame current speed grade L adds and is greater than the speed class value of high permission of main frame after kzl, will on operating rate, be transferred to maximum speed grade at once, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value equals zero, in the time period that is dsj in length, keep existing speed.

7. an IaaS cloud system operating rate kinetic-control system, comprises and it is characterized in that the server (4) of IaaS cloud system: in the server of described IaaS cloud system (4), be provided with IaaS cloud system operating rate device for controlling dynamically (5); Described IaaS cloud system operating rate device for controlling dynamically (5) comprises load monitoring module (1), control decision module (2) and main frame speed control module (3); Comprise data pretreatment unit (201), controlled quentity controlled variable computing unit (202) and control decision package on opportunity (203); The first output of described load monitoring module (1) connects the input of described data pretreatment unit (201), the second output of described load monitoring module (1) connects the first input end of described controlled quentity controlled variable computing unit (202), and the 3rd output of described load monitoring module (1) connects the first input end of described control decision package on opportunity (203); The first output of described data pretreatment unit (201) connects the second input of described controlled quentity controlled variable computing unit (202), and the second output of described data pretreatment unit (201) connects the second input of described control decision package on opportunity (203); The output of described controlled quentity controlled variable computing unit (202) connects the first input end of described main frame speed control module (3), and the output of described control decision package on opportunity (203) connects the second input of described main frame speed control module (3);

Described load monitoring module (1) is for obtaining storage information task time of IaaS cloud system main frame;

Described data pretreatment unit (201) is for having predicted the time of implementation of next task;

Described controlled quentity controlled variable computing unit (202) is for calculating dynamic rate expansion controlled quentity controlled variable;

Described control decision package on opportunity (203) is carried out the control stand-by period of task next time for calculating;

Described main frame speed control module (3) is for controlling the operating rate of IaaS cloud system main frame;

Described load monitoring module (1) is collected the time of implementation sequence ct of L having completed of task on IaaS cloud system main frame ₁, ct ₂... ct _kstill carrying out on IaaS cloud system main frame of task is the time of implementation sequence et of use ₁, et ₂... .et _m, wherein m is still in the quantity of carrying out of task, k and m are positive integer;

Described data pretreatment unit (201) is set the task of having completed successively in the time of implementation sequence on IaaS cloud system main frame, and the logarithm step-length average increment of having rejected after exceptional value is disturbed is loinc;

Calculate loinc=mean{inc _{i, j}| 0 < i < j≤k, yc (i, j)=0) task of obtaining having completed successively the rejecting in the time of implementation sequence on IaaS cloud system main frame the logarithm step-length average increment that disturbs of exceptional value; Described yc (i, j) is judgement inc _{i, j}whether is the function of abnormity point, mean is for gathering the operation of asking arithmetic mean value, inc _{i, j}the logarithm step-length equivalence increment forming between i and j record value in expression sequence, described i and j are positive integer; $\mathrm{in}{c}_{i,j}=\frac{{\mathrm{ct}}_j-{\mathrm{ct}}_i}{\log (j-i)};$

$\mathrm{yc}(i,j)=\left\{\begin{array}{c}1\mathrm{if}\frac{\left|\mathrm{in}{c}_{i,j}\right|}{\mathrm{mn}}>a+1,\left|\mathrm{in}{c}_{i,j}\right|>\max \left\{\right|\mathrm{in}{c}_{t,t+1}\left|\right|i\&le;t\&le;j\}\\ 0\mathrm{else}\end{array}\right.;$

Wherein a is given parameter, 0.1≤a≤0.5;

Mn value is: mn=mean{|inc _{u, v}|| 0 < u < v≤k};

The time of implementation that setting is finished the work is nxt: described data pretreatment unit (201) calculates

$\mathrm{nxt}=\left\{\begin{array}{c}\mathrm{mean}\{c{t}_i+\mathrm{loinc}\&times;e^{k-i}|0<i\&le;k\}\mathrm{if}(\mathrm{loinc}+{\mathrm{ct}}_i)>\max \{{\mathrm{et}}_j|0<j\&le;m\}\\ \min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}+\underset{0<i\&le;k,({\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1})<\min \left\{{\mathrm{et}}_j\right|0<j\&le;m\}}{\mathrm{\&Sigma;}}{\mathrm{ct}}_i+\mathrm{loinc}\&times;e^{k-i+1}\end{array}\right.$

The time of implementation nxt that obtains finishing the work; Described data pretreatment unit (201) is issued nxt value the controlled quentity controlled variable computing unit (202) in control decision module (2) and controls decision package on opportunity (203);

It is qkz that described controlled quentity controlled variable computing unit (202) is set following load breach value,

Described controlled quentity controlled variable computing unit (202) calculates $\mathrm{qkz}=\underset{0<j\&le;m,{\mathrm{et}}_j>\mathrm{nxt}}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})$ Obtain following load breach value;

Described controlled quentity controlled variable computing unit (202) reads the current residing speed class L of processor and the equivalent implementation rate exe (L) corresponding to this speed class of IaaS cloud system main frame, if IaaS cloud system main frame has n processor, its the highest speed class is g level, and L, n and g are positive integer;

Setting dynamic rate expansion controlled quentity controlled variable is kzl,

By calculating $\mathrm{kzl}=\left\{\begin{array}{c}-1\mathrm{ifqkz}=0\\ \min \left\{i\right|n\&GreaterEqual;\frac{\mathrm{qkz}}{\mathrm{exe}(l+i)-\mathrm{exe}{\left(i\right)}^{\&prime;}}0\&le;i\&le;g-l\}\mathrm{else}\end{array}\right.$ Described controlled quentity controlled variable computing unit (202) obtains dynamic rate expansion controlled quentity controlled variable; Described controlled quentity controlled variable computing unit (202) sends to described main frame speed control module (3) by kzl value;

Described control decision package on opportunity (203) determines the stand-by period of once controlling according to the time of implementation sequence of completing the time of implementation of next task and still carrying out on IaaS cloud system main frame of task use, the stand-by period of setup control is dsj;

By calculating $\mathrm{dsj}=\left\{\begin{array}{c}\mathrm{wcif}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})<0\mathrm{or}\underset{0<j\&le;m}{\mathrm{\&Sigma;}}({\mathrm{et}}_j-\mathrm{nxt})>0\\ \mathrm{\&beta;}\&times;\mathrm{wcelse}\end{array}\right.,$ The stand-by period that described control decision package on opportunity (203) obtains controlling next time; Wc is the time in control interval that system initially provides, wc>0,1 < β < 10; Described control decision package on opportunity (203) is issued described main frame speed control module (3) by the stand-by period dsj controlling next time;

Described main frame speed control module (3) sending controling instruction is to IaaS cloud system main frame, to expand controlled quentity controlled variable kzl according to the stand-by period dsj of described control and dynamic rate, in the stand-by period section of controlling, the operating rate of IaaS cloud system main frame is controlled:

If dynamic rate expansion controlled quentity controlled variable kzl is 0, in the time period that is dsj in length, keep the original operating rate of main frame; If kzl value is-1, and main frame is current does not operate in minimum speed grade, operating rate is lowered to one-level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is-1, and main frame is currently operating in minimum speed grade, keeps existing speed in the time period that is dsj in length; If kzl value is greater than zero, and main frame current speed grade L adds and is less than or equal to the speed class value g of high permission of main frame after kzl, at once operating rate raised to kzl level, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value is greater than zero, and main frame current speed grade L adds and is greater than the speed class value of high permission of main frame after kzl, will on operating rate, be transferred to maximum speed grade at once, and the speed after keeping in the time period that is dsj in length adjusting; If kzl value equals zero, in the time period that is dsj in length, keep existing speed.

Images