CN103745092B  Evaluation method and evaluation system for utilization ratio of server  Google Patents
Evaluation method and evaluation system for utilization ratio of server Download PDFInfo
 Publication number
 CN103745092B CN103745092B CN201310722907.4A CN201310722907A CN103745092B CN 103745092 B CN103745092 B CN 103745092B CN 201310722907 A CN201310722907 A CN 201310722907A CN 103745092 B CN103745092 B CN 103745092B
 Authority
 CN
 China
 Prior art keywords
 utilization rate
 correlation properties
 index
 described
 server
 Prior art date
Links
Abstract
Description
Technical field
The present invention relates to network technology, particularly relate to appraisal procedure and the system of a kind of server utilization.
Background technology
Along with the development of network technology, internet, applications is increasingly sophisticated.For LargeScale Interconnected net application system, In order to understand the operation conditions of system, generally require the utilization rate to server and be estimated.
The appraisal procedure of existing a kind of server utilization, it is common that run time institute's record by server Various server resources take historical data (such as CPU usage, memory storage situation, the utilization of resources Situation etc.) estimate server utilization.Such as, server resource is taken historical data and carry out the time Sequence data is analyzed, and such as, the server resource recorded is taken historical data and moves averagely, add The analyses such as power rolling average, exponential smoothing, Trend index smooth, after obtaining analysis result, according to service The analysis result of device resource occupation historical data, takies data to server resource afterwards and is predicted, And take data according to the server resource of prediction, artificially evaluate the utilization rate of the server of current system. Such as, compared with CPU usage before, it was predicted that CPU usage add, then can assess work as The server utilization of front system improves, and the utilization rate of artificial subjective evaluation server.
But, it was found by the inventors of the present invention that the appraisal procedure of existing server utilization, to service The assessment result of the utilization rate of device is inaccurate；It is true that server resource takies data except being serviced The impact of device utilization rate, is also affected by otherwise factor.Such as, the increase of CPU usage can Can be to be caused by the situation causing efficiency to reduce that breaks down of some product in system, and nonserving The increase of device utilization rate is caused；Even, it may be possible to institute in the case of the reduction of server utilization Formed.Therefore, take, according to the server resource of prediction, the standard that data carry out the utilization rate of evaluating server Exactness is the highest.And, the utilization rate of the artificial evaluating server of data is taken according to the server resource of prediction Mode, the accuracy also making assessment result is relatively low.
Therefore, it is necessary to the method providing the utilization rate of a kind of evaluating server, improve server utilization Assessment accuracy.
Summary of the invention
Embodiments provide appraisal procedure and the system of a kind of server utilization, in order to improve clothes The assessment accuracy of business device utilization rate.
According to an aspect of the invention, it is provided the appraisal procedure of a kind of server utilization, including:
According to the achievement data for each index under the utilization rate correlation properties of described server, and respectively The weight of index, that determines described server utilizes assessed value；
Utilize assessed value according to described server, determine the utilization rate of described server；
Wherein, the weight of described index be according to each utilization rate correlation properties between significance level fiducial value, And significance level fiducial value between each index under same utilization rate correlation properties determines.
It is preferred that described basis is for the index of each index under the utilization rate correlation properties of described server Data, and the weight of each index, that determines described server utilizes assessed value, specifically includes:
Assessed value Q that utilizes of server is calculated according to equation below 1:
Wherein, q_{i}For the achievement data of the ith index under the utilization rate correlation properties of described server, r_{i}For The weight of ith index；I value is the natural number of 1～n, and n is the utilization rate correlation properties of described server Under index sum.
It is preferred that described determine described server utilize assessed value before, also include:
According to the significance level fiducial value between each utilization rate correlation properties, calculate the relevant spy of each utilization rate The weight of property；
For each utilization rate correlation properties, according to the weight between each index under these utilization rate correlation properties Want degree fiducial value, calculate in the characteristic of each index after weight, under these utilization rate correlation properties Each index, by the multiplied by weight of weight in the characteristic of this index Yu these utilization rate correlation properties, is somebody's turn to do The weight of index.
It is preferred that described according to the significance level fiducial value between each utilization rate correlation properties, calculate each The weight of utilization rate correlation properties, particularly as follows:
According to the significance level fiducial value between each utilization rate correlation properties, determine between correlation properties important Degree comparator matrix A, the ith row in described A, jth column element a_{ij}Represent ith utilization rate correlation properties Significance level fiducial value relative to jth utilization rate correlation properties；Wherein, i and j is value 1～m Natural number, m is the sum of the utilization rate correlation properties of described server；
After calculating the characteristic vector W of described A, by the ith row element w in described W_{i}Utilize as ith The weight of rate correlation properties.
It is preferred that described according to the significance level fiducial value between each index under these utilization rate correlation properties, Calculate weight in the characteristic of each index, particularly as follows:
According to the significance level fiducial value between each index under these utilization rate correlation properties, determine this profit With significance level comparator matrix B between the index under rate correlation properties；The ith row, jth row in described B are first Element b_{ij}Represent that the ith index under these utilization rate correlation properties compares relative to the significance level of jth index Value；Wherein, i and j is the natural number of value 1～l, and l is each index under these utilization rate correlation properties Sum；
After calculating characteristic vector U of described B, by the ith row element u in described U_{i}As this utilization rate Weight in the characteristic of the ith index under correlation properties.
It is preferred that described according to the significance level fiducial value between each utilization rate correlation properties, calculate Before the weight of each utilization rate correlation properties, also include:
Described A is carried out consistency check；If described A is not by consistency check, then adjust each utilization rate Significance level fiducial value between correlation properties, until described A passes through consistency check；And
Described for each utilization rate correlation properties, according to each index under these utilization rate correlation properties it Between significance level fiducial value, calculate in the characteristic of each index before weight, also include:
For each utilization rate correlation properties, to significance level ratio between the index under these utilization rate correlation properties Relatively matrix carries out consistency check；If significance level comparator matrix is not by consistency check between described index, Then adjust the significance level fiducial value between each index under these utilization rate correlation properties, until described index Between significance level comparator matrix pass through consistency check.
Utilize assessed value it is preferred that described according to described server, determine the utilization of described server Rate, particularly as follows:
Search mapping relations table set in advance, described mapping relations table have recorded and utilize assessed value and profit By the mapping relations between rate；
Find out from described mapping relations table with described server utilize utilization rate corresponding to assessed value, And the utilization rate found out is defined as the utilization rate of described server.
It is preferred that described utilization rate correlation properties specifically include: product attribute, technical characteristic, operation spy Property；And
Index under described product attribute includes: product and the matching degree of existing product, product significance level, Product Maturity；
Index under described technical characteristic includes: system availability, system scalability, research and development speed；
Index under described operational characteristic includes: O&M cost, moving costs, server procurement cycle.
According to another aspect of the present invention, additionally provide the assessment system of a kind of server utilization, bag Include:
Achievement data acquisition module, each for obtain under the utilization rate correlation properties of described server The achievement data of index；
Server assessed value computing module, for each index obtained according to described achievement data acquisition module Achievement data, and the weight of each index, that determines described server utilizes assessed value；
Server utilization determines module, for determine according to described server assessed value computing module Utilize assessed value, determine the utilization rate of described server.
It is preferred that the assessment system of described server utilization also includes:
Index weights determines module, for according to the weight between each utilization rate correlation properties of described server Want the significance level fiducial value between each index under degree fiducial value, and same utilization rate correlation properties Determine the weight of each index.
In the technical scheme of the embodiment of the present invention, according to the significance level ratio between each utilization rate correlation properties Relatively it is worth, and the significance level fiducial value between each index under same utilization rate correlation properties, determine The weight of each index, and according to the weight of each index, determine the assessed value that utilizes of server, Jin Erke To determine the utilization rate of server.Compare existing according to server runtime server resource occupation data The server utilization assessed, technical scheme is according to each index affecting server utilization Carrying out evaluating server utilization rate, accuracy is higher；And, compared to existing artificial evaluating server The method of utilization rate, technical solution of the present invention is come really according to the weight of each index affecting server utilization The method determining the utilization rate of server, further increases accuracy.
Accompanying drawing explanation
Fig. 1 a is the flow chart of the determination method of the weight of each index of the embodiment of the present invention；
Fig. 1 b is the flow chart of the consistency check of the comparator matrix of the embodiment of the present invention；
Fig. 2 is the flow chart of the appraisal procedure of the server utilization of the embodiment of the present invention；
Fig. 3 is the internal structure schematic diagram of the assessment system of the server utilization of the embodiment of the present invention；
Fig. 4 is the internal structure schematic diagram that the index weights of the embodiment of the present invention determines module.
Detailed description of the invention
For making the purpose of the present invention, technical scheme and advantage clearer, develop simultaneously referring to the drawings Going out preferred embodiment, the present invention is described in more detail.However, it is necessary to explanation, in description The many details listed be only used to make reader one or more aspects of the present invention are had one thorough Understand, the aspects of the invention can also be realized even without these specific details.
The term such as " module " used in this application, " system " is intended to include the entity relevant to computer, Such as but not limited to hardware, firmware, combination thereof, software or executory software.Such as, mould Block it may be that it is not limited to: on processor run process, processor, object, journey can be performed Sequence, the thread of execution, program and/or computer.For example, application program calculating equipment run Can be module with this calculating equipment.One or more modules may be located at an executory process and/ Or in thread.
It was found by the inventors of the present invention that affect a lot of because have of the utilization rate of server in actual applications, Such as: product and the matching degree of existing product, product significance level, product Maturity, system availability, System scalability, research and development speed, O&M cost, moving costs, server procurement cycle etc..Therefore, The present inventor takies number it is considered that compare existing server resource when running according to server According to the server utilization assessed, carry out evaluation services according to the principal element of the utilization rate affecting server Device utilization rate, can improve the assessment accuracy of server utilization.
The present inventor is it is further contemplated that affecting between each factor of the utilization rate of server is phase Impact mutually, mutually restriction, therefore, it can each factor of the utilization rate affecting server is carried out level Build, and for each level, the square that structure is made up of the significance level fiducial value between two factors Battle array, thus calculates the weight of each factor；And according to the weight of each factor, determine the utilization of server Assessed value, and then determine the utilization rate of server.Compared to existing artificial evaluating server utilization rate Method, technical solution of the present invention determines clothes according to the weight of each factor of the utilization rate affecting server The utilization rate of business device, substantially increases the accuracy of assessment result.
Specifically, the abovementioned factor of the utilization rate of server can be affected, according to utilization rate correlation properties Reasonably dividing, the correlation properties of the utilization rate affecting server such as marked off include: product Characteristic, technical characteristic and operational characteristic.Then, will be able to produce according to abovementioned utilization rate correlation properties Product and the matching degree of existing product, product significance level, product Maturity are divided into the finger under product attribute Mark；The system availability of system, system scalability, research and development speed are divided into the finger under technical characteristic Mark；O&M cost, moving costs, server procurement cycle are divided into the index under operational characteristic.Real Border application in, each utilization rate correlation properties can as the rule layer influence factor of the utilization rate of server, Each index under these utilization rate correlation properties then can as the indicator layer of the utilization rate of server affect because of Element.In this manner it is possible to according to the significance level fiducial value between each utilization rate correlation properties, and same The significance level fiducial value between each index under utilization rate correlation properties, after determining the weight of each index, According to the weight of each index, determine the assessed value that utilizes of server, and then can determine that server Utilization rate.
Describe technical scheme below in conjunction with the accompanying drawings in detail.
In the embodiment of the present invention, before carrying out the assessment of server utilization, each profit can be preset By each index under the significance level fiducial value between rate correlation properties, and same utilization rate correlation properties Between significance level fiducial value.Specifically, can be according to the significance level ratio of two factors preset Relatively result and the corresponding relation quantified between ratio, basis is preto the influence degree institute of server utilization Significance level comparative result between each utilization rate correlation properties first arranged, is quantified as the relevant spy of utilization rate Significance level fiducial value between property, i.e. one utilization rate correlation properties are correlated with relative to another utilization rate The significance level fiducial value of characteristic.And by according to the influence degree of server utilization preset Significance level comparative result between each index under same utilization rate correlation properties, is quantified as same utilization rate The significance level fiducial value between each index under correlation properties, i.e. one index is relative to another index Significance level fiducial value.
Such as, the significance level comparative result of two factors preset and the corresponding pass quantified between ratio In system, if the significance level comparative result of two factors is of equal importance, then corresponding quantization ratio is 1；If The significance level comparative result of two factors is that one of them is important, then corresponding quantization ratio is 3；If Significance level is between of equal importance and important, then corresponding quantization ratio is 2, by that analogy, Concrete as shown in table 1.
Table 1
Therefore, it can each factor of the utilization rate affecting server is compared twobytwo, according in advance These two the significance level comparative results between factor X and Y arranged, in conjunction with the correspondence shown in table 1 Relation, can obtain the quantization ratio g of significance level comparative result between this X and Y, and g value is The natural number of 1～9.If according to the X that the influence degree of server utilization is preset than Y more Important, then it is set to quantify ratio g, by Y relative to X by the X significance level fiducial value relative to Y Significance level fiducial value be set to 1/g；Correspondingly, if Y is more even more important than X, then by X relative to The significance level fiducial value of Y is set to 1/g, and accordingly, the Y significance level fiducial value relative to X is arranged For g.
Such as, each utilization rate phase according to table 2, the influence degree of server utilization preset Close the significance level comparative result between characteristic.
Table 2
Therefore, in conjunction with table 1 and table 2, the significance level that can arrange between each utilization rate correlation properties compares Value, the most as shown in table 3:
Table 3
In like manner, according to said method, for each utilization rate correlation properties, according to the relevant spy of this utilization rate Significance level comparative result between each index under property, the significance level that can arrange between each index compares Value, specifically as shown in table 4, table 5 and table 6.
Table 4
Table 5
Table 6
Further, in the significance level fiducial value arranged between each utilization rate correlation properties and same profit After the significance level fiducial value between each index under rate correlation properties, it is also possible to according to each utilization rate Between each index under significance level fiducial value between correlation properties, and same utilization rate correlation properties Significance level fiducial value, determine the weight of each index under the utilization rate correlation properties of server, its Concrete method flow, as shown in Figure 1a, may include steps of:
S101: according to the significance level fiducial value between each utilization rate correlation properties, calculate each utilization rate The weight of correlation properties.
Specifically, the significance level fiducial value between each utilization rate correlation properties, correlation properties can be formed Between significance level comparator matrix A.Significance level ratio between therefore, it can according to each utilization rate correlation properties Relatively it is worth, determines A.The ith row in A, jth column element a_{ij}Represent ith utilization rate correlation properties phase Significance level fiducial value for jth utilization rate correlation properties；Wherein, i and j is value 1～m Natural number, m is the sum of the utilization rate correlation properties of server.Wherein, ith utilization rate correlation properties Relative to the significance level fiducial value of jth utilization rate correlation properties, with jth utilization rate correlation properties phase Significance level fiducial value for ith utilization rate correlation properties can be reciprocal each other.
After calculating the characteristic vector W of A, can be by the ith row element w in W_{i}As ith utilization rate The weight of correlation properties.
Such as, according to the significance level fiducial value between each utilization rate correlation properties shown in table 3, permissible Determine A particularly as follows:
Characteristic vector W about A can use those skilled in the art commonly use and method or root method to count Calculate.By with method as a example by, the every string in the A that can first will determine is normalized, and obtains normalization Matrix S.
Wherein, the ith row in S, jth column element s_{ij}Meet:J value be 1～m it Between natural number.So, according to the A of the example above, returning between each utilization rate correlation properties calculated One change matrix S particularly as follows:
Then, to the every a line in S, go summation, obtain quasicharacteristic vector C.
Wherein, the ith row element c in C_{i}Meet:I value is the natural number of 1～m.So, According to the S of the example above, quasicharacteristic vector C between each utilization rate correlation properties obtained particularly as follows:
Finally, to the every string in C, it is normalized, obtains the characteristic vector W of A.
Wherein, the ith row element wi in W meets:I value is the natural number of 1～m. So, according to the C of the example above, the W obtained particularly as follows:
According to the W obtained, can be by the ith row element w in W_{i}As ith utilization rate correlation properties Weight.So, according to the W of the example above, the weight of each utilization rate correlation properties determined particularly as follows: The weight of product attribute p is about 0.54, and the weight of technical characteristic t is about 0.30, and the power of operational characteristic o Weigh about is 0.16.
S102: for each utilization rate correlation properties, according to each index under these utilization rate correlation properties it Between significance level fiducial value, calculate weight in the characteristic of each index.
Specifically, according to the significance level fiducial value between each index under these utilization rate correlation properties, really Make significance level comparator matrix B between index；The ith row in B, jth column element b_{ij}Represent this utilization rate Ith index under correlation properties is relative to the significance level fiducial value of jth index；Wherein, i and j is equal For the natural number of value 1～l, l is the sum of each index under these utilization rate correlation properties.Wherein, ith Index is relative to the significance level fiducial value of jth index, with jth index relative to ith index Significance level fiducial value can be reciprocal each other.
After calculating characteristic vector U of B, by the ith row element u in U_{i}As the relevant spy of this utilization rate Weight in the characteristic of the ith index under property.
Such as, for product attribute p, can according to each index under the product attribute p shown by table 4 it Between significance level fiducial value, significance level comparator matrix between the index under the product attribute p determined B(p_{1},p_{2},p_{3}) it is:
Wherein, p_{1}、p_{2}、p_{3}Represent the index under product attribute p respectively: product and existing product Degree of joining, product significance level, product Maturity.
For technical characteristic t, significance level comparator matrix between the index under technical characteristic t determined B(t_{1},t_{2},t_{3}) it is:
Wherein, t_{1}、t_{2}、t_{3}Index under presentation technology characteristic t respectively: system availability, system scalability Property, research and development speed.
For operational characteristic o, significance level comparator matrix between the index under the operational characteristic o determined B(o_{1},o_{2},o_{3}) it is:
Wherein, o_{1}、o_{2}、o_{3}Represent the index under operational characteristic o respectively: O&M cost, moving costs, Server procurement cycle.
Then, the method that the characteristic vector W of abovementioned A can be used, calculate characteristic vector U of B.Example As, the B (p under product attribute p_{1},p_{2},p_{3}) characteristic vector U (p_{1},p_{2},p_{3}) particularly as follows:
B (t under technical characteristic t_{1},t_{2},t_{3}) characteristic vector U (t_{1},t_{2},t_{3}) particularly as follows:
B (o under operational characteristic o_{1},o_{2},o_{3}) characteristic vector U (o_{1},o_{2},o_{3}) particularly as follows:
So, for product attribute p, according to U (p_{1},p_{2},p_{3}) respectively referring under the product attribute p that determined Weight in target characteristic is particularly as follows: p_{1}Characteristic in weight be 0.25, p_{2}Characteristic in weight be 0.59, p_{3}Characteristic in weight be 0.16.
For technical characteristic t, according to characteristic vector U (t_{1},t_{2},t_{3}) each index under technical characteristic t determined Characteristic in weight particularly as follows: t_{1}Characteristic in weight be 0.30, t_{2}Characteristic in weight be 0.54, t_{3}'s In characteristic, weight is 0.16.
For operational characteristic o, according to U (o_{1},o_{2},o_{3}) characteristic of each index under the operational characteristic o that determined Interior weight is particularly as follows: o_{1}Characteristic in weight be 0.61, o_{2}Characteristic in weight be 0.27, o_{3}Characteristic in Weight is 0.12.
S103: for each utilization rate correlation properties, for each index under these utilization rate correlation properties, By the multiplied by weight of weight in the characteristic of this index Yu these utilization rate correlation properties, obtain the weight of this index.
Specifically, for each utilization rate correlation properties, this utilization rate step S101 calculated is correlated with The weight of characteristic, the characteristic of each index under these utilization rate correlation properties determined with step S102 respectively Interior multiplied by weight, obtains the weight of each index under these utilization rate correlation properties.
Such as, according to the example above, the weight of each index under product attribute p is particularly as follows: p_{1}Weight be 0.135738336, p_{2}Weight be 0.317388167, p_{3}Weight be 0.085834536.
The weight of each index under technical characteristic t is particularly as follows: t_{1}Weight be 0.088362495, t_{2}Weight It is 0.16021041, t_{3}Weight be 0.048685161.
The weight of each index under operational characteristic o is particularly as follows: o_{1}Weight be 0.099572455, o_{2}Power It is heavily 0.044564475, o_{3}Weight be 0.019643733.
In actual application, owing to the significance level between each utilization rate correlation properties is by people in the art Member presets, but due to the complexity of system and the ambiguity of artificial subjective judgment, presets May there is contradiction in the significance level comparative result between each utilization rate correlation properties, cause according to each profit Between each utilization rate correlation properties determined with the significance level comparative result between rate correlation properties Significance level fiducial value and comparator matrix may deviate concordance, finally cause the service evaluated The accuracy of device utilization rate is the highest.
Accordingly, as a kind of more excellent embodiment, before performing abovementioned steps S101, can be to root According to significance level between the correlation properties that the significance level fiducial value between each utilization rate correlation properties is determined Comparator matrix A carries out consistency check.If A is not by consistency check, then adjust the relevant spy of each utilization rate Significance level fiducial value between property, until A passes through consistency check.So, according to consistent, There are not the clothes that the significance level fiducial value between each utilization rate correlation properties of contradiction is determined between data Business device utilization rate, can improve the assessment accuracy of server utilization.
In actual application, if it is determined that A is not by consistency check, then illustrate between each utilization rate correlation properties Significance level fiducial value there is contradiction.And the significance level fiducial value tool between each utilization rate correlation properties Body is the significance level comparative result according to two factors preset and the corresponding pass quantifying between ratio System, and set by the significance level comparative result between each utilization rate correlation properties.Therefore, specifically Ground, can adjust each profit by adjusting the significance level comparative result between each utilization rate correlation properties With the significance level fiducial value between rate correlation properties, value is set.Then, according to each utilization after adjusting Significance level fiducial value between rate correlation properties, determines significance level ratio between new correlation properties Relatively matrix A, and new A is carried out consistency check；If new A is not by consistency check, then continue Adjust the significance level fiducial value between each utilization rate correlation properties, until A passes through consistency check.
About the method flow of the consistency check of the abovementioned A mentioned, as shown in Figure 1 b, it specifically includes Following steps:
S111: calculate the Maximum characteristic root λ of A_{max}。
Specifically, the characteristic vector W for A is used according to equation below 2, calculates the maximum feature of A Root λ_{max}:
In formula 2, (AW)_{i}For the ith row element in the product of A and W, m is the utilization rate phase of server Close the sum of characteristic.So, according to the significance level ratio between each utilization rate correlation properties shown in table 3 Relatively it is worth, the Maximum characteristic root λ of A can be obtained_{max}Particularly as follows: λ_{max}=3.01。
S112: according to the λ calculated_{max}, determine coincident indicator C.I. of A.
Specifically, can be by λ_{max}Substitute into equation below 3, determine coincident indicator C.I. of A:
In formula 3, m is the matrix exponent number of A, and it is equal to the sum of the utilization rate correlation properties of server. So, at m=3 and Maximum characteristic root λ_{max}When value is 3.01, the C.I. of A particularly as follows:
S113: search between the matrix exponent number of Average Random Consistency Index set in advance and comparator matrix Mapping table, therefrom find out the Average Random Consistency Index R.I. corresponding with the matrix exponent number of A.
Specifically, according to those skilled in the art's Average Random Consistency Index set in advance with compare square Battle array matrix exponent number between mapping table, as shown in table 7, can therefrom find out with A matrix The Average Random Consistency Index that exponent number is corresponding, and Average Random Consistency Index the putting down as A that will obtain All random index R.I..Such as, the matrix exponent number of comparator matrix A is 3, therefore, according to table 7, May determine that R.I. value is 0.58.
Table 7
S114: the ratio of the C.I. determined by step S112 and the R.I. found out by step S113 is made Consistency Ratio C.R. for A.
Such as, according to the Consistency Ratio C.R. of C.I. and R.I. of the example above, A particularly as follows:
S115: the Consistency Ratio C.R. of A is compared with consistency check standard value δ set in advance； If C.R. is less than δ, then judge that A passes through consistency check；Otherwise, it is determined that A does not passes through consistency check.
Such as, the concordance that the Consistency Ratio C.R. of the A of the example above commonly uses less than those skilled in the art Touchstone value 0.01, so, it is possible to determine that A passes through consistency check.
In like manner, in the technical scheme that the present invention provides, it is also possible to before performing abovementioned steps S102, right In each utilization rate correlation properties, to significance level comparator matrix between the index under these utilization rate correlation properties Carry out consistency check；If significance level comparator matrix is not by consistency check between this index, then adjust The significance level fiducial value between each index under these utilization rate correlation properties, until important journey between this index Degree comparator matrix passes through consistency check.Specifically can use significance level comparator matrix A between correlation properties The method carrying out consistency check, enters significance level comparator matrix between the index under utilization rate correlation properties Row consistency check, does not repeats them here.
The weight of each index under the utilization rate correlation properties of the server determined based on said method, this The appraisal procedure flow process of a kind of server utilization that inventive embodiments provides, as in figure 2 it is shown, specifically wrap Include following steps:
S201: according to the achievement data for each index under the utilization rate correlation properties of server, and The weight of each index, that determines server utilizes assessed value.
Specifically, assessed value Q that utilizes of server can be calculated according to equation below 1:
Wherein, q_{i}For the achievement data of the ith index under the utilization rate correlation properties of server, r_{i}It is ith The weight of individual index；I value is the natural number of 1～n, and n is the finger under the utilization rate correlation properties of server Mark sum.Wherein, the achievement data for each index under the utilization rate correlation properties of server can be by Those skilled in the art initially set previously according to experience, as shown in table 8:
Table 8
Therefore, according to the weight of each index under the utilization rate correlation properties of the server predefined out, In conjunction with table 8 and formula 1, it may be determined that the assessed value that utilizes going out server is: 5.31.
S202: utilize assessed value according to the server determined by step S201, determines server Utilization rate.
Specifically, the mapping relations table utilized between assessed value and utilization rate set in advance is searched；From institute State mapping relations table finds out with the server determined utilize utilization rate corresponding to assessed value, and will The utilization rate found out is defined as the utilization rate of server.Wherein, mapping relations table have recorded utilization to comment Mapping relations between valuation and utilization rate, and utilizing assessed value is by ability with the mapping relations of utilization rate Field technique personnel be set previously according to experience, and with to utilize assessed value corresponding in these mapping relations Utilization rate can be specially a numerical value, it is also possible to including: utilization rate upper legitimate value, utilization rate are reasonable Value lower limit, utilization rate reasonable value value up to standard.Such as, the server of setting utilize assessed value and utilization rate Mapping relations can be as shown in table 9.
Such as, utilize assessed value 5.31 according to the server determined, and combine table 9, it may be determined that go out The utilization rate of server specifically includes: utilization rate upper legitimate value is 66%, utilization rate reasonable value lower limit is 61%, utilization rate reasonable value value up to standard is 61%.
Table 9
Appraisal procedure based on abovementioned server utilization, the embodiment of the present invention additionally provides a kind of service The assessment system of device utilization rate, as it is shown on figure 3, include: achievement data acquisition module 301, server are commented Valuation computing module 302 and server utilization determine module 303.
Wherein, achievement data acquisition module 301 is for obtaining under the utilization rate correlation properties of server The achievement data of each index.
Server assessed value computing module 302 respectively refers to for obtain according to achievement data acquisition module 301 Target achievement data, and the weight of each index, that determines server utilizes assessed value.
Server utilization determines that module 303 is for determining according to server assessed value computing module 302 Utilize assessed value, determine the utilization rate of server.
As a kind of more excellent embodiment, the assessment system of the server utilization that the embodiment of the present invention provides System also includes: index weights determines module 304.
Index weights determines important for according between each utilization rate correlation properties of server of module 304 The significance level fiducial value between each index under degree fiducial value, and same utilization rate correlation properties is true The weight of fixed each index.
Specifically, index weights determines module 304, as shown in Figure 4, specifically includes: significance level compares Value acquiring unit 401, characteristic weight calculation unit 402, index weights computing unit 403.
Wherein, significance level fiducial value acquiring unit 401 is for the relevant spy of each utilization rate obtaining server Important between each index under significance level fiducial value between property, and same utilization rate correlation properties Degree fiducial value.
Specifically, significance level fiducial value acquiring unit 401 is according to the important journey of two factors preset Degree comparative result and the corresponding relation quantified between ratio, and important between each utilization rate correlation properties Degree comparative result, obtains the significance level fiducial value between each utilization rate correlation properties；According to setting in advance Corresponding relation between significance level comparative result and the quantization ratio of two factors put, and same utilization The significance level comparative result between each index under rate correlation properties, obtains same utilization rate correlation properties Under each index between significance level fiducial value.
Characteristic weight calculation unit 402 is for according to acquired in significance level fiducial value acquiring unit 401 Significance level fiducial value between each utilization rate correlation properties, calculates the weight of each utilization rate correlation properties.
Index weights computing unit 403 is for for each utilization rate correlation properties, according to significance level ratio The significance level ratio between each index under relatively value these utilization rate correlation properties acquired in acquiring unit 401 Relatively it is worth, calculates in the characteristic of each index after weight, for each index under these utilization rate correlation properties, By relevant to this utilization rate calculated by characteristic weight calculation unit 402 for weight in the characteristic of this index special The multiplied by weight of property, obtains the weight of this index.
In technical scheme, each factor affecting the utilization rate of server is divided into corresponding different Each index under utilization rate correlation properties, and compare according to the significance level between each utilization rate correlation properties Value, and the significance level fiducial value between each index under same utilization rate correlation properties, determine each The weight of index, and according to the weight of each index, determine the assessed value that utilizes of server, so that permissible Determine the utilization rate of server.Compare existing according to server runtime server resource occupation data institute The server utilization of assessment, technical scheme is come according to each index affecting server utilization Evaluating server utilization rate, accuracy is higher；And, compared to the profit of existing artificial evaluating server By the method for rate, technical solution of the present invention determines according to the weight of each index affecting server utilization The method of the utilization rate of server, further increases accuracy.
One of ordinary skill in the art will appreciate that all or part of step realizing in abovedescribed embodiment method The program that can be by completes to instruct relevant hardware, and this program can be stored in a computerreadable Take in storage medium, such as: ROM/RAM, magnetic disc, CD etc..
The above is only the preferred embodiment of the present invention, it is noted that general for the art For logical technical staff, under the premise without departing from the principles of the invention, it is also possible to make some improvement and profit Decorations, these improvements and modifications also should be regarded as protection scope of the present invention.
Claims (9)
Priority Applications (1)
Application Number  Priority Date  Filing Date  Title 

CN201310722907.4A CN103745092B (en)  20131224  20131224  Evaluation method and evaluation system for utilization ratio of server 
Applications Claiming Priority (1)
Application Number  Priority Date  Filing Date  Title 

CN201310722907.4A CN103745092B (en)  20131224  20131224  Evaluation method and evaluation system for utilization ratio of server 
Publications (2)
Publication Number  Publication Date 

CN103745092A CN103745092A (en)  20140423 
CN103745092B true CN103745092B (en)  20170111 
Family
ID=50502109
Family Applications (1)
Application Number  Title  Priority Date  Filing Date 

CN201310722907.4A CN103745092B (en)  20131224  20131224  Evaluation method and evaluation system for utilization ratio of server 
Country Status (1)
Country  Link 

CN (1)  CN103745092B (en) 
Citations (4)
Publication number  Priority date  Publication date  Assignee  Title 

CN1860451A (en) *  20031016  20061108  思科技术公司  Policybased network security management 
CN101557344A (en) *  20090521  20091014  南昌航空大学  Dynamic load balancing method based on spatial geographical locations 
CN102461075A (en) *  20090603  20120516  微软公司  Determining server utilization 
CN103365666A (en) *  20130726  20131023  浪潮电子信息产业股份有限公司  Method for evaluating utilization rate of server 

2013
 20131224 CN CN201310722907.4A patent/CN103745092B/en active IP Right Grant
Patent Citations (4)
Publication number  Priority date  Publication date  Assignee  Title 

CN1860451A (en) *  20031016  20061108  思科技术公司  Policybased network security management 
CN101557344A (en) *  20090521  20091014  南昌航空大学  Dynamic load balancing method based on spatial geographical locations 
CN102461075A (en) *  20090603  20120516  微软公司  Determining server utilization 
CN103365666A (en) *  20130726  20131023  浪潮电子信息产业股份有限公司  Method for evaluating utilization rate of server 
Also Published As
Publication number  Publication date 

CN103745092A (en)  20140423 
Similar Documents
Publication  Publication Date  Title 

Pencina et al.  Interpreting incremental value of markers added to risk prediction models  
Wilhelmsson  Spatial models in real estate economics  
Mazanec et al.  Tourism destination competitiveness: second thoughts on the World Economic Forum reports  
MacEachern et al.  A new ranked set sample estimator of variance  
Muckell et al.  SQUISH: an online approach for GPS trajectory compression  
Zhou et al.  Sustainability performance evaluation in industry by composite sustainability index  
Lei  Evaluation of three methods for estimating the Weibull distribution parameters of Chinese pine (Pinus tabulaeformis)  
Shin et al.  Assessment of modified Anderson–Darling test statistics for the generalized extreme value and generalized logistic distributions  
Chen et al.  Project delivery system selection of construction projects in China  
JP2010507146A (en)  Method and apparatus for capacity planning and resource optimization of distributed systems  
CN103514304B (en)  Project recommendation method and device  
US9798575B2 (en)  Techniques to manage virtual classes for statistical tests  
Chen et al.  Combining fuzzy iteration model with dynamic programming to solve multiobjective multistage decision making problems  
Bussmann et al.  The effect of globalization on national income inequality  
Gong et al.  Evolutionary generation of test data for many paths coverage based on grouping  
CN103324566B (en)  A kind of multiversion test method and device  
Xu et al.  A note on group decisionmaking procedure based on incomplete reciprocal relations  
Yin et al.  Sequential sufficient dimension reduction for large p, small n problems  
CN104317681B (en)  For the behavioral abnormal automatic detection method and detecting system of computer system  
CN104200087B (en)  For the parameter optimization of machine learning and the method and system of feature tuning  
US8719211B2 (en)  Estimating relatedness in social network  
Huang et al.  Flood disaster loss comprehensive evaluation model based on optimization support vector machine  
Heo et al.  Approximation of modified Anderson–Darling test statistics for extreme value distributions with unknown shape parameter  
Kun et al.  Inversion model of water distribution systems for nodal demand calibration  
CN102854968A (en)  Realtime energy consumption metering method of virtual machine 
Legal Events
Date  Code  Title  Description 

C06  Publication  
PB01  Publication  
C10  Entry into substantive examination  
SE01  Entry into force of request for substantive examination  
GR01  Patent grant  
GR01  Patent grant 