WO2011093011A1 - 情報処理装置及び情報処理方法 - Google Patents
情報処理装置及び情報処理方法 Download PDFInfo
- Publication number
- WO2011093011A1 WO2011093011A1 PCT/JP2011/000097 JP2011000097W WO2011093011A1 WO 2011093011 A1 WO2011093011 A1 WO 2011093011A1 JP 2011000097 W JP2011000097 W JP 2011000097W WO 2011093011 A1 WO2011093011 A1 WO 2011093011A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- workload
- resource amount
- resource
- amount
- unit
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/505—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the load
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/50—Indexing scheme relating to G06F9/50
- G06F2209/5019—Workload prediction
Definitions
- the present invention relates to an information processing apparatus and an information processing method.
- Patent Document 1 Japanese Patent No. 4240062 discloses that in a virtual server environment in which a virtual server is operated by a plurality of physical servers, the processing capacity margin of each physical server is measured based on the same standard. Are listed.
- Patent Document 2 Japanese Patent Laid-Open No. 2005-227932 estimates the amount of integer computation and the amount of floating-point computation of evaluation target software in a plurality of computers. It describes that the processing time of the evaluation target software is estimated.
- the present invention has been made in view of the above circumstances, and an object of the present invention is to provide an information processing apparatus and an information processing method for estimating a workload that can be executed with a free resource amount on a physical server. .
- the information processing apparatus includes a physical server resource amount allocated to each workload and information indicating each workload when each of the plurality of types of workloads is executed on the physical server.
- the performance information storage means for storing in association, the acquisition means for acquiring the free resource amount of the physical server, and at least one workload is selected from the information stored in the performance information storage means, and the selected workload
- a comparing means for comparing the resource amount associated with the amount of free resources acquired by the acquiring means, and when the amount of resources compared by the comparing means is less than or equal to the free resource amount, the comparing means First extraction means for extracting the selected workload.
- the information processing method includes a resource amount of each physical server allocated to each workload when the computer executes each of the plurality of types of workloads on the physical server, and each workload. Is stored in association with information indicating the physical server, the free resource amount of the physical server is acquired, at least one of the stored workloads is selected, and the resource amount associated with the selected workload is acquired. The selected workload is extracted when the compared resource amount is equal to or less than the free resource amount.
- the program according to an aspect of the present invention includes a physical server resource amount allocated to each workload and information indicating each workload when each of a plurality of types of workloads is executed on the physical server.
- One of the first storage process to be stored in association, the acquisition process to acquire the free resource amount of the physical server, and the workload stored in the first storage process is selected and associated with the selected workload.
- a comparison process comparing the resource amount obtained with the free resource amount acquired in the acquisition process, and the work selected in the comparison process when the resource amount compared in the comparison process is less than or equal to the free resource amount.
- a first extraction process for extracting a load is executed by a computer.
- an information processing apparatus and an information processing method for estimating an executable workload with an amount of free resources on a physical server are provided.
- 1 is a configuration diagram of an information processing system according to a first embodiment. It is a figure which shows the table memorize
- FIG. 1 is a configuration diagram of an information processing system 100 according to the first embodiment.
- the information processing system 100 includes an information processing apparatus 20 and a physical server 30 that can transmit and receive various information, programs, virtual servers, and the like via a communication network 10.
- the information processing apparatus 20 includes an acquisition unit 21, a comparison unit 22, a first extraction unit 23, and a storage unit 24. Note that the information processing apparatus 20 may include components other than the components illustrated here.
- the storage unit 24 stores the resource amount of the physical server 30 allocated to each workload when each of the plurality of types of workload is executed on the physical server 30 and information indicating each workload in association with each other.
- a performance information storage unit 25 is included.
- the performance information storage unit 25 is a part or all of the storage area of the storage unit 24. Further, the information stored in the performance information storage unit 25 may be acquired from the physical server 30, may be acquired from another device not illustrated, or may be acquired from another component not illustrated in the information processing device 20. May be acquired.
- the acquisition unit 21 acquires the free resource amount of the physical server 30.
- the comparison unit 22 selects at least one workload from the information stored in the performance information storage unit 25, the resource amount associated with the selected at least one workload, and the acquisition unit 21 acquires the resource amount. Compare the amount of free resources.
- the first extraction unit 23 extracts at least one workload selected by the comparison unit 22 when the resource amount compared by the comparison unit 22 is equal to or less than the free resource amount.
- at least one workload extracted by the first extraction unit 23 corresponds to a workload that can be executed with the amount of free resources on the physical server 30.
- a workload is a process or a group of processes that are executed by allocating resources. Specifically, floating point arithmetic, UDP (User Datagram Protocol) transmission / reception processing, random access to the hard disk drive, and the like. Note that the resource allocated to the workload may be a physical resource or a virtual resource.
- the resource amount or the free resource amount includes the CPU usage rate [%], the memory usage amount [MB], the read bandwidth and the write bandwidth [MB / s] for the hard disk drive, the network transmission bandwidth / the reception bandwidth. It is a value indicating any one or any combination of widths [bps].
- the resource amount used in the present embodiment may be a part of the resource amounts described here, or other types of resource amounts may be used.
- the process in which the acquisition unit 21 acquires the free resource amount of the physical server 30 may be executed autonomously or may be executed in other ways.
- the acquisition unit 21 may acquire the free resource amount at a predetermined time, or may acquire the free resource amount according to a command generated by the operator of the information processing apparatus 20.
- FIG. 2 is a diagram showing a table stored in the performance information storage unit 25.
- the format shown here is only an example, and the format is not necessarily limited to this format.
- the performance information storage unit 25 may have information equivalent to the table of FIG. 2 by associating a plurality of tables.
- the performance information storage unit 25 stores, for each workload type, information indicating a plurality of workloads and a plurality of resource amounts indicating different values in association with each other. Further, the performance information storage unit 25 stores the resource amount of each resource of a different type in association with each workload. That is, the performance information storage unit 25 stores information indicating a plurality of workloads with different amounts of at least one resource among a plurality of types of resources with respect to the same workload type.
- the comparison unit 22 selects at least two workloads from the information stored in the performance information storage unit 25, the total amount of resources associated with the selected workloads, and the free space acquired by the acquisition unit 21 Compare resource amount. More specifically, the comparison unit 22 selects resource amounts associated with at least two selected workloads, sums the selected resource amounts for each resource type, and is calculated for each resource type. The total amount of each resource and the free resource amount acquired by the acquisition unit 21 are compared for each resource type.
- the first extraction unit 23 extracts a plurality of workloads selected by the comparison unit 22 when the sum of the resource amounts compared by the comparison unit 22 is equal to or less than the free resource amount. More specifically, the first extraction unit 23 selects the workload and the resource amount selected by the comparison unit 22 when the total of the resource amounts compared by the comparison unit 22 is equal to or less than the free resource amount in all resource types. And are extracted respectively.
- All or part of the acquisition unit 21, the comparison unit 22, and the first extraction unit 23 described above may be realized by hardware, or may be realized by a program (or program code) that causes a processor to execute processing. Also good.
- the processor reads and executes the program from a recording medium such as a nonvolatile memory.
- the storage medium that stores the program may be the storage unit 24.
- the program When the acquisition unit 21, the comparison unit 22, or the first extraction unit 23 is implemented by a program, the program is stored in a processor (computer) readable storage medium. Then, the program causes the processor to execute the following processing.
- (B) The free resource amount of the physical server 30 is acquired.
- At least one workload is selected from the information stored in the performance information storage process, and the resource amount associated with the selected workload is compared with the free resource amount acquired by the acquisition unit 21. To do.
- D When the compared resource amount is equal to or less than the free resource amount, at least one workload selected in the comparison process is extracted.
- FIG. 3 is a flowchart showing an operation flow of the information processing apparatus 20 of the present embodiment.
- the information processing apparatus 20 associates the resource amount of the physical server 30 allocated to each workload when each of the plurality of types of workload is executed on the physical server 30 with information indicating each workload.
- the information processing apparatus 20 acquires the free resource amount of the physical server 30 (step S2).
- the information processing apparatus 20 selects one workload from the data stored in step S1, the resource amount associated with the selected workload, the free resource amount acquired by the acquisition unit 21, and Are compared (step S3).
- the resource amount compared in step S3 is equal to or less than the free resource amount (YES in step S4)
- one workload selected in step S3 is extracted (step S5). If the comparison process in step S3 has been executed for each workload indicated by the information stored in step S1 and all combinations of two or more workloads (YES in step S6), this flowchart is ended. Become.
- step S3 when the resource amount compared in step S3 is larger than the free resource amount (NO in step S4), the process proceeds to step S6. Further, in this flowchart, if the comparison process in step S3 is not executed for each of the workloads indicated by the information stored in step S1 or a part of all combinations of two or more workloads. (NO in step S6), the process proceeds to step S3.
- the workload extracted by the first extraction unit 23 is a workload or a combination of workloads that can be executed with the amount of free resources on the physical server 30.
- the physical server 30 does not overload even if it accepts the processing of the extracted workload.
- FIG. 4 is a configuration diagram of the information processing system 200 of the present embodiment.
- the information processing system 200 includes a plurality of physical servers 31 to 3m (m is a positive number) and an autonomous control server 40 that manages each virtual server running on the physical servers 31 to 3m. It is.
- the autonomous control server 40 and the physical servers 31 to 3m are connected via the communication network 10.
- 311 to 31n are assigned to virtual servers operating on the physical server 31, and 3m1 to 3mn are assigned to virtual servers operating on the physical server 3m for identification (n is a positive number).
- the physical servers 31 to 3m are realized by, for example, a computer that executes processing according to a program stored in a recording medium (magnetic disk, semiconductor memory, optical disk, etc.).
- a recording medium magnetic disk, semiconductor memory, optical disk, etc.
- the autonomous control server 40 includes a performance information collection unit 41, a performance measurement execution unit 42, a measurement result reception unit 43, a performance characteristic value generation unit 44 (interpolation means), a workload estimation unit 45, and a storage unit 46.
- the storage unit 46 includes a performance information storage unit 47.
- the performance information collection unit 41 receives the CPU usage rate, the memory usage rate, and the input / output performance value of the recording medium (for example, the maximum data transfer rate) from the physical servers 31 to 3m and the virtual server or the performance measurement virtual servers 311 to 3mn. Information on performance such as the ratio of input / output data transfer rate) and communication performance to the communication network 10 (transmission speed, bandwidth used, etc.) is acquired as a load index. Further, the performance information storage unit 47 stores the load index values of the physical servers 31 to 3m acquired at regular intervals as history data.
- the performance measurement execution unit 42 instructs the performance measurement virtual servers 311 to 3m1 on the physical servers 31 to 3m to measure performance information.
- the contents instructed here include the type of workload to be executed, the load value of the workload to be executed, and the like.
- information indicating the load level of the workload is referred to as a load value.
- the performance measurement virtual server 311 receives a measurement instruction from the performance measurement execution unit 42, the performance measurement virtual server 311 operates as follows. First, the performance measurement virtual server 311 executes the designated type of workload with the designated load value. Subsequently, the performance measurement virtual server 311 measures the resource amount on the physical server 31 allocated to the workload.
- the performance measurement virtual server 311 associates the information indicating the executed workload, the load value of the workload, and the measured resource amount, respectively, and the autonomous control server 40 as a measurement result of the performance information.
- the method for measuring the performance information is realized by using a conventional technique such as execution of a benchmark program or measurement using another dedicated program, and is not particularly limited here.
- the workload load value varies depending on the type of workload. For example, if the workload to be executed is a floating-point operation, it corresponds to the number of times. If the workload to be executed is UDP transmission and reception, the network transmission bandwidth and reception bandwidth correspond to the workload load value.
- the measurement result receiving unit 43 receives the measurement result of the performance information measured on the physical servers 31 to 3m.
- the performance information storage unit 47 stores the measurement result of the received performance information. For example, the performance information storage unit 47 associates information indicating a network reception workload, a load value assigned to the workload when the workload is executed, and a measured network reception bandwidth. May be stored.
- the performance characteristic value generation unit 44 can be assigned to the workload associated with the load value and the resource amount based on the correlation between the load value and the resource amount stored in association with the performance information storage unit 47. A performance characteristic value that is an interpolation value of the resource amount is calculated. More specifically, the performance characteristic value generation unit 44 generates an approximate function from the measurement result of the performance information stored in the performance information storage unit 47, and generates a performance characteristic value that can be taken by each workload. The performance information storage unit 47 stores the generated performance characteristic value.
- FIG. 5 is a diagram showing the relationship between the actually measured resource amount and the calculated performance characteristic value.
- the “ ⁇ ” point indicates the measured resource amount
- the “ ⁇ ” point indicates the performance characteristic value complemented by the approximation function.
- the workload (workload 1) that generates the network load and the network reception bandwidth are illustrated as having a proportional relationship, but the present invention is not necessarily limited to this.
- the performance characteristic value stored in the performance information storage unit 47 is a combination of the actually measured value and the complemented value.
- the approximate function is assumed to be a linear function or other function in which the coefficient is determined by the least square method or the like, but may be calculated by other methods.
- the workload estimation unit 45 uses the resource amount or performance characteristic value of the physical servers 31 to 3m stored in the performance information storage unit 47 to calculate the workload that can be processed by the free resource portion on each of the physical servers 31 to 3m. It is estimated and stored in the performance information storage unit 47.
- the workload estimation unit 45 acquires from the performance information storage unit 47 the resource amount of the virtual server used by the virtual servers 311 to 3m1 and the allowable resource amount that the physical servers 31 to 3m can tolerate. Then, a load index corresponding to the free resource amount of each of the physical servers 31 to 3m is calculated and obtained from the difference. That is, the workload estimation unit 45 functions in the same manner as the acquisition unit 21 in the first embodiment.
- the workload estimation unit 45 selects at least two of the workloads stored in the performance information storage unit 47, and selects a resource amount or a performance characteristic value associated with the selected workload for each workload. Then, the selected resource amount or performance characteristic value is totaled for each resource type, and the calculated resource amount or performance characteristic value for each resource type and the free resource amount acquired by the acquisition unit 21 are obtained. Compare for each resource type. That is, the workload estimation unit 45 functions in the same manner as the comparison unit 22 in the first embodiment.
- the workload estimation unit 45 extracts a combination of the selected workload and the resource amount or the performance characteristic value when the total of the compared resource amounts is equal to or less than the free resource amount in all resource types. To do. That is, the workload estimation unit 45 functions in the same manner as the first extraction unit 23 in the first embodiment.
- the workload estimation unit 45 compares the resource amount or the performance characteristic value included in the combination with the free resource amount to obtain an evaluation index. Calculate and extract the combination with the smallest evaluation index.
- the performance information storage unit 47 stores the load index collected by the performance information collection unit 41, the measurement result of the performance information received by the measurement result reception unit 43, and the performance characteristic value generated by the performance characteristic value generation unit 44. To do. In other words, the performance information storage unit 47 holds at least information equivalent to the information included in the performance information storage unit 25 in the first embodiment.
- FIG. 6 is a flowchart showing an operation flow of the autonomous control server 40 of the present embodiment.
- the autonomous control server 40 acquires the load index values of the physical servers 31 to 3m at regular intervals as described above, and stores them in the performance information storage unit 47 as history data. It shall be.
- the autonomous control server 40 calls the performance measurement virtual servers 311 to 3m1 of the physical servers 31 to 3m, and instructs measurement of performance characteristics (step A1). Subsequently, the autonomous control server 40 receives the measurement results notified from the performance measurement virtual servers 311 to 3m1 of the physical servers 31 to 3m, and stores the received measurement results in the performance information storage unit 47 ( Step A2). Then, the autonomous control server 40 generates an approximate function from the measurement result of the performance characteristic, calculates the performance characteristic value by complementing the value between the actual measurement values, and stores the calculated performance characteristic value in the performance information storage unit 47. Store (step A3).
- the autonomous control server 40 uses the performance characteristic values of the physical servers 31 to 3m stored in the performance information storage unit 47 to estimate the workload that can be processed by the free resource portion on each physical server, and is estimated. Information indicating the selected workload is stored in the performance information storage unit 47 (step A4).
- the autonomous control server 40 calculates the difference between the resource amount or performance characteristic value included in the extracted combination and the free resource amount for each resource type, and normalizes the calculated difference.
- the evaluation index can be obtained by adding the values obtained. In more detail, it formulates like the following formula
- FIG. 7 is a flowchart showing a procedure for obtaining a combination of workload values that minimizes the evaluation index G described above.
- a possible combination of workloads is selected (step B1).
- step B1 a possible combination of workloads is selected (step B2).
- step B3 For each of the m types of loads indicator, it calculates a performance characteristic value f ij when the workload x i (x i) (step B3).
- the calculated load index value is added to the total value of the load index for workloads x 1 to x i ⁇ 1, and the total value of the load index up to workload x i is calculated. (Step B4).
- step B4 When the total value of the m types of load indices calculated in step B4 does not exceed the maximum value (NO in step B5), and all the workloads x 1 to x n are not determined ( (NO in step B6), 1 is added to i (step B7), and the process of step B2 is executed for the next workload x i + 1 .
- step B3 In addition, among the total values of the m types of load indexes calculated in step B3, there are those exceeding the maximum value (YES in step B5), and the processing in step B1 has been completed for some workload combinations. If not (NO in step B9), the process proceeds to step B2 to call the next set of workloads.
- step B10 When the processing of step B1 has been completed for all the combinations of workloads (YES in step B9), among the returned workload combinations, the workload combinations with the smallest evaluation index G are estimated as workloads. Extracted as a value (step B10).
- the operator of the autonomous control server 40 can view the combination of the workload extracted by the workload estimation unit 45 and the resource amount or performance characteristic value. Thus, the operator can compare the processing performance of the virtual servers based on the estimated workload values when the virtual servers are operated using the free resources of the physical servers 31 to 3m.
- FIG. 8 is a configuration diagram of the information processing system 300 of this embodiment.
- the autonomous control server 50 included in the information processing system 300 includes a workload estimation unit 51, a load index estimation unit 52, and a virtual server operation unit 53, as compared with the autonomous control server 40 of the second embodiment. Different. Other components are the same as those in the second embodiment, and thus detailed description thereof is omitted.
- the workload estimation unit 51 can execute the same processing as the workload estimation unit 45 of the second embodiment. Then, the workload estimation unit 51 uses the virtual server 3ip (2 ⁇ p ⁇ n) used on the physical server 3i (1 ⁇ i ⁇ m) among the plurality of physical servers 31 to 3m, for example. The load index (resource amount) of the server is acquired, and the free resource amount on another physical server 3j (1 ⁇ j ⁇ m, i ⁇ j) is obtained.
- the workload estimation unit 51 selects at least one of the workloads stored in the performance information storage unit 47, the resource amount associated with the selected workload, the acquired load index of the virtual server, Compare Then, the workload estimation unit 51 extracts the selected workload when the compared load index approximates the virtual server load index.
- the workload extracted by the workload estimation unit 51 is estimated as the workload corresponding to the virtual server 3ip.
- the load index estimation unit 52 extracts the load index stored in the performance information storage unit 47, associated with the physical server 3j, and associated with the word load extracted by the workload estimation unit 51, and the extracted load index. Store in the performance information storage unit 47.
- the load index extracted by the load index estimation unit 52 is an estimated value of the load index when the virtual server 3iq is executed on the physical server 3j.
- the virtual server operating unit 53 operates the virtual server 3iq on the physical server 3j when the free resource amount on the physical server 3j acquired by the workload estimation unit 51 is larger than the load index extracted by the load index estimation unit 52.
- FIG. 9 is a flowchart showing an operation flow of the autonomous control server 50 of the present embodiment. However, it is assumed that the autonomous control server 50 has collected the measurement results of the performance characteristics of the installed physical servers 31 to 3m and stores them in the performance information storage unit 47.
- the autonomous control server 50 acquires the performance characteristic value of the source physical server 3i and the load index of the virtual server 3ip from the performance information storage unit 47 (step C1). Then, the autonomous control server 50 estimates the workload for the virtual server 3ip as described above based on the performance characteristic value and the load index acquired in step C1, and displays information indicating the estimated workload as performance information. It memorize
- the autonomous control server 50 operates the virtual server 3iq on the physical server 3j (Step C5).
- the autonomous control server 50 stops the movement of the virtual server 3iq to the physical server 3j ( Step C6).
- the virtual server by generating a performance characteristic value from the correspondence relationship between the actually measured workload and the load index, and estimating the workload and the load index from them, the virtual server can be used on the physical server having different performance. Since the load index on the destination physical server can be estimated with high accuracy when the server is moved, it is possible to avoid insufficient capacity of the virtual server and generation of useless free resources.
- the operator of the autonomous control server 50 can view the estimated value of the load index stored in the performance information storage unit 47 as the increase amount of the load index generated on the physical server 3j that is the movement destination.
- a plurality of physical servers capable of sending and receiving virtual servers via a communication network; For each physical server, when each of the multiple types of workloads is executed on each physical server, the resource amount of each physical server allocated to each workload is associated with information indicating each workload.
- Performance information storage means for storing; A first acquisition means for acquiring a virtual server resource amount used by a virtual server operating on the first physical server among the plurality of physical servers; Select at least one workload from the information stored in the performance information storage means, and determine the resource amount associated with the selected workload and the virtual server resource amount acquired by the first acquisition means.
- a comparison means for comparing A first extraction unit that extracts a workload selected by the comparison unit when the resource amount compared by the comparison unit is equal to or less than the virtual server resource amount; A second extraction unit that calculates an evaluation index of a resource amount and a virtual server resource amount associated with the workload extracted by the first extraction unit, and extracts a workload having the smallest evaluation index; Third extraction means for extracting a resource amount stored in the performance information storage means, associated with a second physical server different from the first physical server, and associated with a workload extracted by the second extraction means;
- the performance information storage means associates the resource amount of the first resource and the resource amount of the second resource assigned to each workload when each of the plurality of types of workloads is executed, and Remember about each load, Interpolated values of the first resource and the second resource that can be allocated to the workload based on the correlation between the resource amount of the first resource and the resource amount of the second resource stored in the performance information storage unit Is further provided with an estimation means for calculating for each workload,
- the performance information storage means stores the interpolation value calculated by the estimation means in association with each of the plurality of types of workloads
- the comparison means selects the resource amount or interpolation value associated with the selected workload for each workload, sums the selected resource amount or interpolation value for each resource type, and calculates the calculated resource amount or Comparing the total of the interpolation values and the virtual server resource amount acquired by the first acquisition means for each resource type,
- the first extracting means when the total of the respective resource amounts compared by the comparing means is equal to or less than the virtual server resource amount in all resource types, the workload
- (Appendix 3) Second acquisition means for acquiring the amount of free resources of the second physical server; Virtual server operating means for operating the virtual server on the second physical server when the free resource amount acquired by the second acquiring means is larger than the resource amount extracted by the third extracting means;
- the performance information storage unit stores, for each of the plurality of types of workloads, information indicating the plurality of workloads and a plurality of resource amounts corresponding to the same type of resources and different from each other in association with each other.
- the comparison means selects the resource amount associated with the selected workload for each workload, compares the total amount of the selected resource amount with the free resource amount,
- the first extraction unit extracts the workload and the resource amount selected by the comparison unit, respectively, when the sum of the resource amounts compared by the comparison unit is equal to or less than the free resource amount.
- the information processing apparatus according to any one of 1 to 3.
- the performance information storage means stores, for each of the plurality of types of workloads, information indicating the plurality of workloads and resource amounts of different types of resources in association with each other,
- the acquisition unit acquires a free resource amount corresponding to at least one of a plurality of types of resources corresponding to the resource amount stored in the performance information storage unit,
- the comparison means totals the resource amounts associated with the selected workload for each resource type, and calculates the total resource amount and the free resource amount acquired by the acquisition means for each resource type.
- the first extraction unit is configured to calculate a plurality of workloads selected by the comparison unit when the total amount of each resource type compared by the comparison unit is equal to or less than the free resource amount in all resource types.
- the information processing apparatus according to any one of appendices 1 to 4, wherein the information processing apparatus is extracted.
- the fourth extraction unit calculates the difference between the resource amount or the interpolation value included in the combination extracted by the first extraction unit and the free resource amount for each resource type, and normalizes the calculated difference.
- the information processing apparatus according to any one of supplementary notes 1 to 5, wherein the evaluation index is obtained by adding the values.
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Debugging And Monitoring (AREA)
Abstract
Description
まず、第1の実施形態の情報処理システム100について説明する。図1は、第1の実施形態に係る情報処理システム100の構成図である。情報処理システム100は、通信網10を介して種々の情報やプログラム、仮想サーバ等を送受信可能な情報処理装置20と物理サーバ30とから構成される。
(a)物理サーバ30上で複数種のワークロードの各々が実行されたときに各ワークロードに割り当てられた物理サーバ30のリソース量と各ワークロードを示す情報とを関連付けて記憶する。
(b)物理サーバ30の空きリソース量を取得する。
(c)性能情報記憶処理で記憶された情報の中から少なくとも1つのワークロードを選択し、選択されたワークロードに関連付けられたリソース量と、取得部21によって取得された空きリソース量とを比較する。
(d)比較されたリソース量が空きリソース量以下であるとき、比較処理で選択された少なくとも1つのワークロードを抽出する。
続いて、第2の実施形態に係る情報処理システム200について説明する。図4は、本実施形態の情報処理システム200の構成図である。図4に示すように、情報処理システム200は、複数の物理サーバ31~3m(mは正数)と、物理サーバ31~3mで稼動する各仮想サーバを管理する自律制御サーバ40とを有する構成である。自律制御サーバ40と物理サーバ31~3mとは通信網10を介して接続される。図4では、物理サーバ31で稼動する仮想サーバに符号311~31nを付与し、物理サーバ3mで稼動する仮想サーバに符号3m1~3mnを付与して識別している(nは正数)。
n:ワークロードの種別数
m:単一のワークロードに関連付けられた負荷指標の数
xi:ある種別のワークロードiを表す変数
fij(x):ワークロードiの性能特性値がxのときの負荷指標jの値
G:評価指標
次に、第3の実施形態について図面を参照して詳細に説明する。図8は、本実施形態の情報処理システム300の構成図である。情報処理システム300に含まれる自律制御サーバ50は、第2の実施形態の自律制御サーバ40と比較して、ワークロード推定部51と負荷指標推定部52と仮想サーバ稼動部53とを有する点で異なる。その他の構成要素については、第2の実施形態と同等であるので、詳細な説明は割愛する。
通信網を介して仮想サーバを授受することができる複数の物理サーバと、
各々の物理サーバに関し、当該各物理サーバ上で複数種のワークロードの各々が実行されたときに各ワークロードに割り当てられた各物理サーバのリソース量と各ワークロードを示す情報とをそれぞれ関連付けて記憶する性能情報記憶手段と、
上記複数の物理サーバのうち、第1物理サーバ上で稼動した仮想サーバによって使用された仮想サーバリソース量を取得する第1取得手段と、
上記性能情報記憶手段に記憶された情報の中から少なくとも1つのワークロードを選択し、選択されたワークロードに関連付けられたリソース量と、上記第1取得手段によって取得された仮想サーバリソース量とを比較する比較手段と、
上記比較手段によって比較されたリソース量が上記仮想サーバリソース量以下であるとき、上記比較手段によって選択されたワークロードを抽出する第1抽出手段と、
上記第1抽出手段によって抽出されたワークロードに関連付けられたリソース量と仮想サーバリソース量との評価指標を算出し、評価指標が最小であるワークロードを抽出する第2抽出手段と、
上記性能情報記憶手段に記憶され、上記第1物理サーバと異なる第2物理サーバに関連付けられ、上記第2抽出手段によって抽出されたワークロードと関連付けられたリソース量を抽出する第3抽出手段と、
を備えることを特徴とする情報処理システム。
上記性能情報記憶手段は、上記複数種のワークロードの各々が実行されたときに各ワークロードに割り当てられた第1リソースのリソース量と、第2リソースのリソース量とを関連付けて、上記各ワークロードに関しそれぞれ記憶し、
上記性能情報記憶手段に記憶された上記第1リソースのリソース量と、上記第2リソースのリソース量との相関関係に基づいて、当該ワークロードに割り当て可能な第1リソースおよび第2リソースの補間値を、ワークロードごとに算出する推測手段を更に備え、
上記性能情報記憶手段は、上記推測手段によって算出された補間値を、上記複数種のワークロードの夫々に関連付けて記憶し、
上記比較手段は、選択されたワークロードに関連付けられたリソース量または補間値をワークロードごとに選択し、選択されたリソース量または補間値をリソースの種別ごとに合計し、算出されたリソース量または補間値の合計と、上記第1取得手段によって取得された仮想サーバリソース量とを、リソースの種別ごとに比較し、
上記第1抽出手段は、上記比較手段によって比較された各リソース量の合計が、全てのリソース種別において仮想サーバリソース量以下である場合、上記比較手段によって選択されたワークロードと、リソース量または補間値とを組み合わせて抽出し、
上記第2抽出手段は、上記第1抽出手段によって抽出された組合せに含まれるリソース量または補間値と仮想サーバリソース量とを比較して評価指標を算出し、評価指標が最小である組合せを抽出することを特徴とする付記1に記載の情報処理システム。
第2物理サーバの空きリソース量を取得する第2取得手段と、
上記第2取得手段によって取得された空きリソース量が、上記第3抽出手段によって抽出されたリソース量より大きい場合、上記仮想サーバを上記第2物理サーバで稼動させる仮想サーバ稼動手段と、
を備えることを特徴とする付記1又は2に記載の情報処理システム。
上記性能情報記憶手段は、当該複数種のワークロードの各々に関し、複数のワークロードを示す情報と、同一種のリソースに対応し、かつ互いに異なる複数のリソース量と、を関連付けて記憶し、
上記比較手段は、選択されたワークロードに関連付けられたリソース量をワークロードごとに選択し、選択されたリソース量の総和と空きリソース量とを比較し、
上記第1抽出手段は、上記比較手段によって比較されたリソース量の総和が空きリソース量以下であるとき、上記比較手段によって選択されたワークロードとリソース量とを夫々抽出することを特徴とする付記1から3のいずれか1つに記載の情報処理装置。
上記性能情報記憶手段は、当該複数種のワークロードの各々に関し、複数のワークロードを示す情報と、異なる複数種のリソースのリソース量とをそれぞれ関連付けて記憶し、
上記取得手段は、上記性能情報記憶手段に記憶されたリソース量に対応する複数種のリソースのうち少なくとも一種に対応する空きリソース量を取得し、
上記比較手段は、選択されたワークロードに関連付けられたリソース量をリソースの種別ごとに合計し、算出されたリソース量の合計と、上記取得手段によって取得された空きリソース量とをリソースの種別ごとに比較し、
上記第1抽出手段は、上記比較手段によって比較されたリソース種別ごとの各リソース量の合計が、全てのリソース種別において空きリソース量以下であるとき、上記比較手段によって選択された複数のワークロードを抽出することを特徴とする付記1から4のいずれか1つに記載の情報処理装置。
上記第4抽出手段は、上記第1抽出手段によって抽出された組合せに含まれるリソース量または補間値と空きリソース量との差分を、リソースの種別ごとに算出し、算出された差分を正規化した値を合算して当該評価指標を求めることを特徴とする付記1から5のいずれか1つに記載の情報処理装置。
Claims (10)
- 物理サーバ上で複数種のワークロードの各々が実行されたときに各ワークロードに割り当てられた前記物理サーバのリソース量と各ワークロードを示す情報とを関連付けて記憶する性能情報記憶手段と、
前記物理サーバの空きリソース量を取得する取得手段と、
前記性能情報記憶手段に記憶された情報の中から少なくとも1つのワークロードを選択し、選択されたワークロードに関連付けられたリソース量と、前記取得手段によって取得された空きリソース量とを比較する比較手段と、
前記比較手段によって比較されたリソース量が前記空きリソース量以下である場合に、前記比較手段によって選択されたワークロードを抽出する第1抽出手段と、
を備えることを特徴とする情報処理装置。 - 請求項1に記載の情報処理装置であって、
前記取得手段は、物理サーバ上で仮想サーバを稼動したとき、当該仮想サーバによって使用された仮想サーバリソース量と、当該物理サーバが許容可能なリソース許容量と、を取得し、取得された当該リソース許容量と当該仮想サーバリソース量との差分から前記空きリソース量を得ることを特徴とする情報処理装置。 - 請求項2に記載の情報処理装置であって、
前記比較手段は、さらに、選択された前記ワークロードに関連付けられたリソース量と、前記取得手段によって取得された前記仮想サーバリソース量とを比較し、
前記比較手段によって比較されたリソース量が前記仮想サーバリソース量に近似する場合に、前記比較手段によって選択されたワークロードを抽出する第2抽出手段を更に備えることを特徴とする情報処理装置。 - 請求項3に記載の情報処理装置であって、
前記性能情報記憶手段は、通信網を介して前記仮想サーバを授受することができる複数の物理サーバの各々に関し、前記物理サーバ上で複数種のワークロードの各々が実行されたときに各ワークロードに割り当てられたリソース量と各ワークロードを示す情報と、を関連付けて記憶し、
前記取得手段は、前記複数の物理サーバのうち、第1物理サーバ上で稼動した仮想サーバによって使用された仮想サーバリソース量を取得し、さらに第2物理サーバ上の空きリソース量を取得し、
前記比較手段は、前記性能情報記憶手段に記憶された情報の中から少なくとも1つのワークロードを選択し、選択されたワークロードに関連付けられたリソース量の総和と、前記取得手段によって取得された空きリソース量とを比較し、
前記第2抽出手段は、前記比較手段によって比較された前記リソース量が前記仮想サーバリソース量に近似する場合に、前記比較手段によって選択された前記ワークロードを抽出し、
前記性能情報記憶手段に記憶され、前記第2物理サーバに関連付けられ、前記第2抽出手段によって抽出されたワークロードと関連付けられたリソース量を抽出する第3抽出手段と、
前記取得手段によって取得された第2物理サーバ上の空きリソース量が、前記第3抽出手段によって抽出されたリソース量より大きい場合に、前記仮想サーバを前記第2物理サーバ上で稼動させる仮想サーバ稼動手段と、
を更に備えることを特徴とする情報処理装置。 - 請求項1乃至4いずれか1項に記載の情報処理装置であって、
前記比較手段は、前記性能情報記憶手段に記憶されたワークロードのうち少なくとも二つを選択し、選択されたワークロードに関連付けられたリソース量の総和と、前記取得手段によって取得された空きリソース量とを比較し、
前記第1抽出手段は、前記比較手段によって比較された前記リソース量の総和が前記空きリソース量以下である場合に、前記比較手段によって選択された複数のワークロードを抽出することを特徴とする情報処理装置。 - 請求項5に記載の情報処理装置であって、
前記物理サーバ上の性能計測用仮想サーバに対して、複数種のワークロードの実行と、実行させるワークロードの負荷値と、を指示する計測実行手段を備え、
前記性能計測用仮想サーバは、指示されたワークロードを、指示された負荷値で実行したとき、当該ワークロードに割り当てられる前記物理サーバ上のリソース量を計測し、実行されたワークロードを示す情報と、当該ワークロードに対応する負荷値と、計測されたリソース量と、を夫々関連付けて通知し、
前記性能情報記憶手段は、前記性能計測用仮想サーバによって関連付けて通知された前記ワークロードを示す情報と前記負荷値と前記リソース量とを夫々記憶することを特徴とする情報処理装置。 - 請求項6に記載の情報処理装置であって、
前記性能情報記憶手段に関連付けられて記憶された前記負荷値と前記リソース量との相関関係に基づいて、当該負荷値と当該リソース量に関連付けられたワークロードに割り当て可能なリソース量の補間値を算出する補間手段を更に備え、
前記性能情報記憶手段は、前記補間手段によって算出された補間値を、前記複数種のワークロードの夫々に関連付けて記憶し、
前記比較手段は、選択された前記ワークロードに関連付けられた前記リソース量または前記補間値をワークロードごとに選択し、選択されたリソース量または補間値をリソースの種別ごとに合計し、算出されたリソースの種別ごとのリソース量または補間値の合計と、前記取得手段によって取得された空きリソース量とを、リソースの種別ごとに比較し、
前記第1抽出手段は、前記比較手段によって比較された各リソース量の合計が、全てのリソース種別において前記空きリソース量以下である場合、前記比較手段によって選択されたワークロードと、リソース量または補間値とを抽出することを特徴とする情報処理装置。 - 請求項7に記載の情報処理装置であって、
前記第1抽出手段は、前記比較手段によって選択されたワークロードと、リソース量または補間値とを組み合わせて抽出し、
前記第1抽出手段によって抽出されたワークロードとリソース量または補間値との組合せごとに、当該組合せに含まれるリソース量または補間値と空きリソース量とを比較して評価指標を算出し、当該評価指標が最小である組合せを抽出する第4抽出手段を更に備えることを特徴とする情報処理装置。 - コンピュータが、
物理サーバ上で複数種のワークロードの各々が実行されたときに各ワークロードに割り当てられた前記物理サーバのリソース量と各ワークロードを示す情報とを関連付けて記憶し、
前記物理サーバの空きリソース量を取得し、
前記記憶されたワークロードの少なくとも1つを選択し、
前記選択されたワークロードに関連付けられたリソース量と、前記取得された空きリソース量とを比較し、
前記比較されたリソース量が前記空きリソース量以下である場合に、前記選択されたワークロードを抽出する、
ことを含む情報処理方法。 - 物理サーバ上で複数種のワークロードの各々が実行されたときに各ワークロードに割り当てられた前記物理サーバのリソース量と各ワークロードを示す情報とを関連付けて記憶する第1記憶処理と、
前記物理サーバの空きリソース量を取得する取得処理と、
前記第1記憶処理で記憶されたワークロードの少なくとも1つを選択し、選択されたワークロードに関連付けられたリソース量と、前記取得処理で取得された空きリソース量とを比較する比較処理と、
前記比較処理で比較されたリソース量が前記空きリソース量以下である場合に、前記比較処理で選択されたワークロードを抽出する第1抽出処理と、
をコンピュータに実行させることを特徴とするプログラム。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2011551735A JPWO2011093011A1 (ja) | 2010-01-29 | 2011-01-12 | 情報処理装置及び情報処理方法 |
US13/575,890 US9218214B2 (en) | 2010-01-29 | 2011-01-12 | Information processing apparatus and information processing method |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2010-018990 | 2010-01-29 | ||
JP2010018990 | 2010-01-29 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2011093011A1 true WO2011093011A1 (ja) | 2011-08-04 |
Family
ID=44319010
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2011/000097 WO2011093011A1 (ja) | 2010-01-29 | 2011-01-12 | 情報処理装置及び情報処理方法 |
Country Status (3)
Country | Link |
---|---|
US (1) | US9218214B2 (ja) |
JP (1) | JPWO2011093011A1 (ja) |
WO (1) | WO2011093011A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2013132735A1 (ja) * | 2012-03-08 | 2013-09-12 | 日本電気株式会社 | 仮想マシン管理装置及び仮想マシン管理方法 |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120198063A1 (en) * | 2009-10-09 | 2012-08-02 | Nec Corporation | Virtual server system, autonomous control server thereof, and data processing method and computer program thereof |
JP6044248B2 (ja) * | 2012-10-05 | 2016-12-14 | 富士ゼロックス株式会社 | 情報処理装置、応用プログラム導入装置およびプログラム |
US9940167B2 (en) * | 2014-05-20 | 2018-04-10 | Red Hat Israel, Ltd. | Identifying memory devices for swapping virtual machine memory pages |
SG11202009256SA (en) * | 2018-03-23 | 2020-10-29 | Carolina Cloud Exchange Inc | Quantifying usage of disparate computing resources as a single unit of measure |
KR102488614B1 (ko) * | 2021-01-18 | 2023-01-17 | 주식회사 텐 | 가상화된 리소스를 관리하는 방법, 장치 및 컴퓨터 프로그램 |
JP2022187285A (ja) * | 2021-06-07 | 2022-12-19 | 富士通株式会社 | 管理装置,ストレージシステム及び情報処理方法 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10283210A (ja) * | 1997-04-01 | 1998-10-23 | Hitachi Ltd | 仮想計算機システム間の仮想計算機移動制御方式 |
JP2007323245A (ja) * | 2006-05-31 | 2007-12-13 | Nec Corp | 計算機システムおよび性能計測方法ならびに管理サーバ装置 |
WO2008062864A1 (en) * | 2006-11-24 | 2008-05-29 | Nec Corporation | Virtual machine locating system, virtual machine locating method, program, virtual machine management device and server |
WO2008102739A1 (ja) * | 2007-02-23 | 2008-08-28 | Nec Corporation | 仮想サーバシステム及び物理サーバ選択方法 |
JP2008293117A (ja) * | 2007-05-22 | 2008-12-04 | Hitachi Ltd | 仮想計算機の性能監視方法及びその方法を用いた装置 |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6802062B1 (en) | 1997-04-01 | 2004-10-05 | Hitachi, Ltd. | System with virtual machine movable between virtual machine systems and control method |
US7171668B2 (en) * | 2001-12-17 | 2007-01-30 | International Business Machines Corporation | Automatic data interpretation and implementation using performance capacity management framework over many servers |
JP2005227932A (ja) | 2004-02-12 | 2005-08-25 | Nippon Telegr & Teleph Corp <Ntt> | ソフトウェアの処理時間推定方法および処理時間推定装置 |
US8024736B1 (en) * | 2005-01-28 | 2011-09-20 | Hewlett-Packard Development Company, L.P. | System for controlling a distribution of unutilized computer resources |
US8365182B2 (en) * | 2006-10-02 | 2013-01-29 | International Business Machines Corporation | Method and system for provisioning of resources |
US8370472B2 (en) * | 2008-09-02 | 2013-02-05 | Ca, Inc. | System and method for efficient machine selection for job provisioning |
-
2011
- 2011-01-12 WO PCT/JP2011/000097 patent/WO2011093011A1/ja active Application Filing
- 2011-01-12 US US13/575,890 patent/US9218214B2/en active Active
- 2011-01-12 JP JP2011551735A patent/JPWO2011093011A1/ja active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10283210A (ja) * | 1997-04-01 | 1998-10-23 | Hitachi Ltd | 仮想計算機システム間の仮想計算機移動制御方式 |
JP2007323245A (ja) * | 2006-05-31 | 2007-12-13 | Nec Corp | 計算機システムおよび性能計測方法ならびに管理サーバ装置 |
WO2008062864A1 (en) * | 2006-11-24 | 2008-05-29 | Nec Corporation | Virtual machine locating system, virtual machine locating method, program, virtual machine management device and server |
WO2008102739A1 (ja) * | 2007-02-23 | 2008-08-28 | Nec Corporation | 仮想サーバシステム及び物理サーバ選択方法 |
JP2008293117A (ja) * | 2007-05-22 | 2008-12-04 | Hitachi Ltd | 仮想計算機の性能監視方法及びその方法を用いた装置 |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2013132735A1 (ja) * | 2012-03-08 | 2013-09-12 | 日本電気株式会社 | 仮想マシン管理装置及び仮想マシン管理方法 |
JPWO2013132735A1 (ja) * | 2012-03-08 | 2015-07-30 | 日本電気株式会社 | 仮想マシン管理装置及び仮想マシン管理方法 |
US9600311B2 (en) | 2012-03-08 | 2017-03-21 | Nec Corporation | Virtual-machine managing device and virtual-machine managing method |
Also Published As
Publication number | Publication date |
---|---|
JPWO2011093011A1 (ja) | 2013-05-30 |
US9218214B2 (en) | 2015-12-22 |
US20120311600A1 (en) | 2012-12-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2011093011A1 (ja) | 情報処理装置及び情報処理方法 | |
JP5218390B2 (ja) | 自律制御サーバ、仮想サーバの制御方法及びプログラム | |
CN103383655B (zh) | 用于在qos感知云中管理合并的工作负载的性能干扰模型 | |
US10289440B2 (en) | Capacity risk management for virtual machines | |
JP5827678B2 (ja) | 仮想コンテナのシステムにおけるリソース容量評価のための方法および装置 | |
US9600343B2 (en) | System and method for analyzing computing system resources | |
US9645909B2 (en) | Operation management apparatus and operation management method | |
JP2008027442A (ja) | サブタスク・プロセッサの分散スケジューリング | |
Zhang et al. | CPU load predictions on the computational grid | |
US20130054492A1 (en) | Determining an option for decommissioning or consolidating software | |
JP6299759B2 (ja) | 予測関数作成装置、予測関数作成方法、及びプログラム | |
US9612876B2 (en) | Method and apparatus for estimating a completion time for mapreduce jobs | |
US20130290499A1 (en) | Method and system for dynamic scaling in a cloud environment | |
JP2014032674A (ja) | 仮想マシンのリソース配置システム及びその方法 | |
JP7006607B2 (ja) | 分散処理システム、分散処理方法、及び記録媒体 | |
KR20140117905A (ko) | 클라우드 서비스의 가상자원 할당을 위한 퍼지 로직 기반의 자원평가 장치 및 방법 | |
US20220100548A1 (en) | Network performance assurance system and network performance assurance method | |
JP5445739B2 (ja) | リソース割当装置、リソース割当方法、及びプログラム | |
JP2014164425A (ja) | 情報処理装置、リソース制御方法及びプログラム | |
US9983911B2 (en) | Analysis controller, analysis control method and computer-readable medium | |
US9501321B1 (en) | Weighted service requests throttling | |
Kumar et al. | Load balancing algorithm to minimize the makespan time in cloud environment | |
WO2012093469A1 (ja) | 性能評価装置及び性能評価方法 | |
NO20101110A1 (no) | Fremgangsmater og systemer med anvendelse av maksimal entropi | |
US20150256421A1 (en) | Information processing method and information processing apparatus |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11736728 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2011551735 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13575890 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 11736728 Country of ref document: EP Kind code of ref document: A1 |