WO2013011624A1 - Système de gestion de machine virtuelle et procédé de gestion de machine virtuelle - Google Patents

Système de gestion de machine virtuelle et procédé de gestion de machine virtuelle Download PDF

Info

Publication number
WO2013011624A1
WO2013011624A1 PCT/JP2012/003793 JP2012003793W WO2013011624A1 WO 2013011624 A1 WO2013011624 A1 WO 2013011624A1 JP 2012003793 W JP2012003793 W JP 2012003793W WO 2013011624 A1 WO2013011624 A1 WO 2013011624A1
Authority
WO
WIPO (PCT)
Prior art keywords
virtual machine
server device
resource amount
load characteristic
value
Prior art date
Application number
PCT/JP2012/003793
Other languages
English (en)
Japanese (ja)
Inventor
雅也 藤若
Original Assignee
日本電気株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 日本電気株式会社 filed Critical 日本電気株式会社
Publication of WO2013011624A1 publication Critical patent/WO2013011624A1/fr

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

Definitions

  • the present invention relates to a technology for rearranging virtual machines in a virtual environment.
  • Patent Document 1 resources are allocated by deploying virtual machines based on the time periodicity of past performance information (resource information) of each virtual machine so that the total performance of the combination of virtual machines is maximized. It has been proposed to use it efficiently.
  • Patent Document 2 it is proposed to use resources efficiently by using the correlation between virtual machines and consolidating virtual machines with high correlation on the same server.
  • the conventional method as described above cannot be applied when the performance information (resource information) of the virtual machine has no time periodicity or correlation with other virtual machines.
  • the present invention has been made in view of the above-described matters, and provides a virtual machine placement determination technique that can effectively use the resources of the server device while maintaining the service level of the virtual machine.
  • the first aspect relates to a virtual machine management apparatus.
  • the virtual machine management device includes a first information acquisition unit that acquires time-series data of usage resource amounts for each server device, a requested resource amount for a virtual machine, and the requested resource amount as an average value.
  • a second information acquisition unit that acquires a moving average width indicating a time width for calculation, and a virtual machine that acquires time series data of each server device acquired by the first information acquisition unit by the second information acquisition unit
  • An analysis unit that generates load characteristic data indicating a probability density distribution of the amount of resource used for each server device based on each time-series data averaged with a moving average width for each, and the analysis unit
  • Al and a selector for selecting at least one are Depending on the generated load characteristic data of each server device and the requested resource amount of the virtual machine acquired by the second information acquisition unit.
  • the second aspect relates to a virtual machine management system including the virtual machine management apparatus according to the first aspect and a plurality of server apparatuses.
  • each server device collects information on the amount of used resources in its own server device, and transmits the collected information on the amount of used resources to the virtual machine management device;
  • a control unit that allocates resources to a virtual machine running on the server device.
  • the third aspect relates to a virtual machine management method.
  • the computer acquires time-series data of the used resource amount for each server device, and calculates the requested resource amount and the requested resource amount as an average value for the virtual machine.
  • the moving average width indicating the time width of each server device is obtained, the time series data of each server device is averaged with the moving average width for each virtual machine, and the use of each server device is based on the averaged time series data.
  • Generating load characteristic data indicating a probability density distribution of resource amounts, and selecting at least one of a plurality of server devices according to the load characteristic data of each server device and the requested resource amount of a virtual machine; including.
  • Another aspect of the present invention may be a computer program that causes a computer to implement the configuration of the first aspect described above, or a computer-readable recording medium that records such a program. Good.
  • This recording medium includes a non-transitory tangible medium.
  • FIG. 1 is a diagram conceptually illustrating a configuration example of a virtual machine management system according to an embodiment of the present invention.
  • FIG. 2 is a diagram illustrating an example of time-series data of CPU usage rate (CPU USAGE).
  • FIG. 3 is a diagram illustrating an example of time-series data of the CPU usage rate (CPU USAGE) averaged by the moving average width.
  • FIG. 4 is a diagram illustrating an example of load characteristics obtained from time-series data of averaged CPU usage rate (CPU USAGE).
  • FIG. 5 is a diagram conceptually illustrating a processing configuration example of the virtual machine management device and the server device in the first embodiment.
  • FIG. 6 is a diagram illustrating the concept of the expected value calculation formula.
  • FIG. 7 is a diagram illustrating an example of the value function.
  • FIG. 8 is a flowchart illustrating an operation example of the virtual machine management apparatus according to the first embodiment.
  • FIG. 9 is a diagram conceptually illustrating a processing configuration example of the virtual machine management device and the server device in the second embodiment.
  • FIG. 10 is a flowchart illustrating an operation example of the virtual machine management apparatus according to the second embodiment.
  • FIG. 11 is a diagram illustrating time-series data of the usage resource amount (unit: 1 second) that is not averaged of the server device 20 (# 1).
  • FIG. 12 is a diagram illustrating time-series data of the usage resource amount (unit: 1 second) that is not averaged by the server device 20 (# 2).
  • FIG. 11 is a diagram illustrating time-series data of the usage resource amount (unit: 1 second) that is not averaged of the server device 20 (# 2).
  • FIG. 12 is a diagram illustrating time-series data of the usage resource amount (unit: 1 second) that is not averaged by the server device 20 (# 2).
  • FIG. 13 is a diagram illustrating time-series data of the usage resource amount averaged over the moving average width (20 seconds) regarding the server device 20 (# 1).
  • FIG. 14 is a diagram illustrating time-series data of the usage resource amount averaged over the moving average width (20 seconds) regarding the server device 20 (# 2).
  • FIG. 15 is a diagram illustrating the load characteristics indicated by the load characteristic data generated for the server 20 (# 1).
  • FIG. 16 is a diagram illustrating the load characteristics indicated by the load characteristic data generated for the server 20 (# 2).
  • FIG. 17 is a diagram illustrating load characteristics obtained from time-series data of the usage resource amount that is not averaged regarding the server 20 (# 1).
  • FIG. 18 is a diagram illustrating load characteristics obtained from time-series data of the usage resource amount that is not averaged regarding the server 20 (# 2).
  • FIG. 19 is a diagram showing the degree to which the virtual machine A can secure resources when the virtual machine A is deployed on the server 20 (# 1) in a free use type.
  • FIG. 20 is a diagram illustrating the degree to which the virtual machine A can secure resources when the virtual machine A is deployed on the server 20 (# 2) in a free-use manner.
  • FIG. 21 is a diagram illustrating time-series data of the resource amount of the virtual machine A.
  • FIG. 22 is a diagram illustrating time-series data obtained by adding the time-series data of the resource amounts of the server 20 (# 2) and the virtual machine A.
  • FIG. 23 is a diagram illustrating load characteristics of the server 20 (# 2) obtained from time-series data obtained by adding the time-series data of the resource amounts of the server 20 (# 2) and the virtual machine A.
  • FIG. 24 is a diagram illustrating the degree to which the virtual machine B can secure resources when the virtual machine B is deployed to the server 20 (# 1) in a free-use manner.
  • FIG. 25 is a diagram illustrating the degree to which the virtual machine B can secure resources when the virtual machine B is deployed to the server 20 (# 2) in a free-use type.
  • FIG. 26 shows the degree to which the virtual machine B can secure resources when the virtual machine B is deployed on the server 20 (# 1) in a mixed type (fixed allocation type: 0.5 (GHz), free use type: remaining).
  • FIG. 27 shows the degree to which the virtual machine B can secure resources when the virtual machine B is deployed on the server 20 (# 2) in a mixed type (fixed allocation type: 0.5 (GHz), free use type: remaining).
  • FIG. FIG. 28A is a diagram illustrating load characteristic data of the server device 20 (# 1).
  • FIG. 28B is a diagram showing load characteristic data of the server device 20 (# 2).
  • FIG. 29 is a diagram illustrating load characteristic data of a virtual machine.
  • FIG. 30A is a diagram illustrating the load characteristic data before correction of the server device 20 (# 1) after placement of the virtual machine.
  • FIG. 30B is a diagram illustrating load characteristic data before correction of the server device 20 (# 2) after placement of the virtual machine.
  • FIG. 1 is a diagram conceptually illustrating a configuration example of a virtual machine management system according to an embodiment of the present invention.
  • the virtual machine management system 1 includes a virtual machine management device 10 and a plurality of server devices 20 that can operate a virtual machine to be managed.
  • the virtual machine management system 1 is abbreviated as the management system 1
  • the virtual machine management apparatus 10 is abbreviated as the management apparatus 10.
  • the management apparatus 10 selects one server apparatus 20 that is optimal for operating the target virtual machine from among the plurality of server apparatuses 20.
  • the management device 10 is a so-called computer, and includes, for example, a CPU (Central Processing Unit) 11, a memory 12, an input / output interface (I / F) 14, and the like that are connected to each other via a bus 15.
  • the memory 12 is a RAM (Random Access Memory), a ROM (Read Only Memory), a hard disk, a portable storage medium, or the like.
  • the input / output I / F 14 is connected to a communication device that communicates with each server device 20 via a communication network.
  • the input / output I / F 14 may be connected to a user interface device such as a display device or an input device.
  • Each server device 20 is also a so-called computer, and has, for example, a hardware configuration as shown in FIG.
  • Each server device 20 only needs to have a hardware configuration capable of operating a plurality of virtual machines, and does not need to have the same hardware configuration.
  • the present embodiment does not limit the hardware configuration of the management device 10 and each server device 20.
  • each server device 20 realize each processing unit by causing the CPU 11 to execute a program stored in the memory 12, for example.
  • each server device 20 will be described as one server device 20 unless it is necessary to distinguish each server device 20 from one another. Further, the present embodiment does not limit the number of server devices 20.
  • the management device 10 calculates the requested resource amount and the requested resource amount as an average value for the first information acquisition unit that obtains time series data of the used resource amount for each server device 20 and the target virtual machine.
  • a second information acquisition unit that acquires a moving average width indicating a time width for the purpose, and a virtual machine in which time series data of each server device 20 acquired by the first information acquisition unit is acquired by the second information acquisition unit
  • An analysis unit that generates each of the load characteristic data indicating the probability density distribution of the use resource amount of each server device 20 based on each time series data averaged by the moving average width, and the analysis unit
  • a plurality of server devices 20 Comprising a selection unit for selecting at least one, and from within.
  • the requested resource amount and moving average width are set for the target virtual machine.
  • the requested resource amount indicates the total amount of resources that can be required per predetermined unit time (moving average width) according to each resource type of the server device 20.
  • a CPU Central Processing Unit
  • 1 gigahertz (GHz) or the like is set as the requested resource amount.
  • 100 megabps or the like is set as the requested resource amount.
  • MB megabytes
  • the present embodiment does not limit the resource type of the requested resource amount of the virtual machine.
  • resource amounts of a plurality of resource types may be included. In the following description, a case where the resource amount of the CPU resource is used as the requested resource amount will be described as an example.
  • the moving average width indicates the width (time width) of the section for obtaining the average value when the requested resource amount is secured as an average value.
  • the corresponding service level is maintained. Accordingly, the smaller the value of the moving average width, the higher the corresponding service level.
  • the server device 20 at the placement destination secures a minimum resource amount ( ⁇ ⁇ ⁇ ) in ⁇ seconds. Need to be done.
  • the management device 10 averages the time series data (see FIG. 2) of the amount of resource used for each server device 20 with the moving average width for the virtual machine. Subsequently, load characteristic data (see FIG. 4) indicating the probability density distribution of the used resource amount is generated for each server device 20 based on the averaged time-series data (see FIG. 3).
  • the load characteristic data represents, in a form of probability density distribution, the history (time series data) of the amount of used resources of the server device 20 averaged with the moving average width ⁇ .
  • FIG. 2 shows an example of time-series data of the CPU usage rate (CPU USAGE) as the used resource amount of each server device 20 acquired by the first information acquisition unit.
  • FIG. 4 shows an example of load characteristics obtained from the averaged time series data shown in FIG. In the example of FIG. 4, the horizontal axis indicates the amount of resource used for each moving average width ⁇ , and the vertical axis indicates the probability.
  • the load characteristic data generated by the management apparatus 10 is formed by the resource use amount (utilization resource amount) and the probability thereof.
  • the load characteristics data of each server apparatus 20 reflecting the service level (moving average width) of the virtual machine and the requested resource amount of the virtual machine in this way At least one is selected.
  • the selected at least one server device 20 is determined as a placement destination of the virtual machine.
  • the present embodiment it is possible to select the server device to which the virtual machine is arranged in consideration of the service level of the virtual machine. Therefore, according to the present embodiment, it is possible to determine the placement of the virtual machine so that the resources of the server device 20 can be effectively utilized while maintaining the service level of the virtual machine. Furthermore, according to the present embodiment, since the server device 20 is selected using the load characteristic data from which information related to time is excluded, the time periodicity between the load of the server device 20 and the requested resource of the virtual machine Even when there is no time correlation, it is possible to determine an appropriate placement destination of the virtual machine.
  • FIG. 5 is a diagram conceptually illustrating a processing configuration example of the virtual machine management device 10 and the server device 20 in the first embodiment.
  • the management apparatus 10 includes a first information acquisition unit 101, a second information acquisition unit 102, an analysis unit 103, a selection unit 104, a resource information database (DB) 109, and the like.
  • the server device 20 includes a control unit 201, a measurement unit 202, a virtual machine 205, and the like.
  • the management device 10 and the server device 20 each implement the processing units illustrated in FIG. 5 by executing a program stored in the memory 12 by the CPU 11.
  • server equipment In the server device 20, one or more virtual machines 205 are executed. This embodiment does not limit the number of virtual machines 205 executed (operated) in each server device 20.
  • the control unit 201 executes resource allocation to the virtual machine 205.
  • the control unit 201 sets the CPU resource use priority for each virtual machine 205.
  • the CPU resource use priority is for setting a CPU resource amount guaranteed at a minimum.
  • the server device 20 in which the total CPU resource amount is 10 (GHz)
  • the CPU resource use priority when the CPU resource use priority is set to 30 (%) for the virtual machine A, the virtual machine A is 3 ( GHz) CPU resources can be used at a minimum.
  • assigning the minimum guaranteed resource amount to a certain virtual machine in the server device 20 as in the setting of the CPU resource use priority will be referred to as fixed assignment type assignment, fixed assignment, or the like. . Therefore, when a certain resource amount is allocated to the virtual machine A in the fixed allocation type, the virtual machine A can always use the resource in the allocated range.
  • the control unit 201 can also perform a free-use type assignment for the fixed assignment type assignment.
  • the free use type is a borrowing type resource allocation method. That is, when a free-use resource allocation is performed for a certain virtual machine, the control unit 201 does not allocate a new resource to the virtual machine, but performs other tasks (other processes, other virtual machines). A free resource that is not used by a machine or the like is allocated to the virtual machine.
  • the resources allocated in the free usage type are limited resources that can be used only for the free time, and are not guaranteed to be used continuously. In other words, resources allocated in the free usage type are pre-allocated to other virtual machines with a request from another virtual machine allocated in the fixed allocation type.
  • the control unit 201 sets the CPU resource use priority related to the virtual machine A to 0 (%) when all the requested resource amounts related to the CPU resource of the virtual machine A are allocated in the free use type.
  • the measuring unit 202 collects information on the amount of resources used in the server device 20 itself.
  • the information on the amount of used resources is information on the amount of resources used on the server 20 for each resource type, and can also be called performance information.
  • the resource usage information includes, for example, at least one of a CPU usage rate, a network usage amount, a memory usage amount, and the like.
  • the information on the used resource amount may be information on the used resource amount for each virtual machine operating on the own server device 20.
  • the information on the used resource amount includes at least one of the free resource amount and the total resource amount for each resource type.
  • the measurement unit 202 sends the usage resource amount information collected in this way to the virtual machine management apparatus 10 at an arbitrary timing.
  • the measurement unit 202 may send the used resource amount information to the virtual machine management device 10, or may send the information spontaneously at a predetermined cycle.
  • the present embodiment does not limit the way in which the usage resource amount information is sent from the server device 20 to the virtual machine management device 10.
  • the first information acquisition unit 101 acquires the used resource amount information regarding each server device 20 from each server device 20, and stores the acquired used resource amount information of each server device 20 in the resource information DB 109, respectively.
  • the resource information DB 109 stores the used resource amount information of each server device 20 as time series data. For the time series data, the time when the used resource amount information is acquired from each server device 20 by the first information acquisition unit 101 may be used, and the used resource amount information is collected in each server device 20. Time may be used.
  • the second information acquisition unit 102 acquires the requested resource amount and the moving average width for the target virtual machine.
  • the requested resource amount and moving average width related to a virtual machine can also be said to indicate the service level of the virtual machine, hereinafter, the requested resource amount and moving average width are referred to as service level information related to the virtual machine.
  • the second information acquisition unit 102 acquires the service level information when the user inputs using a user interface device connected to the input / output I / F 14. In this case, the second information acquisition unit 102 displays a screen for inputting service level information on a display device (not shown), and acquires data input on the screen as the service level information. Further, the second information acquisition unit 102 may acquire the service level information from another device via communication, or may acquire it from a portable storage medium that stores the service level information.
  • the analysis unit 103 uses the moving average width related to the virtual machine included in the service level information acquired by the second information acquisition unit 102 to store the resource information time series data of each server device 20 stored in the resource information DB 109. Analyze. Specifically, the analysis unit 103 averages the time series data of each server device 20 with the moving average width related to the virtual machine, and uses the resource used for each server device 20 based on each averaged time series data. Load characteristic data indicating the probability density distribution of the quantity is respectively generated. The load characteristic indicated by the load characteristic data is shown in FIG. 4, for example. The load characteristic data is realized as, for example, array data of the amount of resource used and its probability.
  • the analysis unit 103 may generate the load characteristic data of all the target server devices 20 or may generate the load characteristic data of only the designated server device 20. Further, the generation of the load characteristic data by the analysis unit 103 may be executed at a timing when the resource information DB 109 is updated, may be executed at a predetermined cycle, or may be executed by a request from another processing unit. May be.
  • the selection unit 104 includes an expected value calculation unit 106, a value calculation unit 107, and the like.
  • the expected value calculation unit 106 determines, for each server device 20, the free space allocated to other virtual machines in a fixed allocation type. Expected values indicating the amount of resources that can be used by the virtual machine in the form of use (free use type) are calculated. The expected value calculation unit 106 calculates the expected value using the following (Equation 1).
  • ⁇ represents the total amount of resources already allocated in the server apparatus 20 in the fixed allocation type.
  • ⁇ indicates the resource amount that is desired to be used in the free usage type among the requested resource amount of the virtual machine, and the requested resource amount ⁇ of the virtual server is set as the initial value.
  • ⁇ indicates a width for dividing the horizontal axis (utilization resource amount) of the load characteristic of the server device 20, and a predetermined value larger than the moving average width of the virtual server is set for ⁇ .
  • L indicates the number of divisions on the horizontal axis (resource amount) of the load characteristics of the server device 20. That is, ⁇ and L indicate the accuracy for obtaining the expected value.
  • A indicates the width (number of divisions) of the horizontal axis of the load characteristic of the server device corresponding to the resource amount ⁇ .
  • P ( ⁇ n) indicates the probability that the amount of used resources is ⁇ (n ⁇ 1) to ⁇ n.
  • the selection unit 104 sets the ratio of the expected value calculated by the expected value calculation unit 106 to the resource amount that is desired to be used in the free usage type in the requested resource amount of the virtual machine is a predetermined ratio (for example, 95%). It is determined whether or not there are more server devices 20. In the initial state, in order to select the server device 20 that can allocate all the requested resource amounts in the free usage type, the selection unit 104 determines whether there is a server device 20 in which the ratio of the expected value to the requested resource amount exceeds a predetermined rate. Determine. The selection unit 104 extracts the server device 20 corresponding to this determination as a placement destination candidate. Note that the selection unit 104 may select the server device 20 having the largest calculated expected value as the placement destination among the plurality of server devices 20 extracted as placement destination candidates.
  • the value calculation unit 107 uses the use resource in each of the load characteristic data of the server device 20 when the use resource amount is larger than the first predetermined amount or smaller than the second predetermined amount indicating a value equal to or less than the first predetermined amount.
  • First value data indicating the value of the load characteristic for each server device 20 by applying a value function that outputs a smaller value than when the amount falls between the first predetermined amount and the second predetermined amount. are calculated respectively.
  • the first value data shows a high value for the load characteristic in which the total resource usage is expressed with an average probability, and shows a small value for the load characteristic with a bias in the resource usage.
  • the value calculation unit 107 calculates the first value data using, for example, the following (Formula 2).
  • the value function for obtaining the value of the load characteristic is set to a function as shown in FIG.
  • FIG. 7 is a diagram illustrating an example of the value function.
  • P (), ⁇ and L in the following (Formula 2) are the same as the above (Formula 1).
  • the division number L affects the calculation accuracy such that the larger the value, the larger the calculation amount. Therefore, the value of the division number L is held so as to be adjustable in advance in consideration of the calculation amount and the calculation accuracy.
  • This value function f is a function that outputs a high value (value) when the distribution of resource usage is close to the ideal state, and its domain is set to 0 or more and 1 or less, for example.
  • the ideal state indicates a state where the overload state is small, resources are used without waste, and the average resource usage is high.
  • the value function f is defined so as to output the highest value for the CPU usage rate of 60% and to output a lower value as the distance from the value increases. Just do it.
  • the value function f in the above (Equation 2) is an example, and the value function f should be a function that gives a higher value to a used resource amount as it approximates a predetermined amount indicating a desired average used resource amount. That's fine.
  • the value function f in (Expression 2) may be a function that outputs a negative value or a function that outputs a positive value.
  • the predetermined amount may be determined for each resource type and may be held in advance so as to be adjustable, or may be acquired from each server device 20.
  • the value calculation unit 107 estimates the load characteristic data after the virtual machine is arranged for each server device, and applies the above (Equation 2) to each estimated load characteristic data, thereby Second value data relating to the server device is respectively calculated.
  • the second value data indicates the value of the load characteristic of the server device 20 after the placement of the virtual machine.
  • the value calculation unit 107 uses the division number ⁇ used for calculating the second value data as the division number L, which is a predetermined value, and the total amount ⁇ of resource amounts allocated by the fixed allocation type in the server device 20. , Based on the above (Equation 1). ⁇ used for calculating the second value data is calculated by adding the fixed allocation amount ⁇ allocated to the placement target virtual machine to the total resource amount ⁇ used when calculating the first value data. Is done.
  • the load characteristic data after placement of the virtual machine is acquired by multiplying the load characteristic data related to the virtual machine and the load characteristic data of each server device 20.
  • the load characteristic data regarding the virtual machine may be acquired by the second information acquisition unit 102 as one of the service level information of the virtual machine, or may be data generated based on the requested resource amount of the virtual machine. .
  • the load characteristic data of the virtual machine is acquired in advance by, for example, simulation.
  • load characteristic data indicating that the requested resource amount of the virtual machine is secured with a probability of 100 (%) may be used.
  • the selection unit 104 determines the placement of the virtual machine according to the increase amount from the first value data corresponding to the virtual machine before placement of the virtual machine to the second value data corresponding to the placement of the virtual machine calculated by the value calculation unit 107. Select the server device as the destination. For example, the selection unit 104 selects the server device 20 that maximizes the increase from the first value data to the second value data as the placement destination of the virtual machine. Thus, by selecting the server device having the largest increase in the value of the load characteristics before and after the placement, it is possible to level the load characteristic deviation of the server device 20. In addition, it is possible to share resources that have been allocated in the fixed allocation type effectively in a balanced manner among a plurality of virtual machines.
  • the selection unit 104 may select the server device 20 that is the placement destination of the virtual machine in accordance with either the first value data or the second value data calculated by the value calculation unit 107. .
  • the selection unit 104 When there is no target server device 20 due to the selection using the expected value as described above, the selection unit 104, that is, the server device 20 that satisfies the service level of the virtual machine in complete free-use resource allocation. If it does not exist, the server device 20 that is the placement destination of the virtual machine is selected as follows.
  • the selection unit 104 selects the server device 20 that can allocate a part of the requested resource amount of the virtual machine in the fixed allocation type and can allocate the remaining part in the free usage type.
  • a method of allocating a part of the requested resource amount with the fixed allocation type and allocating the remaining part with the free use type may be referred to as a mixed allocation type.
  • the selection unit 104 determines a predetermined amount ⁇ of the requested resource amount ⁇ as a resource amount that can be allocated in a fixed allocation type.
  • the resource amount allocated in the fixed allocation type may be referred to as a fixed allocation amount.
  • the predetermined amount ⁇ is a fixed allocation amount. This predetermined amount ⁇ may be information that is stored in advance, or may be determined as a predetermined ratio of the requested resource amount ⁇ .
  • the selection unit 104 extracts the server device 20 having a free resource amount larger than the fixed allocation amount ⁇ from all the server devices 20.
  • the selection unit 104 recalculates the expected value using the remaining requested resource amount ( ⁇ ) obtained by subtracting the fixed allocation amount ⁇ as the resource amount ⁇ desired to be used in the free use type, and the recalculated value is calculated. Based on the expected value, one of the extracted server devices 20 is selected.
  • the selection unit 104 sequentially increases the fixed allocation amount ⁇ until one server device 20 as a virtual machine placement destination is found in this way. If the fixed allocation amount ⁇ is equal to the requested resource amount ⁇ , the server device 20 that can allocate resources in a fixed allocation type is selected.
  • FIG. 8 is a flowchart illustrating an operation example of the virtual machine management apparatus 10 according to the first embodiment.
  • the usage resource amount information collected by the first information acquisition unit 101 is already stored in the resource information DB 109 for each server device 20, and each server is analyzed by the analysis unit 103 based on the time series data of the usage resource amount. It is assumed that the load characteristic data of the device 20 is generated. Furthermore, it is assumed that the requested resource amount and the moving average width related to the target virtual machine have already been acquired by the second information acquisition unit 102.
  • the selection unit 104 sets the resource amount ⁇ desired to be used in the free usage type to the requested resource amount ⁇ of the virtual machine (S10). This means that an attempt is made to select a server device 20 that can use all of the requested resource amount of the virtual machine in a free use type.
  • the selection unit 104 refers to the time series data of the resource information stored in the resource information DB 109 and the service level information of the virtual machine, so that the total resource amount ⁇ allocated in the fixed allocation type is determined as the resource amount.
  • a server device 20 larger than ⁇ is extracted (S11).
  • the selection unit 104 requests the expected value calculation unit 106 to calculate an expected value. Note that, when the number of extracted server devices 20 is one, the selection unit 104 determines the server device 20 as a virtual machine placement destination (not illustrated).
  • the expected value calculation unit 106 calculates an expected value for each server device 20 extracted in the process (S11) based on the load characteristic data and the resource amount ⁇ of each server device 20 (S14). ).
  • the expected value calculated here indicates, for example, the minimum resource amount that can be used by the virtual machine in a free use state in a state where all resources are assigned to another virtual machine in a fixed assignment type.
  • the selection unit 104 determines that the sum of the expected value calculated by the expected value calculation unit 106 and the fixed allocation amount ⁇ is a predetermined ratio of the requested resource amount ⁇ . It is determined whether or not there are more server devices 20 (for example, 95%) (S15). This determination means that the server device 20 that can use the resource amount ⁇ desired to be used in the free use type with a high probability corresponding to the predetermined ratio is detected. Therefore, when the resource amount ⁇ is the requested resource amount ⁇ , the server device 20 that satisfies the service level of the virtual machine with a completely free use type is detected.
  • the selection unit 104 determines the server device 20 as a virtual machine placement destination (S21). On the other hand, when there are a plurality of corresponding server devices 20 (S20; YES), the selection unit 104 requests the value calculation unit 107 to calculate value data.
  • the value calculation unit 107 In response to this, the value calculation unit 107 generates load characteristic data after the virtual machine is arranged for each server device 20 extracted in the process (S15) (S22). In this generation, the value calculation unit 107 multiplies the load characteristic data before placement of the virtual machine generated by the analysis unit 103 and the load characteristic data corresponding to the service level information of the virtual machine.
  • the value calculation unit 107 applies the value function f to the load characteristic data before placement of the virtual machine and the load characteristic data after placement of the virtual machine, so that the first value is related to each server device 20 extracted in the process (S15). Data and second value data are calculated (S23).
  • the first value data indicates the value of the load characteristic before placement of the virtual machine
  • the second value data indicates the value of the load characteristic after placement of the virtual machine.
  • the selection unit 104 selects the server device 20 that maximizes the increase from the first value data to the second value data as the placement destination of the virtual machine (S24).
  • the selection unit 104 operates as follows.
  • the selection unit 104 determines a part (predetermined amount ⁇ ) of the requested resource amount ⁇ as a resource amount (fixed allocation amount) allocated in a fixed allocation type (S30). The selection unit 104 extracts server devices 20 having free resources larger than the fixed allocation amount ⁇ from all the server devices 20 (S32).
  • the selection unit 104 When there is no server device 20 having a free resource larger than the fixed allocation amount ⁇ in the corresponding server device 20 (S33; NO), the selection unit 104 has a complete fixed allocation type, a complete free usage type, and a mixed type. Whichever resource allocation method is used, it is determined that there is no server device 20 capable of maintaining the service level of the virtual machine (resource shortage determination) (S34). Such a determination result may be output from a display device or the like.
  • the selection unit 104 determines the resource amount ⁇ desired to be used in the free usage type, A value obtained by subtracting the fixed allocation amount ⁇ from the requested resource amount ⁇ is set (S36). This setting means that the resource amount ⁇ is fixedly allocated to the virtual machine and the remaining ( ⁇ ) is allocated in a free use type in the server apparatus 20 at the placement destination.
  • the selection unit 104 determines that the total resource amount ⁇ allocated in the fixed allocation type is the server device 20 extracted in the process (S32). A new server device 20 larger than the changed resource amount ⁇ is extracted (S11). Thereafter, the processing (S11) and subsequent steps are executed. On the other hand, if the changed resource amount ⁇ is 0 (S37; YES), the selection unit 104 means that there is a server device 20 that can allocate the requested resource amount ⁇ in a completely fixed allocation type. The process (S20) and subsequent steps are executed.
  • the selection unit 104 sequentially increases the fixed allocation amount ⁇ until one server device 20 as a virtual machine placement destination is found (S21 or S24) or until a resource shortage is determined (S34). Repeat the process.
  • a part of the requested resource amount ⁇ is a fixed allocation type. Assuming that the allocation is made, the server apparatus 20 is selected with a high possibility that a part of the requested resource amount ⁇ can be secured in the free use type. Thereby, the certainty of selection of the placement destination of the virtual machine can be improved while obtaining the above effects.
  • the value data of the load characteristics before and after the virtual machine placement in each server device 20 is calculated.
  • a server device that has the greatest increase in load characteristic value due to the placement of the virtual machine is selected as the placement destination.
  • the server device 20 that can stably provide resources according to the service level of the virtual machine is selected as the placement destination of the virtual machine, the number of times of relocation of the virtual machine is reduced. As a result, the performance degradation of the entire system due to the relocation of the virtual machine can be suppressed.
  • FIG. 9 is a diagram conceptually illustrating a processing configuration example of the virtual machine management device 10 and the server device 20 in the second embodiment.
  • the management device 10 further includes a rearrangement determination unit 301 in addition to the configuration of the first embodiment.
  • the processing configuration of the server device 20 is the same as that of the first embodiment.
  • the hardware configurations of the management device 10 and the server device 20 are the same as those in the first embodiment (see FIG. 1).
  • the second information acquisition unit 102 further acquires virtual machine relocation time information in addition to the information described in the first embodiment.
  • This rearrangement time indicates the time required to rearrange the virtual machine from the server device 20 to another server device 20.
  • the second information acquisition unit 102 may acquire the relocation time information when the user inputs using a user interface device connected to the input / output I / F 14.
  • the 2nd information acquisition part 102 displays the screen for inputting rearrangement time on a display apparatus (not shown), and acquires the data input into this screen.
  • the second information acquisition unit 102 may acquire the information on the rearrangement time from another device via communication, or may acquire it from a portable storage medium that stores the information.
  • the moving average width is ⁇
  • the relocation time is ⁇
  • the resource amount per unit time that can be secured on the server device 20 to be relocated is ⁇ .
  • a virtual machine indicating ⁇ > ⁇ and ⁇ ( ⁇ ) ⁇ ⁇ ⁇ ( ⁇ ⁇ ⁇ ) cannot similarly maintain the service level when rearranged. This is because the resource amount ⁇ ( ⁇ ) ⁇ ⁇ that can be secured after the rearrangement is smaller than the maximum resource amount ( ⁇ ⁇ ⁇ ).
  • a virtual machine with a high service level that cannot be relocated itself will be referred to as a non-relocatable virtual machine, and a relocatable virtual machine may be referred to as a relocatable virtual machine.
  • the relocation determination unit 301 determines whether or not the target virtual machine can be relocated based on the relationship between the service level information regarding the virtual machine and the relocation time.
  • the rearrangement determination unit 301 may perform the above determination using only the relationship between the moving average width ⁇ and the rearrangement time ⁇ , or the requested resource amount ⁇ , the moving average width ⁇ , the rearrangement time ⁇ , and the rearrangement.
  • the determination may be performed using the relationship of the resource amount ⁇ per unit time that can be secured on the server device 20.
  • the resource amount ⁇ per unit time that can be secured on the server device 20 to be rearranged is acquired from the resource information stored in the resource information DB 109. This resource amount ⁇ agrees with the free resource amount that can be fixedly allocated in the server device 20 to be rearranged.
  • the predetermined amounts A and B are predetermined values of 0 or more that are determined in advance in consideration of errors, and are stored in advance.
  • the selection unit 104 selects a server device 20 that can allocate all of the requested resource amount ⁇ in a fixed allocation type to a virtual machine that is determined to be unrearranged by the relocation determination unit 301. There is a possibility that sufficient resources may not be allocated to a virtual machine to which resources are allocated in the free usage type or the mixed type. Generally, when the service level set for a virtual machine cannot be maintained, the virtual machine is relocated to another server. However, a virtual machine that cannot be relocated cannot maintain its service level when relocated. Therefore, the selection unit 104 maintains the service level of the virtual machine by selecting the server device 20 that can allocate all of the requested resource amount ⁇ in a fixed allocation type.
  • the selection unit 104 selects the server device 20 that is the placement destination for the virtual machine determined to be relocatable by the same method as in the first embodiment. For example, in the case of a virtual machine having a large moving average width, even if resources are not allocated for a certain period of time, abundant resources can be allocated on another server device 20 after the rearrangement.
  • FIG. 10 is a flowchart illustrating an operation example of the virtual machine management apparatus 10 according to the second embodiment.
  • the same processes as those in the first embodiment are denoted by the same reference numerals as those in FIG.
  • the relocation determination unit 301 determines whether or not a virtual machine can be relocated based on the relationship between the service level information of the virtual machine and the relocation time (S91).
  • the rearrangement determination unit 301 determines that the target virtual machine can be rearranged (S92; NO)
  • the processing (S10) and subsequent steps are executed as in the first embodiment.
  • the relocation determination unit 301 determines that the target virtual machine cannot be relocated (S92; YES)
  • the relocation determination unit 301 notifies the selection unit 104 to that effect.
  • the selection unit 104 extracts a server device 20 having a free resource amount larger than the requested resource amount ⁇ of the virtual machine (S93). Subsequently, the selection unit 104 executes the processing (S20) and subsequent steps as in the first embodiment.
  • the second embodiment it is possible to select the server apparatus 20 that is unlikely to be rearranged as a placement destination for a virtual machine whose rearrangement itself leads to a violation of the service level. As a result, according to the second embodiment, the service level of the virtual machine can be maintained.
  • the operation of the management apparatus 10 is started, for example, when the user operates the user interface.
  • the selection unit 104 checks whether it is possible to allocate resources to the virtual machine A in a free usage type.
  • FIGS. 11 and 12 are diagrams showing time-series data of the usage resource amounts of the server device 20 (# 1) and the server device 20 (# 2), respectively. Since the moving average width of the analysis unit 103 and the virtual machine A is 20 seconds, the time series data of the used resource amount of each server device is averaged by the moving average width (20 seconds).
  • FIGS. 13 and 14 are diagrams illustrating an example of averaged time-series data regarding the server apparatuses 20 (# 1) and 20 (# 2).
  • the analysis unit 103 generates load characteristic data of the server devices 20 (# 1) and 20 (# 2) based on the averaged time series data. Specifically, the analysis unit 103 obtains a probability density distribution with the CPU usage rate as the x-axis for the averaged time-series data.
  • 15 and 16 are diagrams illustrating the load characteristics indicated by the load characteristic data generated for the server devices 20 (# 1) and 20 (# 2). The maximum value (right end) of the x-axis in each load characteristic of FIGS. 15 and 16 is the total amount of resources allocated in a fixed allocation type to other virtual machines already running on each server device. ⁇ is shown.
  • FIG. 17 and FIG. 18 show load characteristics (probability density distribution) obtained from time-series data of the usage resource amounts that are not averaged for the server devices 20 (# 1) and 20 (# 2).
  • the probability density distribution (FIGS. 15 and 16) of the average used resource amount is higher than the probability density distribution (FIGS. 17 and 18) before being averaged. It can be seen that the probability of increasing is decreasing. Since the required resource amount of the virtual machine A is 0.5 (GHz), the larger the sum of the probability densities of the portions surrounded by the broken-line squares shown in FIG. 19 and FIG. The expected value of becomes smaller.
  • the expected value calculation unit 106 calculates an expected value that can secure all of the requested resource amount of the virtual machine A by the free use type for each of the server devices 20 (# 1) and 20 (# 2).
  • the analysis unit 103 calculates each expected value using, for example, the above (Equation 1).
  • the selection unit 104 determines a server device for which an expected value equal to or greater than a predetermined ratio of the requested resource amount is calculated as a placement destination candidate.
  • the expected value calculated for the server device 20 (# 2) is larger than that of the server device 20 (# 1). This is because the server device 20 (# 1) has a high probability of a large amount of resources (high load) (see FIG. 19), and therefore the probability that an empty resource can be provided to the virtual machine A is low. (# 2) is because the probability that the resource amount is large is low (see FIG. 20), and thus the probability that an empty resource can be provided to the virtual machine A is increased. 19 and 20, the server device 20 (# 2) is determined as a placement destination candidate.
  • the value calculation unit 107 When a plurality of server devices are selected as placement destination candidates, the value calculation unit 107 relates to each server device that is a placement destination candidate, load property value data (first value data) before placement of the virtual machine, and Value data (second value data) of the load characteristics after the placement of the virtual machines is calculated.
  • the value data of the load characteristics of each server device is calculated using, for example, the above (Equation 2).
  • the selection unit 104 selects the server device having the largest increase amount from the value data before the virtual machine placement to the value data after the virtual machine placement as the placement destination server device.
  • FIG. 21 shows an example of the time series data of the resource amount related to the virtual machine A
  • FIG. 22 shows the example of FIG. 21 in the time series data of the used resource amount of the server device 20 (# 2) shown in the example of FIG.
  • the time series data obtained by adding the time series data of the resource amount related to the virtual machine A shown in FIG. 22 is shown
  • FIG. 23 shows the load characteristics related to the time series data after the addition shown in the example of FIG. If the resource amount information of the virtual machine A does not exist, the resource amount of the virtual machine A may be estimated on the assumption that the virtual machine A always uses the requested resource amount ⁇ .
  • the resource amount ⁇ per unit time that can be secured on the server device 20 to be rearranged is the same as that of the virtual machine A.
  • the relocation determination unit 301 determines that the virtual machine B is also a relocatable virtual machine. Further, as in the case of the virtual machine A, the expected value calculation unit 106 sets the expected values that can secure all of the requested resource amounts of the virtual machine B by the free use type, in each of the server devices 20 (# 1) and 20 (# Calculate 2).
  • the requested resource amount was 0.5 (GHz), so the expected value calculated for server device 20 (# 2) is larger than that of server device 20 (# 1), and the result As a result, the server device 20 (# 2) is determined as the placement destination.
  • the requested resource amount is 1.0 (GHz)
  • both of the server apparatuses 20 (# 1) and 20 (# 2) Expected value decreases. In other words, in both the server apparatuses 20 (# 1) and 20 (# 2), it is determined that there is a possibility that sufficient resources may not be allocated to the virtual machine B in the free use type.
  • the selection unit 104 tries to apply the mixed resource allocation method because all the server apparatuses cannot secure resources in the free use type. If 0.5 (GHz) is assigned by the fixed allocation type in the server device, the resource amount that needs to be secured by the free resource is 0.5 (GHz). Therefore, since the situation is the same as that of the virtual machine A, the server device 20 (# 2) (FIG. 27) can secure the server, but the server device 20 (# 1) (FIG. 26) may not be able to secure it sufficiently. Recognize.
  • the fixed allocation amount is set to 0.5 (GHz) units, but may be changed to 0.1 (GHz) units.
  • the virtual machine C is determined as a virtual machine that cannot be relocated because the moving average width is 1 second. Since no resources are allocated to the virtual machine C during the relocation, the resource amount (CPU usage rate) allocated in 1 second becomes 0, and the service level of the virtual machine C can be maintained. Because it disappears.
  • the selection unit 104 determines a fixed allocation type server device that can secure a 1 (GHz) CPU resource as a placement destination candidate.
  • the server apparatus 20 (# 1) having the load characteristics shown in FIG. 28A and the server apparatus 20 (# 2) having the load characteristics shown in FIG. 28B have the load characteristics shown in FIG. A case where a virtual machine is arranged will be described as an example.
  • the server device 20 (# 1) the total resource usage appears with an average probability, whereas in the server device 20 (# 2), the resource usage is biased.
  • each of the load characteristic data of the server apparatuses 20 (# 1) and (# 2) is generated from time-series data averaged with the same average movement width as that of the virtual machine load characteristic data.
  • the value calculation unit 107 calculates the first value data by applying the above (formula 2) to the load characteristic data (FIGS. 28A and 28B) of the server devices 20 (# 1) and (# 2).
  • the predetermined division number L is held in advance so as to be adjustable. However, the amount of calculation increases as the predetermined division number L increases.
  • the predetermined division number L is set to four.
  • each total resource amount ⁇ already allocated in the fixed allocation type in the server apparatuses 20 (# 1) and (# 2) is 100. Accordingly, the division width ⁇ is calculated based on the above (Equation 1), and is set to 25 here.
  • the first value data of the server device 20 (# 1) is calculated as follows.
  • the first value data of the server device 20 (# 2) is calculated as follows.
  • the value calculation unit 107 multiplies the load characteristic data of the virtual machine (FIG. 29) and the load characteristic data (FIGS. 28A and 28B) of the server devices 20 (# 1) and (# 2), thereby The load characteristic data after placement of the virtual machines for the devices 20 (# 1) and (# 2) are calculated.
  • the value calculation unit 107 corrects each load characteristic data calculated in this way by using the division width ⁇ or the predetermined division number L.
  • the predetermined division number L is held in advance so as to be adjustable.
  • the predetermined division number L used here may be the same as or different from that used in calculating the first value data.
  • the predetermined division number L is set to four.
  • FIG. 30A shows the load characteristic data before correction of the server device 20 (# 1) after the placement of the virtual machine
  • FIG. 30B shows the load characteristic data before correction of the server device 20 (# 2) after the placement of the virtual machine.
  • the load characteristic data of the server apparatus 20 (# 2) after machine arrangement is shown.
  • the value calculation unit 107 calculates the second value data of the server devices 20 (# 1) and (# 2) by applying the load characteristic data thus corrected to the above (Equation 2). .
  • the selection unit 104 calculates the increase amount from the first value data corresponding to the virtual machine before the placement of the virtual machine calculated by the value calculation unit 107 to the second value data corresponding to the placement of the virtual machine, from the server device 20 (# 1). ) And (# 2).
  • the selection unit 104 determines the server device 20 (# 2) having a large increase amount as the placement destination of the virtual machine.
  • the right end of the horizontal axis of the value function f is set to the total amount ⁇ of resources already allocated in the fixed allocation type in the server device 20. It was. Thereby, the bias of the load characteristic is leveled with respect to the resources already allocated in the server device 20.
  • the right end of the horizontal axis of the value function f may be set to the total resource amount of the server device 20. In this case, it is possible to level the load characteristic bias with respect to the total resources including unallocated resources.
  • the fixed allocation amount is assumed on the assumption that any one of the resource allocation methods of the fixed allocation type, the free usage type, or the mixed type is applied when selecting the server device 20 where the virtual machine is to be arranged. And other resource amounts were determined.
  • the determined fixed allocation amount and other resource amounts may be notified as resource allocation information from the management device 10 to the server device 20 determined as the placement destination of the virtual machine.
  • the control unit 201 of the server device 20 may perform the resource allocation based on the resource allocation information sent from the management device 10. For example, regarding the CPU resource, the control unit 201 may determine the CPU resource use priority corresponding to the notified fixed allocation amount and set this for the virtual machine. In this way, it is possible to reduce the burden of setting work for resource control required in the server device 20.
  • the virtual machine placement determination method in each of the above-described embodiments may be applied to a scene where a virtual machine is newly placed in the server device 20, or a virtual machine that is already running on the server device 20 is rearranged. You may apply to the scene which determines a destination.
  • Each above-mentioned embodiment does not limit the application scene.

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

L'invention concerne un dispositif de gestion de machine virtuelle qui comprend : une unité d'analyse qui fait la moyenne d'unités de données en série chronologique pour une quantité de ressources utilisées par rapport à chaque dispositif serveur sur des durées moyennes de mouvement par rapport à des machines virtuelles, et sur la base de chacune des unités de données en série chronologique moyennes, par rapport à chaque dispositif serveur, génère des données caractéristiques de charge respectives indiquant une distribution de densité de probabilité pour la quantité de ressources utilisées ; et une unité de sélection pour sélectionner au moins un dispositif serveur parmi une pluralité de dispositifs serveurs en fonction des données caractéristiques de charge pour chaque dispositif serveur qui ont été générées par l'unité d'analyse et des quantités de ressources demandées pour les machines virtuelles.
PCT/JP2012/003793 2011-07-15 2012-06-11 Système de gestion de machine virtuelle et procédé de gestion de machine virtuelle WO2013011624A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2011-156925 2011-07-15
JP2011156925 2011-07-15

Publications (1)

Publication Number Publication Date
WO2013011624A1 true WO2013011624A1 (fr) 2013-01-24

Family

ID=47557828

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2012/003793 WO2013011624A1 (fr) 2011-07-15 2012-06-11 Système de gestion de machine virtuelle et procédé de gestion de machine virtuelle

Country Status (2)

Country Link
JP (1) JPWO2013011624A1 (fr)
WO (1) WO2013011624A1 (fr)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2015152984A (ja) * 2014-02-12 2015-08-24 日本電信電話株式会社 仮想マシン配置装置及び方法及びプログラム
KR101740490B1 (ko) 2015-12-29 2017-05-26 경희대학교 산학협력단 클라우드 컴퓨팅 환경에서의 사전 오토 스케일링 시스템 및 그 방법
CN107179949A (zh) * 2016-12-16 2017-09-19 重庆大学 一种用于移动设备中操作系统内存分配流畅度的量化方法
JP2020003929A (ja) * 2018-06-26 2020-01-09 富士通株式会社 運用管理装置、移動先推奨方法及び移動先推奨プログラム
KR102470081B1 (ko) * 2021-10-28 2022-11-23 오케스트로 주식회사 자원 적정화를 기반한 가상 머신 배치 시스템 및 가상 머신 배치 방법

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6550822B2 (ja) * 2015-03-20 2019-07-31 日本電気株式会社 情報処理資源管理装置、情報処理資源管理方法、及び、情報処理資源管理プログラム
JP2018190355A (ja) * 2017-05-11 2018-11-29 日本電信電話株式会社 リソース管理方法
JP7235296B2 (ja) * 2019-03-08 2023-03-08 Ｎｅｃソリューションイノベータ株式会社 セッション管理方法、セッション管理装置、プログラム

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010050249A1 (fr) * 2008-10-30 2010-05-06 株式会社日立製作所 Appareil de gestion de fonctionnement de système de traitement d'informations
JP2011013822A (ja) * 2009-06-30 2011-01-20 Nec Corp 情報システム、制御装置、そのデータ処理方法およびプログラム

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010050249A1 (fr) * 2008-10-30 2010-05-06 株式会社日立製作所 Appareil de gestion de fonctionnement de système de traitement d'informations
JP2011013822A (ja) * 2009-06-30 2011-01-20 Nec Corp 情報システム、制御装置、そのデータ処理方法およびプログラム

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2015152984A (ja) * 2014-02-12 2015-08-24 日本電信電話株式会社 仮想マシン配置装置及び方法及びプログラム
KR101740490B1 (ko) 2015-12-29 2017-05-26 경희대학교 산학협력단 클라우드 컴퓨팅 환경에서의 사전 오토 스케일링 시스템 및 그 방법
CN107179949A (zh) * 2016-12-16 2017-09-19 重庆大学 一种用于移动设备中操作系统内存分配流畅度的量化方法
CN107179949B (zh) * 2016-12-16 2020-11-24 重庆大学 一种用于移动设备中操作系统内存分配流畅度的量化方法
JP2020003929A (ja) * 2018-06-26 2020-01-09 富士通株式会社 運用管理装置、移動先推奨方法及び移動先推奨プログラム
JP7040319B2 (ja) 2018-06-26 2022-03-23 富士通株式会社 運用管理装置、移動先推奨方法及び移動先推奨プログラム
KR102470081B1 (ko) * 2021-10-28 2022-11-23 오케스트로 주식회사 자원 적정화를 기반한 가상 머신 배치 시스템 및 가상 머신 배치 방법

Also Published As

Publication number Publication date
JPWO2013011624A1 (ja) 2015-02-23

Similar Documents

Publication Publication Date Title
WO2013011624A1 (fr) Système de gestion de machine virtuelle et procédé de gestion de machine virtuelle
US10896055B2 (en) Capacity risk management for virtual machines
EP3161632B1 (fr) Attribution de ressources globales et équilibrage de charge intégrés
CN109643251B (zh) 基于计算系统中的利用模式的资源过度订阅
JP5400482B2 (ja) 管理計算機、リソース管理方法、リソース管理プログラム、記録媒体および情報処理システム
JP4982578B2 (ja) リソース割当装置、リソース割当方法、およびリソース割当制御プログラム
US20140282520A1 (en) Provisioning virtual machines on a physical infrastructure
US20160316003A1 (en) Balancing resources in distributed computing environments
US7206890B2 (en) System and method for reducing accounting overhead during memory allocation
US8104038B1 (en) Matching descriptions of resources with workload requirements
US7664859B2 (en) Resource assigning management apparatus and resource assigning method
US20160070602A1 (en) Workload distribution management apparatus and control method
US11496413B2 (en) Allocating cloud computing resources in a cloud computing environment based on user predictability
US11216311B1 (en) Resource overprovisioning in a virtual machine environment
CN104158841B (zh) 计算资源分配方法
CN110636388A (zh) 一种业务请求分配方法、系统、电子设备及存储介质
Chen et al. Considering resource demand misalignments to reduce resource over-provisioning in cloud datacenters
JP2010224754A (ja) リソース割当装置、リソース割当方法、及びプログラム
US9563532B1 (en) Allocation of tasks in large scale computing systems
JP2016110248A (ja) 仮想化実行装置、仮想化システム、および、リソース最適化方法
KR20160139082A (ko) 역경매 방식 자원할당 장치를 포함하는 하이브리드 클라우드 서버 및 그 자원 할당 방법 및 시스템
US9367439B2 (en) Physical memory usage prediction
JP2005309868A (ja) 処理割当管理装置、処理割当管理装置の制御方法、及びプログラム
Rumi et al. Optimization techniques within the hadoop eco-system: A survey
KR100946696B1 (ko) 타임 시리즈 기법을 이용한 스트리밍 소프트웨어 서비스시스템 및 그 방법

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: 12814725

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2013524582

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12814725

Country of ref document: EP

Kind code of ref document: A1