Device, assignment method, and program

Abstract

This virtual machine provisioning device provisions a virtual machine to a plurality of sites to which electric power is supplied through renewable energy generation, wherein the virtual machine provisioning device comprises: a prediction unit that acquires, for each site, a prediction value for the amount of power generated in the renewable energy generation and a prediction value for the amount of power consumed; and a provisioning unit that selects, from among predetermined time durations segmented by a plurality of time intervals at the plurality of sites, a site and a time interval for which surplus power in the renewable energy generation is maximum, that provisions a virtual machine selected from a control-target virtual machine group comprising one or more mobile virtual machines, to the selected site and the selected time interval, and that repeatedly executes processing to remove the provisioned virtual machine from the control-target virtual machine group.

Description

Technical Field

[0001] The present invention relates to a technique for allocating virtual machines to a plurality of bases that utilize electric power generated by renewable energy power generation.

Background Art

[0002] In recent years, various network services and applications have been provided by virtual machines (hereinafter referred to as VMs) operating on servers (physical machines) in a plurality of bases (for example, data centers).

[0003] In addition, for the purpose of reducing electricity purchase costs and decarbonization, etc., each base often receives two power supplies: renewable energy power generation such as solar power generation and commercial power supplied from an electric power company. Here, such a case is assumed.

[0004] In renewable energy power generation, the power supply amount fluctuates every moment. Therefore, for example, when the supply amount of renewable energy at base A decreases, by moving the VM at base A to base B where there is a surplus in the supply amount of renewable energy, the renewable energy at base B can be utilized without waste, and the amount of power purchased from the electric power company at base A can be reduced.

[0005] Non-Patent Document 1 discloses a vector-based algorithm for calculating where each VM should be placed in order to use the renewable energy of each base as efficiently as possible.

Prior Art Documents

Non-Patent Documents

[0006]

Non-Patent Document 1

[0007] However, the prior art disclosed in Non-Patent Document 1 does not take into account the surplus renewable energy at each site, which may result in the waste of renewable energy at certain times of day or at certain sites.

[0008] This invention has been made in view of the above points, and aims to provide a technology to minimize surplus electricity generated by renewable energy at multiple locations where electricity is supplied by renewable energy generation. [Means for solving the problem]

[0009] According to the disclosed technology, a device for allocating virtual machines to multiple locations where power is supplied, Within a predetermined time period at the aforementioned multiple locations, the time when the surplus power from renewable energy generation is maximized is selected, and a virtual machine selected from a group of controllable virtual machines consisting of one or more mobile virtual machines is allocated to the selected time. Then, the assigned virtual machine is removed from the group of controlled virtual machines. The device is provided. [Effects of the Invention]

[0010] According to the disclosed technology, it is possible to reduce surplus electricity generated from renewable energy sources at multiple locations where electricity is supplied by renewable energy generation. [Brief explanation of the drawing]

[0011] [Figure 1] This is an overall system configuration diagram in an embodiment of the present invention. [Figure 2] This is a configuration diagram of a virtual machine allocation device in an embodiment of the present invention. [Figure 3] This is a flowchart illustrating the operation in Example 1. [Figure 4] This is a diagram to explain S101. [Figure 5] This is a diagram to explain S102. [Figure 6] This is a diagram to explain S103. [Figure 7] This is a diagram to explain S103. [Figure 8] This is a diagram to explain S104-S105. [Figure 9] This is a diagram to explain S106-S108. [Figure 10] This is a flowchart illustrating the operation in Example 2. [Figure 11] This figure shows an example of the device's hardware configuration. [Modes for carrying out the invention]

[0012] Hereinafter, embodiments of the present invention (this embodiment) will be described with reference to the drawings. The embodiments described below are merely examples, and the embodiments to which the present invention is applied are not limited to the embodiments described below.

[0013] (System configuration, operation overview) Figure 1 shows an example of the overall system configuration in this embodiment. As shown in Figure 1, multiple locations equipped with server clusters are connected to the physical network 200. In addition, the virtual machine allocation device 100 is connected to the physical network 200.

[0014] At each site, commercial power supplied by the power company is supplied, and power generated by renewable energy is also supplied. Note that there may be sites where only commercial power supplied by the power company is supplied, or sites where only power generated by renewable energy is supplied.

[0015] Each server at each site is equipped with 0 or more VMs. Each VM can be moved between sites by a setting command from the virtual machine allocation device 100. For the movement of VMs (which may also be called migration), various existing technologies can be used.

[0016] The virtual machine allocation device 100 derives the allocation (assignment) of each VM to each site in consideration of the surplus power of renewable energy generation at each site and the power consumption of each VM, and transmits a setting command to move the VM to the corresponding site according to the result of the allocation process. The virtual machine allocation device 100 can obtain information (such as the amount of renewable energy generation and the power consumption of each VM) necessary for the allocation calculation from each site (or a monitoring system, etc.).

[0017] Prior to deriving the allocation, from the perspective of minimizing the surplus power of renewable energy generation as much as possible, the virtual machine allocation device 100 first selects the VMs to be moved among all the VMs. In deriving the allocation, it is assumed that an approximate guarantee is obtained by sequentially allocating the VMs with the largest power consumption to the site and time period with the largest surplus power.

[0018] (Device configuration example) Fig. 2 shows a configuration example of the virtual machine allocation device 100 in the present embodiment. As shown in Fig. 2, the virtual machine allocation device 100 includes a schedule management unit 110, a power generation amount prediction unit 120, a power consumption prediction unit 130, a physical resource acquisition unit 140, a traffic prediction unit 150, a service request management unit 160, an allocation unit 170, and a setting command unit 180. Note that the power generation amount prediction unit 120 and the power consumption prediction unit 130 may be collectively referred to as a prediction unit.

[0019] The schedule management unit 110 manages the schedule for executing the control pattern calculated by the allocation unit 170 based on the update status of the prediction information and the control time constant of the physical network 200. For example, if it is known from the control time constant that control takes 30 minutes, and the desired VM placement is to be achieved at a certain time T, the schedule is managed to start the control (VM placement calculation, command transmission) 30 minutes before time T.

[0020] The power generation forecasting unit 120 forecasts the amount of power generated by renewable energy power generation facilities at each location based on weather information and other data. The power consumption forecasting unit 130 forecasts the power consumption of each mobile load and non-mobile load based on past power consumption data and other data.

[0021] The physical resource acquisition unit 140 acquires and manages information related to physical resources (such as the CPU capacity of each site and the link bandwidth of each link).

[0022] The traffic prediction unit 150 acquires the traffic volume (predicted traffic volume) during the time virtual network control is performed. The service request management unit 160 manages service requests.

[0023] The allocation unit 170 calculates the optimal virtual network control pattern according to a procedure described later, for example, in response to instructions from the schedule management unit 110.

[0024] The configuration command unit 180 sends virtual network configuration commands to each physical resource (for example, servers and various communication devices) in order to realize the optimal virtual network control pattern calculated by the allocation unit 170.

[0025] The virtual machine allocation device 100 may consist of one device (computer) or multiple devices. Furthermore, the virtual machine allocation device 100 may only include the function of determining the allocation of VMs to locations, with other devices performing the configuration commands for the actual VM movement. Also, for example, the virtual machine allocation device 100 may include a prediction unit and an allocation unit 170, with other functions provided outside the virtual machine allocation device 100.

[0026] (Example 1) Next, the detailed processing by the virtual machine allocation device 100 in Example 1 will be explained following the steps in the flowchart of Figure 3. Figures 4 to 9 will be referred to as appropriate in the explanation of the procedure.

[0027] In this embodiment, a predetermined time length is divided into multiple time intervals, and a VM is allocated to each time interval. The predetermined time length is, for example, 24 hours. Also, a 1-hour interval is, for example, 30 minutes. A time interval is called a "frame" or "time frame."

[0028] Furthermore, in this embodiment, it is assumed that control to each location will be performed at predetermined time intervals (for example, every 3 hours), and before the predetermined time for actual control, allocations will be made in units of a predetermined time length (e.g., 24 hours). In the diagrams used in the following explanation, for convenience, units with a time length of approximately 3 to 4 hours are used to make the diagrams easier to understand.

[0029] <s101> In step S101, the power consumption prediction unit 130 obtains predicted power consumption values ​​for the mobile load and non-mobile load for each location and each time slot. The power generation prediction unit 120 also obtains predicted power generation values ​​for renewable energy for each location and each time slot. The acquired power information is stored in a memory or other storage unit of the virtual machine allocation device 100 and is read from memory and used for processing in subsequent operations.

[0030] A mobile load is a load whose power consumption can be moved between locations, and corresponds to a VM (Virtual Mass). A non-mobile load is a load whose power consumption cannot be moved between locations, and corresponds to, for example, the air conditioning system of a building. Hereafter, the power consumption of the mobile load will be referred to as mobile load power, and the power consumption of the non-mobile load will be referred to as non-mobile load power. In addition, in the following explanation, unless otherwise specified, power will be assumed to be a predicted value even if it is not explicitly stated to be a predicted value.

[0031] Figure 4 shows an example of information about bases A, B, and C acquired in S101. As shown in the figure, mobile load power, non-mobile load power, and power generation amount are acquired for each base. In Figure 4, each square in the mobile load power section represents the mobile load power of one VM in that section. It can also be seen that each square in the mobile load power section represents one VM in that section.

[0032] <s102> In S102, the allocation unit 170 manages the VM groups that are the source of the mobile load power at all locations as the controlled VM group. In other words, for example, it stores information about the controlled VM group in memory. Figure 5 shows the mobile load power for each segment of the controlled VM group. In Figure 5, the shading indicates which of the locations A to C shown in Figure 4 the VM (mobile load power) belongs to.

[0033] <s103> In S103, the allocation unit 170 excludes VMs from the group of VMs to be controlled based on the predicted amount of renewable energy generation, which will generate (or increase) surplus renewable energy when moved.

[0034] Let's illustrate with an example referring to Figure 6. In case (a) of Figure 6, the VMs (mobile load power) shown as A, B, and C each use electricity generated from renewable energy sources. Therefore, if any of these are moved to another location, surplus renewable energy power will be generated. The same applies to D, E, and F in (b).

[0035] In the case shown in Figure 4, for bases A and B, the VMs (moving load power) in the three central squares correspond to those that generate surplus renewable energy when moved, so these are excluded from the group of VMs to be controlled. Figure 7 shows the group of VMs to be controlled after excluding these from the total group of VMs to be controlled shown in Figure 5.

[0036] <s104> In S104, the allocation unit 170 compares the amount of power generated with "non-moving load power + moving load power" for each base and each slot to determine whether there is surplus power. Here, moving load power is the sum of the power of VMs that have been removed from the control target (i.e., the power of VMs that are not moved) and the power of VMs that have been reassigned in the allocation process described later. Figure 8 shows the "non-moving load power + moving load power" at this point for bases A and B.

[0037] The allocation unit 170 determines that there is surplus power if "power generation > (non-mobile load power + mobile load power)".

[0038] <s105> In S105, the allocation unit 170 selects the location / time slot with the largest surplus renewable energy generation for each time slot at all locations. In the example shown in Figure 8, the surplus power at location A in the indicated time slot is the largest among all locations / time slots, so it is selected.

[0039] <S106~S108> In S106, the allocation unit 170 selects the VM with the highest power consumption among the time slots (frames) selected in S105 from the group of VMs to be controlled.

[0040] In S107, the allocation unit 170 allocates the VMs selected in S106 to the locations selected in S105, and removes the VMs from the group of VMs to be controlled. In S108, the power of the allocated VMs is added to the power consumption of the location.

[0041] Figure 9 shows a specific example of steps S106 to S108. In the example in Figure 9, the VM with the largest power among the slots selected in S105 is the VM indicated by A, which is allocated to base A, and its power is added to the corresponding slot.

[0042] <S104、S109> The allocation unit 170 repeats operations S105 to S108 until there are no more VMs in the controlled VM group or until there is no surplus power at all locations.

[0043] <s110> In S110, the allocation unit 170 allocates the VMs (mobile load power) in the controlled VM group after the surplus power has been depleted back to their original locations.

[0044] Subsequently, when the control time managed by the schedule management unit 110 arrives, the setting command unit 180 sends setting commands to each location based on the allocation results calculated by the allocation unit 170.

[0045] (Example 2) Next, we will describe Example 2. In Example 2, after the placement of the controlled VM group to locations with surplus power has been completed, when the remaining VMs of the controlled VM group are placed at locations, the placement is designed to improve communication quality.

[0046] Specifically, in Example 2, S110 in Figure 3 is replaced by S111 to S114 shown in Figure 10. S101 to S109 are the same as in Example 1. The processing details will be explained according to the steps in the flowchart of Figure 10.

[0047] <s111> In S111, the allocation unit 170 randomly selects one VM from the remaining group of controlled VMs to determine the destination.

[0048] <s112> In S112, the allocation unit 170 determines the location with the smallest round-trip delay time as the destination for the VM selected in S111. Here, the round-trip delay time is, for example, the round-trip delay time of communication between the user terminal accessing the VM and the VM. The round-trip delay time may be estimated based on the distance between the user terminal and the VM, or it may be obtained or estimated by previously measured values ​​for communication between the user terminal and the location, or by other methods.

[0049] <s113> In S113, the allocation unit 170 removes the VM whose placement destination has been determined from the group of VMs to be controlled.

[0050] <s114> In S114, the allocation unit 170 determines whether there are any VMs remaining in the controlled VM group. If there are, it returns to S111; otherwise, it terminates the process. The allocation unit 170 repeats steps S111 to S114 until there are no more VMs left in the controlled VM group.

[0051] In the example above, round-trip delay time was used as an example of a communication quality parameter, but this is not the only one. For example, the VM destination could be determined by considering the server resources (CPU, memory, storage, etc.) at each location and the available bandwidth of each link. Alternatively, multiple parameters could be incorporated to determine the VM destination.

[0052] The physical resource acquisition unit 140 acquires server resources, and the traffic prediction unit 150 acquires available bandwidth.

[0053] For example, VMs used to run applications that process large amounts of data are placed at locations where the remaining memory is greater than the threshold, and VMs used to run applications that require large amounts of communication are placed at locations where the available bandwidth is greater than the threshold.

[0054] (Example hardware configuration) The virtual machine allocation device 100 can be implemented, for example, by having a computer run a program. This computer may be a physical computer or a virtual machine on the cloud.

[0055] In other words, the virtual machine allocation device 100 can be realized by using hardware resources such as the CPU and memory built into the computer to execute a program corresponding to the processing performed by the virtual machine allocation device 100. The above program can be recorded on a computer-readable recording medium (such as portable memory), saved, and distributed. It is also possible to provide the above program via a network such as the Internet or email.

[0056] Figure 11 shows an example of the hardware configuration of the computer described above. The computer in Figure 11 has a drive device 1000, an auxiliary storage device 1002, a memory device 1003, a CPU 1004, an interface device 1005, a display device 1006, an input device 1007, an output device 1008, etc., all of which are interconnected by a bus BS.

[0057] The program that enables processing on the computer is provided, for example, on a recording medium 1001 such as a CD-ROM or memory card. When the recording medium 1001 containing the program is set in the drive device 1000, the program is installed from the recording medium 1001 to the auxiliary storage device 1002 via the drive device 1000. However, the program does not necessarily have to be installed from the recording medium 1001; it may also be downloaded from another computer via a network. The auxiliary storage device 1002 stores the installed program as well as necessary files and data.

[0058] The memory device 1003 reads and stores a program from the auxiliary storage device 1002 when a program startup command is received. The CPU 1004 implements the functions related to the virtual machine allocation device 100 according to the program stored in the memory device 1003. The interface device 1005 is used as an interface for connecting to a network, etc. The display device 1006 displays a GUI (Graphical User Interface) etc. generated by a program. The input device 1007 consists of a keyboard and mouse, buttons, or a touch panel etc., and is used to input various operation commands. The output device 1008 outputs the calculation results.

[0059] (Effects of the embodiment) The technology according to this embodiment makes it possible to reduce surplus renewable energy at all locations and to supply more electricity with renewable energy.

[0060] Furthermore, increasing the proportion of renewable energy will contribute to corporate social responsibility (CSR) initiatives and help reduce electricity purchase costs from power companies.

[0061] (Note) This specification discloses at least the following virtual machine allocation devices, virtual machine allocation methods, and programs. (Section 1) A virtual machine allocation device that allocates virtual machines to multiple locations where electricity is supplied by renewable energy generation, A prediction unit that acquires predicted values ​​for renewable energy generation and predicted values ​​for power consumption for each site, An allocation unit repeatedly performs the following processes: selecting a location and time interval within a predetermined time period divided into multiple time intervals at the aforementioned locations where the surplus power from renewable energy generation is maximized; assigning a virtual machine selected from a group of controllable virtual machines consisting of one or more mobile virtual machines to the selected location and time interval; and removing the assigned virtual machine from the group of controllable virtual machines. A virtual machine allocation device equipped with the following features. (Section 2) The allocation unit designates one or more virtual machines from the virtual machines at each time interval across the multiple locations as the group of virtual machines to be controlled, excluding virtual machines that generate surplus renewable energy when moved. The virtual machine allocation device described in paragraph 1. (Section 3) The allocation unit selects the virtual machine with the highest power consumption from the group of controlled virtual machines within the selected time interval, and allocates that virtual machine to the selected site and time interval. The virtual machine allocation device described in paragraph 2. (Section 4) When there is no surplus power at any of the locations, the allocation unit allocates each virtual machine remaining in the controlled virtual machine group to its original location. A virtual machine allocation device as described in any one of paragraphs 1 through 3. (Section 5) When there is no surplus power at any of the locations, the allocation unit determines the location to which each virtual machine remaining in the controlled virtual machine group will be allocated in a manner that improves the communication quality related to that virtual machine. A virtual machine allocation device as described in any one of paragraphs 1 through 3. (Section 6) A virtual machine allocation method executed by a virtual machine allocation device that allocates virtual machines to multiple locations where electricity is supplied by renewable energy generation, A prediction step to obtain predicted values ​​for renewable energy generation and predicted values ​​for power consumption for each location, The assignment step involves repeatedly performing the following steps: selecting a location and time interval within a predetermined time period divided into multiple time intervals at the aforementioned locations where the surplus power from renewable energy generation is maximized; assigning a virtual machine selected from a group of controllable virtual machines consisting of one or more mobile virtual machines to the selected location and time interval; and removing the assigned virtual machine from the group of controllable virtual machines. A virtual machine allocation method comprising the following features. (Section 7) A program for causing a computer to function as a component of a virtual machine allocation device as described in any one of paragraphs 1 through 5.

[0062] Although this embodiment has been described above, the present invention is not limited to this specific embodiment, and various modifications and changes are possible within the scope of the gist of the invention as described in the claims. [Explanation of Symbols]

[0063] 100 Virtual Machine Allocation Devices 110 Schedule Management Department 120 Power generation forecasting unit 130 Power Consumption Prediction Unit 140 Physical Resource Acquisition Unit 150 Traffic Prediction Unit 160 Service Request Management Department 170 Allocation Section 180 Setting command section 200 physical networks 1000 drive unit 1001 Recording media 1002 Auxiliary storage 1003 Memory device 1004 CPU 1005 Interface device 1006 Display device 1007 Input device 1008 Output device

Claims

1. A device for allocating virtual machines to multiple locations where power is supplied, A device that selects a time within a predetermined time period at the aforementioned multiple locations when the surplus power from renewable energy generation is maximized, assigns a virtual machine selected from a group of controllable virtual machines consisting of one or more mobile virtual machines to the selected time, and removes the assigned virtual machine from the group of controllable virtual machines.

2. The device further selects a location where the surplus power from renewable energy generation is maximized, and assigns a virtual machine selected from the group of controllable virtual machines consisting of one or more mobile virtual machines to the selected location and time. The apparatus according to claim 1.

3. The group of controlled virtual machines is one or more virtual machines obtained by excluding virtual machines that generate surplus electricity from renewable energy generation when moved from the virtual machines at the plurality of locations. The apparatus according to claim 1 or 2.

4. A device for allocating virtual machines to multiple locations where electricity is supplied by renewable energy generation, A prediction unit that acquires predicted values ​​for renewable energy generation and predicted values ​​for power consumption for each site, An allocation unit repeatedly performs the following processes: selecting a location and time interval within a predetermined time period divided into multiple time intervals at the aforementioned locations where the surplus power from renewable energy generation is maximized; assigning a virtual machine selected from a group of controllable virtual machines consisting of one or more mobile virtual machines to the selected location and time interval; and removing the assigned virtual machine from the group of controllable virtual machines. A device equipped with the following features.

5. The allocation unit designates one or more virtual machines from the virtual machines at each time interval in the multiple locations as the group of virtual machines to be controlled, excluding virtual machines that generate surplus renewable energy when moved. The apparatus according to claim 4.

6. The allocation unit selects the virtual machine with the highest power consumption from the group of controlled virtual machines within the selected time interval, and allocates that virtual machine to the selected site and time interval. The apparatus according to claim 5.

7. When there is no surplus power at any of the locations, the allocation unit allocates each virtual machine remaining in the controlled virtual machine group to its original location. The apparatus according to any one of claims 4 to 6.

8. When there is no surplus power at any of the locations, the allocation unit determines the location to which each virtual machine remaining in the controlled virtual machine group will be allocated in a manner that improves the communication quality related to that virtual machine. The apparatus according to any one of claims 4 to 6.

9. An allocation method performed by a device that allocates virtual machines to multiple locations where power is supplied, Within a predetermined time period at the aforementioned multiple locations, the time when the surplus power from renewable energy generation is maximized is selected, and a virtual machine selected from a group of controlled virtual machines consisting of one or more movable virtual machines is assigned to the selected time, and the assigned virtual machine is removed from the group of controlled virtual machines. Allocation method.

10. The group of virtual machines to be controlled is one or more virtual machines obtained by excluding virtual machines that generate surplus electricity from renewable energy generation when moved from the virtual machines at the plurality of locations. The assignment method according to claim 9.

11. A program for causing a computer to function as the device described in any one of claims 1 to 8.

Images