JP2018206062A - Control program, control device, and control method - Google Patents

Abstract

To provide a control program, a control device, and a control method capable of leveling the load of each physical machine.SOLUTION: Based on load information of a plurality of physical machines, a physical machine of an arrangement destination, at which a specific virtual machine is arranged to distribute a request accepted by any one of a plurality of virtual machines respectively built in any one of the plurality of physical machines, is specified. Then, based on load information of the specified physical machine, the number of virtual machines to which the request accepted by the specific virtual machine can be distributed, out of the plurality of virtual machines, is calculated. According to the arrangement of the specific virtual machine to the specified physical machine, the number of virtual machines corresponding to the calculated number, out of the plurality of virtual machines, are designated as virtual machines of a distribution destination to which the request accepted by the specific virtual machine is distributed.SELECTED DRAWING: Figure 9

Description

  The present invention relates to a control program, a control device, and a control method.

  For example, a business provider (hereinafter simply referred to as a business operator) that provides a service to a user constructs and operates a business system (hereinafter also referred to as an information processing system) for providing the service.

  Specifically, for example, a business operator can process a plurality of virtual machines (hereinafter also referred to as VM: Virtual Machine) that perform processing in response to a processing request from a user, and process requests from a user to a plurality of virtual machines. A business system having a virtual machine (hereinafter also referred to as a load balancer) that performs distribution is constructed. Then, for example, the load balancer refers to the current load of each virtual machine, and causes the virtual machine with a relatively low load to perform processing in response to a processing request from the user. Accordingly, the business operator can provide a service to the user while leveling the load on each virtual machine constituting the business system (see, for example, Patent Documents 1 to 4).

International Publication No. 2011/114623 Japanese Patent Laid-Open No. 2015-022501 Special table 2013-506908 gazette JP 2006-0486680 A

  In the business system as described above, when the number of processing requests from the user increases, there is a case where the processing requests cannot be distributed to each virtual machine with one load balancer.

  Therefore, for example, the business operator prepares a plurality of load distribution devices as necessary. The operator causes one of the plurality of load distribution apparatuses to function as a load distribution apparatus (hereinafter also simply referred to as a management apparatus) that distributes processing requests from users to other load distribution apparatuses. Further, the business operator causes each of the load distribution apparatuses other than the management apparatus to function as a load distribution apparatus that distributes the processing request distributed from the management apparatus to each virtual machine. Thereby, even if a plurality of load balancers operate, the business can provide services to the user while leveling the load on each virtual machine constituting the business system. .

  Here, for example, the management apparatus may add or delete a load distribution apparatus in addition to the distribution of processing requests from users. Specifically, the management apparatus acquires information indicating the current load (hereinafter also referred to as load information) from each physical machine on which each virtual machine operates. Then, for example, when there is a physical machine whose current load is higher than a predetermined threshold, the management apparatus determines that the load of the load balancer operating on the physical machine is high, and operates on the physical machine. Add a load balancer. As a result, the management apparatus can determine whether or not a load balancer should be added to each physical machine without acquiring load information for each virtual machine. The number of virtual machines to be distributed can be adjusted.

  However, in a physical machine on which each virtual machine operates, a virtual machine (hereinafter also referred to as another virtual machine) in which the load distribution apparatus does not distribute processing requests may operate. Therefore, when adding a load distribution device based on the load information for each physical machine, the management device may add more load distribution devices than necessary, and the processing load of a physical machine with a large load increases further. There is a case.

  Accordingly, in one aspect, an object of the present invention is to provide a control program, a control apparatus, and a control method that can level the load of each physical machine.

  In one aspect of the embodiment, a specific virtual machine that distributes a request received to any of a plurality of virtual machines, each of which is constructed on any of a plurality of physical machines, is disposed on any of the plurality of physical machines In the control program, the physical machine of the placement destination where the specific virtual machine is arranged is specified based on the load information of the plurality of physical machines, and the plurality of virtual machines are determined based on the load information of the specified physical machine. Among them, the number of virtual machines that can distribute the request received by the specific virtual machine is calculated, and the calculation is performed among the plurality of virtual machines according to the arrangement of the specific virtual machine on the specified physical machine. The number of virtual machines according to the number is set as a distribution-destination virtual machine that distributes requests received by the specific virtual machine. To be executed by a computer.

  According to one aspect, the load on each physical machine can be leveled.

FIG. 1 is a diagram illustrating an overall configuration of the information processing system 10. FIG. 2 describes a specific example of processing in the management apparatus 1. FIG. 3 illustrates a specific example of processing in the management apparatus 1. FIG. 4 describes a specific example of processing in the management apparatus 1. FIG. 5 is a diagram illustrating a hardware configuration of the management apparatus 1. FIG. 6 is a diagram illustrating a hardware configuration of the physical machine 2. FIG. 7 is a block diagram of functions of the management apparatus 1. FIG. 8 is a block diagram of functions realized by the load balancer 3 a among the functions of the physical machine 2. FIG. 9 is a flowchart for explaining the outline of the control processing in the first embodiment. FIG. 10 is a diagram for explaining the outline of the control processing in the first embodiment. FIG. 11 is a diagram for explaining the outline of the control processing in the first embodiment. FIG. 12 is a diagram for explaining the outline of the control processing in the first embodiment. FIG. 13 is a diagram for explaining the outline of the control processing in the first embodiment. FIG. 14 is a diagram for explaining the outline of the control processing in the first embodiment. FIG. 15 is a diagram for explaining the outline of the control processing in the first embodiment. FIG. 16 is a flowchart for explaining the details of the control processing in the first embodiment. FIG. 17 is a flowchart for explaining the details of the control processing in the first embodiment. FIG. 18 is a flowchart for explaining the details of the control processing in the first embodiment. FIG. 19 is a flowchart for explaining the details of the control processing in the first embodiment. FIG. 20 is a flowchart for explaining the details of the control processing in the first embodiment. FIG. 21 is a flowchart for explaining the details of the control processing in the first embodiment. FIG. 22 is a flowchart for explaining the details of the control processing in the first embodiment. FIG. 23 is a flowchart for explaining the details of the control processing in the first embodiment. FIG. 24 is a diagram illustrating specific examples of the distribution management information 131 and the host management information 132. FIG. 25 is a diagram illustrating a specific example of the VM management information 133. FIG. 26 is a diagram illustrating specific examples of the distribution management information 131 and the host management information 132. FIG. 27 is a diagram illustrating a specific example of the VM management information 133. FIG. 28 is a diagram illustrating specific examples of the distribution management information 131 and the host management information 132. FIG. 29 is a diagram illustrating specific examples of the distribution management information 131 and the host management information 132.

[Configuration of information processing system]
FIG. 1 is a diagram illustrating an overall configuration of the information processing system 10. An information processing system 10 illustrated in FIG. 1 is a business system for providing a service to a user, for example. In the information processing system 10 shown in FIG. 1, a management apparatus 1 and a physical machine 2 are provided in a data center (not shown). The business entity terminal 5 is accessible to the data center via a network such as the Internet or an intranet.

  The physical machine 2 is composed of a plurality of physical machines, for example. Each physical machine has a CPU (Central Computing Unit), a memory (DRAM: Dynamic Random Access Memory), and a large-capacity memory such as a hard disk (HDD: Hard Disk Drive). The physical resources of the physical machine 2 are allocated to a plurality of virtual machines 3 that execute processes for providing services to users.

  The virtualization software 4 is basic software that operates the virtual machine 3 by assigning the CPU, memory, and hard disk of the physical machine 2. The virtualization software 4 operates on the physical machine 2, for example.

  For example, when the user receives a processing request transmitted via a user terminal (not shown), the management device 1 distributes the received processing request to one of the virtual machines 3. The management device 1 may be composed of, for example, a virtual machine 3. Hereinafter, a specific example of processing of the management apparatus 1 will be described.

[Specific example of processing in management device]
2 to 4 describe specific examples of processing in the management apparatus 1. In the following description, it is assumed that the physical machine 2 includes physical machines 21, 22, and 23. Also, a part of the virtual machine 3 operating on the physical machines 21, 22, and 23 distributes processing requests to other virtual machines 3, load distribution devices 31 a, 32 a, and 33 a (hereinafter collectively referred to as load distribution). The following description will be made assuming that each device operates as a device 3a. Further, the load balancer 31a distributes the processing requests to the VMs 31b, 31c, and 31d, the load balancer 32a distributes the processing requests to the VMs 32b, 32c, and 32d, and the load balancer 33a uses the VMs 33b and 33c. In the following description, processing requests are distributed.

  When the management device 1 receives a processing request from a user terminal (not shown) or the like, the management device 1 distributes the received processing request to any one of the load distribution devices 31a, 32a, and 33a.

  Specifically, as illustrated in FIG. 2, the management apparatus 1 acquires information (hereinafter also simply referred to as load information) indicating the load of each physical machine as a whole from the physical machines 21, 22, and 23, for example. Then, for example, the management device 1 identifies the physical machine 2 with the smallest current load based on the acquired load information. Thereafter, as shown in FIG. 3, the management device 1 distributes processing requests to, for example, the load distribution device 3 a operating in the identified physical machine 2. As a result, the management device 1 can level the load on each physical machine 2.

  Further, the management device 1 determines whether or not it is necessary to add or delete a load balancer in each physical machine 2 at regular timing, for example.

  Specifically, as shown in FIG. 3, for example, the management apparatus 1 determines whether or not there is a physical machine 2 in which the current load exceeds a predetermined threshold value based on each acquired load information. judge. When there is a physical machine 2 that exceeds a predetermined threshold, the management apparatus 1 determines that the load of the load balancer 3a that operates in the physical machine 2 is high, and a new load balancer 3a. Add. As a result, the management device 1 can prevent a response to a processing request from being delayed due to, for example, a high load on the load distribution device 3a. Further, in this case, the management device 1 can determine whether or not to add a new load distribution device 3a without acquiring the load information of each virtual machine 3 generated in each physical machine 2. It becomes possible.

  Here, in the physical machine 2, a virtual machine 3 (other virtual machine 3) in which the load distribution device 3a does not distribute the processing request may be generated. Specifically, in the example illustrated in FIG. 4, VMs 31 e, 31 f and 31 g which are other virtual machines 3 are generated on the physical machine 21, and a VM 32 e which is another virtual machine 3 is generated on the physical machine 22. ing. Therefore, when adding the load distribution device 3a based on the load information of each physical machine 2, the management device 1 may add more load distribution devices 3a than necessary, and the processing load of each physical machine 2 is increased. May increase more.

  Therefore, the management apparatus 1 according to the present embodiment distributes the processing request received to any of the plurality of virtual machines 3 constructed in any of the plurality of physical machines 2 based on the load information of the plurality of physical machines 2. An arrangement destination physical machine 2 in which a specific load distribution apparatus 3a (hereinafter also referred to as a specific virtual machine 3) is arranged is specified. Based on the load information of the identified physical machine 2, the management apparatus 1 calculates the number of virtual machines 3 that can distribute the processing requests received by the specific load distribution apparatus 3 among the plurality of virtual machines 3.

  Furthermore, the management apparatus 1 assigns the number of virtual machines 3 corresponding to the calculated number out of the plurality of virtual machines 3 according to the arrangement of the specific load distribution apparatus 3 a to the specified physical machine 2. A distribution destination virtual machine 3 that distributes a processing request received by the distribution apparatus 3a is assumed.

  That is, when adding a new load balancer 3a, the management device 1 adds a new load balancer 3a to a physical machine 2 having a relatively small load. Then, the management apparatus 1 distributes the virtual machine 3 (part of the virtual machine 3) to which the load distribution apparatus 3a generated in the physical machine 2 having a large load distributes the processing request to the new load distribution apparatus. Let 3a do.

  As a result, the management apparatus 1 can prevent a new load distribution apparatus 3a from being added to the physical machine 2 with a large load. In addition, the management apparatus 1 can suppress the load on the physical machine 2 having a large load by reducing the number of virtual machines 3 to which the load distribution apparatus 3a generated in the physical machine 2 having a large load distributes. Become. Therefore, the management apparatus 1 can level the load on each physical machine 2.

[Hardware configuration of information processing system]
Next, the hardware configuration of the information processing system 10 will be described. FIG. 5 is a diagram illustrating a hardware configuration of the management apparatus 1. FIG. 6 is a diagram illustrating a hardware configuration of the physical machine 2.

  As illustrated in FIG. 5, the management apparatus 1 includes a CPU 101 that is a processor, a memory 102, an external interface (hereinafter also referred to as an I / O unit) 103, and a storage medium 104. Each unit is connected to each other via a bus 105.

  The storage medium 104 stores, for example, a program 110 for performing processing (hereinafter also referred to as control processing) for managing the load distribution device 3a in a program storage area (not shown) in the storage medium 104. In addition, the storage medium 104 includes, for example, an information storage area 130 (hereinafter also referred to as a storage unit 130) that stores information used when performing control processing.

  When executing the program 110, the CPU 101 loads the program 110 from the storage medium 104 to the memory 102 and performs control processing in cooperation with the program 110. The external interface 103 communicates with the physical machine 2, for example.

  Further, as illustrated in FIG. 6, the physical machine 2 includes a CPU 201 that is a processor, a memory 202, an external interface (hereinafter also referred to as an I / O unit) 203, and a storage medium 204. Each unit is connected to each other via a bus 205.

  The storage medium 204 stores, for example, a program 210 for performing the control device in a program storage area (not shown) in the storage medium 204. In addition, the storage medium 204 includes, for example, an information storage area 230 (hereinafter also referred to as a storage unit 230) that stores information used when performing control processing.

  When executing the program 210, the CPU 201 loads the program 210 from the storage medium 204 to the memory 202 and performs control processing in cooperation with the program 210. The external interface 203 communicates with, for example, the management device 1 and the operator terminal 5.

[Information processing system functions]
Next, functions of the information processing system 10 will be described. FIG. 7 is a block diagram of functions of the management apparatus 1. FIG. 8 is a block diagram of functions realized by the load balancer 3 a among the functions of the physical machine 2.

[Management device functions]
First, the function of the management apparatus 1 will be described. As shown in FIG. 7, the CPU 101 of the management apparatus 1 cooperates with the program 110 to process the process request accepting unit 111, the process request distributing unit 112, the information acquiring unit 113, the information managing unit 114, It operates as an arrangement destination specifying unit 115, a machine number calculation unit 116, and a distribution destination determination unit 117. Further, in the information storage area 130, for example, distribution management information 131, host management information 132, VM management information 133, and threshold information 134 are stored.

  For example, the processing request reception unit 111 receives a processing request transmitted from a user terminal (not shown). Then, for example, the processing request distribution unit 112 distributes the processing request received by the processing request reception unit 111 to one of the load distribution devices 3a.

  For example, the information acquisition unit 113 acquires load information from each physical machine 2 at regular timing. Then, the information management unit 114 updates the distribution management information 131, the host management information 132, and the VM management information 133 stored in the information storage area 130 based on the load information acquired by the information acquisition unit 113. The distribution management information 131 is information related to the management device 1 and each load distribution device 3a, and the host management information 132 is information related to each physical machine 2. The VM management information 133 is information regarding each virtual machine 3 to which the load distribution apparatus 3a distributes processing requests.

  Based on the load information of each physical machine 2, the placement destination specifying unit 115 specifies the placement destination physical machine 2 on which the new load distribution device 3a is placed. Specifically, for example, when there is a physical machine 2 whose current load exceeds the threshold information 134, the placement destination specifying unit 115 specifies the placement destination physical machine 2 in which the new load distribution device 3a is placed. .

  Based on the load information of the physical machine 2 specified by the placement destination specifying unit 115, the number-of-machines calculation unit 116 calculates the number of virtual machines 3 among the virtual machines 3 to which the new load balancer 3a can distribute processing requests. calculate.

  The distribution destination determination unit 117 sets the number calculated by the number-of-machines calculation unit 116 of each virtual machine 3 according to the arrangement of the new load distribution device 3a on the physical machine 2 specified by the arrangement destination specification unit 115. The corresponding number of virtual machines 3 are determined as virtual machines 3 to which the new load distribution device 3a distributes processing requests.

[Physical machine (load balancer) functions]
Next, of the functions of the physical machine 2, functions realized by the load distribution device 3a will be described. The CPU 201 of the physical machine 2 functions as a processing request reception unit 311 and a processing request distribution unit 312 as shown in FIG. For example, as in the information storage area 130 in the management apparatus 1, for example, distribution management information 131, host management information 132, and VM management information 133 are stored in the information storage area 230.

  For example, the processing request reception unit 311 receives a processing request transmitted from the management device 1. Then, for example, the processing request distribution unit 312 distributes the processing request received by the processing request reception unit 311 to one of the virtual machines 3.

[Outline of First Embodiment]
Next, an outline of the first embodiment will be described. FIG. 9 is a flowchart for explaining the outline of the control processing in the first embodiment. FIG. 10 to FIG. 15 are diagrams for explaining the outline of the control processing in the first embodiment. The outline of the control processing in the first embodiment shown in FIG. 9 will be described with reference to FIGS.

[Load distribution apparatus in first embodiment]
First, the load distribution device 3a in the first embodiment will be described.

  As illustrated in FIG. 10, the load distribution apparatus 3 a according to the first embodiment is configured to process requests distributed by the management apparatus 1 to each of a plurality of virtual machines 3 generated in one or more physical machines 2. Sort out.

  That is, the load balancer 3a according to the first embodiment applies not only to the virtual machine 3 generated in the same physical machine 2 as the own apparatus but also to the virtual machine 3 generated in another physical machine 2. Also distributes processing requests.

  Specifically, in the example illustrated in FIG. 10, the load distribution apparatus 3 a generated in the physical machine 21 is generated in the VMs 31 b, 31 c, and 31 d that are the virtual machines 3 generated in the physical machine 21 and the physical machine 22. The processing requests distributed by the management apparatus 1 are distributed to the VMs 32b, 32c, and 32d that are the virtual machines 3. In the example illustrated in FIG. 10, the load distribution device 3 c generated in the physical machine 21 receives the processing requests distributed by the management device 1 to the VMs 33 b and 31 c that are the virtual machines 3 generated in the physical machine 23. Sort out.

  Thereby, for example, when the load of the load distribution device 3a generated in each physical machine 2 is small, the management device 1 distributes the processing requests to the virtual machines 3 generated in the plurality of physical machines 2. This can be performed by the load balancer 3a. Therefore, the management apparatus 1 can reduce the processing load on each physical machine 2. Hereinafter, one or more physical machines 2 on which one or more virtual machines 3 on which one load distribution apparatus 3a distributes processing requests are also called a group.

[Flowchart of Outline of First Embodiment]
Next, the flowchart of the outline of the first embodiment will be described.

  As shown in FIG. 9, the management apparatus 1 waits until the virtual machine addition determination timing (NO in S1). The virtual machine addition determination timing may be, for example, a timing when the presence of the physical machine 2 in which the current load exceeds the threshold information 134 stored in advance in the information storage area 130 is detected. Specifically, for example, as illustrated in FIG. 11, the management apparatus 1 detects that the load on the physical machine 21 has exceeded the threshold information 134 in association with the generation of the VM 31e that is another virtual machine 3. In this case, it may be determined that the virtual machine addition timing has come.

  When the virtual machine addition timing comes (YES in S1), the management apparatus 1 starts a new load distribution apparatus from the plurality of physical machines 2 based on the load information of the plurality of physical machines 2, as shown in FIG. The placement-destination physical machine 2 on which 3a is placed is specified (S2). Further, as shown in FIG. 12, the management apparatus 1 loads the load distribution apparatus 3a among the plurality of virtual machines 3 constructed in the plurality of physical machines 2 based on the load information of the physical machine 2 specified in the process of S2. Calculates the number of virtual machines 3 that can distribute the processing request received (S3).

  Thereafter, the management apparatus 1 determines the number of virtual machines corresponding to the number calculated in the process of S3 among the plurality of virtual machines 3 in accordance with the arrangement of the load distribution apparatus 3a on the physical machine 2 specified in the process of S2. 3 is the virtual machine 3 that is the distribution destination to which the load distribution apparatus 3a distributes the processing request (S4).

  Specifically, for example, when the physical machine 22 is specified in the processing of S2, the management apparatus 1 generates a new load distribution apparatus 32a, which is a load distribution apparatus 3a, in the physical machine 22, as shown in FIG. . For example, the management apparatus 1 causes the load distribution apparatus 32a, which is a new load distribution apparatus 3a, to distribute processing requests to the VMs 32b, 32c, and 32d operating on the physical machine 22.

  Further, for example, when the number of virtual machines 3 calculated in the process of S3 is 6 or more, the management apparatus 1 adds to the VMs 32b, 32c, and 32d generated in the physical machine 22 as shown in FIG. Then, the load distribution device 32a, which is a new load distribution device 3a, distributes processing requests to the VMs 31b, 31c, and 31d generated in the physical machine 21. In this case, for example, as illustrated in FIG. 14, the management device 1 deletes the load distribution device 31 a in which the virtual machine 3 that is the distribution destination of the processing request disappears.

  On the other hand, for example, when the physical machine 23 is specified in the process of S2 and the number of virtual machines 3 calculated in the process of S3 is five or more, the management apparatus 1 makes the physical machine 23 as shown in FIG. In addition to the generated VMs 33b and 33c, the load distribution device 33a, which is the existing load distribution device 3a, distributes processing requests to the VMs 32b, 32c, and 32d generated in the physical machine 22.

  That is, when adding a new load balancer 3a, the management device 1 adds a new load balancer 3a to a physical machine 2 having a relatively small load. Then, the management apparatus 1 distributes the virtual machine 3 (part of the virtual machine 3) to which the load distribution apparatus 3a generated in the physical machine 2 having a large load distributes the processing request to the new load distribution apparatus. Let 3a do.

  As a result, the management apparatus 1 can prevent a new load distribution apparatus 3a from being added to the physical machine 2 with a large load. In addition, the management apparatus 1 can suppress the load on the physical machine 2 having a large load by reducing the number of virtual machines 3 to which the load distribution apparatus 3a generated in the physical machine 2 having a large load distributes. Become. Therefore, the management apparatus 1 can level the load on each physical machine 2.

[Details of First Embodiment]
Next, details of the first embodiment will be described. FIG. 16 to FIG. 23 are flowcharts for explaining the details of the control processing in the first embodiment. FIGS. 24 to 29 are diagrams for explaining the details of the control processing in the first embodiment. Details of the control process shown in FIGS. 16 to 23 will be described with reference to FIGS. 24 to 29.

[Distribution processing]
First, a process (hereinafter also referred to as a distribution process) performed when a process request is transmitted from a user terminal (not shown) or the like will be described. 16 and 17 are diagrams for explaining the distribution process.

  As illustrated in FIG. 16, the processing request reception unit 111 of the management device 1 waits until a processing request transmitted by a user via a user terminal (not shown) is received (NO in S11).

  When a processing request is received (YES in S11), the processing request distribution unit 112 of the management apparatus 1 refers to the host management information 132 and calculates the sum of the loads on the physical machines 2 in each group (S12). . Thereafter, the process request distribution unit 112 refers to the distribution management information 131 and identifies a load balancer corresponding to the group having the smallest sum calculated in the process of S12 (S13). Hereinafter, specific examples of the distribution management information 131 and the host management information 132 will be described.

[Specific examples of distribution management information and host management information]
24, 26, 28, and 29 are diagrams illustrating specific examples of the distribution management information 131 and the host management information 132. FIG. Specifically, FIGS. 24A, 26A, 28A, and 29A are diagrams illustrating specific examples of the distribution management information 131. FIG. FIGS. 24B, 26B, 28B, and 29B are diagrams illustrating specific examples of the host management information 132. FIG. Note that the contents of the distribution management information 131 and the host management information 132 shown in FIG. 24 correspond to the states of the devices described with reference to FIG.

  The distribution management information 131 shown in FIG. 24A and the like includes a “group ID” that identifies each group, a “flag” that indicates whether each device is the management device 1 or the load distribution device 3a, The items include “device ID” for identifying each device and “IP address” in which the IP address of each device is set. In the “flag”, either “true” indicating that each device is the management device 1 or “false” indicating that each device is the load balancer 3a is set. Further, the distribution management information 131 shown in FIG. 24A or the like indicates the number of load distribution apparatuses 3a to which the management apparatus 1 distributes processing requests, or the virtual machine 3 to which the load distribution apparatus 3 distributes processing requests. And the upper limit value of the number of load distribution devices 3a to which the management device 1 distributes processing requests, or the number of virtual machines 3 to which the load distribution device 3 distributes processing requests. An “upper limit value” is set in which an upper limit value is set. Furthermore, the distribution management information 131 shown in FIG. 24A and the like includes “load information” in which the current load information of each device is set, and “host ID” that identifies the physical machine 2 in which each device is generated. Is set. In the “load information”, for example, the current average usage rate of the CPU assigned to each device is set.

  Specifically, in the distribution management device 1 shown in FIG. 24A, “true” is set as the “flag” and “1000” is set as the “device ID” in the information whose “group ID” is “0000”. Is set, and “100.100.100.50” is set as the “IP address”. In the distribution management apparatus 1 shown in FIG. 24A, “2 (units)” is set as the “number of distributions” in the information whose “group ID” is “0000”, and the “upper limit value” is set. “4 (unit)” is set, “10% / second” is set as “load information”, and “2000” is set as “host ID”. Also, in the distribution management device 1 shown in FIG. 24A, “false” is set as “flag” and “1001” is set as “device ID” in the information whose “group ID” is “0001”. Is set, and “100.100.100.51” is set as the “IP address”. In the distribution management apparatus 1 shown in FIG. 24A, “6 (units)” is set as the “number of distributions” and “upper limit value” is set in the information whose “group ID” is “0001”. “6 (unit)” is set, “10% / second” is set as “load information”, and “2001” is set as “host ID”. Description of other information included in FIG.

  In the host management information 132 shown in FIG. 24B and the like, a “group ID” for identifying each group, a “host ID” for identifying each physical machine 2, and the IP address of each physical machine 2 are set. “IP address” as an item. In the host management information 132 shown in FIG. 24B and the like, the “number of VMs” in which the number of virtual machines 3 generated in each physical machine 2 is set, the load balancer 3a, and the load balancer 3a. The item includes “number of own service VMs” in which the total number of virtual machines 3 to which processing requests are distributed is set, and “load information” in which the current load information of each physical machine 2 is set.

  Specifically, in the host management information 132 shown in FIG. 24B, the information “IP ID” is “100.100” in the information whose “group ID” is “00001” and “host ID” is “20001”. 100.100.1 ”is set,“ 4 (units) ”is set as the“ number of VMs ”,“ 4 (units) ”is set as the“ number of own service VMs ”, and“ 30% ”is set as the“ load information ”. / Sec "is set. Also, in the host management information 132 shown in FIG. 24B, the information “IP ID” is “100.100.” In the information where “Group ID” is “00001” and “Host ID” is “20002”. “100.2” is set, “3 (units)” is set as the “number of VMs”, “3 (units)” is set as the “number of own service VMs”, and “20% / second” is set as the “load information” "Is set. Description of other information included in FIG.

  Then, in the process of S12, the process request distribution unit 112 refers to, for example, the host management information 132 described with reference to FIG. 24B, and sets the “load information” of the information whose “group ID” is “0001”. “50% / second”, which is the sum of the set information “30% / second” and “20% / second”, is calculated. Further, the processing request distribution unit 112 refers to, for example, the host management information 132 described with reference to FIG. 24B, and is information set to “load information” of the information whose “group ID” is “0002”. A certain “30% / second” is specified. That is, the processing request distribution unit 112 calculates “50% / second” as the sum of the loads of the physical machines 2 included in the group whose “group ID” is “0001”, and the “group ID” is “0002”. "30% / second" is calculated as the sum of the loads of the physical machines 2 included in the group "."

  Thereafter, the process request distribution unit 112 refers to the distribution management information 131 described with reference to FIG. 24A in the process of S13, and calculates “30% / second” calculated in the process of S12 (calculated in the process of S12). "0001" corresponding to "50% / second" and "30% / second") is the information set in "device ID" of the information set in "group ID" “10001” is specified.

  Returning to FIG. 16, the process request distribution unit 112 transmits the process request received in the process of S11 to the load distribution apparatus 3a specified in the process of S13 (S14). Specifically, when the “group ID” of the load balancer 3a specified in the process of S13 is “10001”, the load balancer 3a whose “group ID” is “10001” is accepted in the process of S11. Send the processing request.

  As a result, the management apparatus 1 can level the processing load on each load distribution apparatus 3a.

  On the other hand, as shown in FIG. 17, the processing request reception unit 311 of the load distribution device 3a waits until it receives a processing request transmitted from the management device 1 (NO in S21).

  When the processing request is received (YES in S21), the processing request distribution unit 312 refers to the VM management information 133, and the virtual machine 3 with the smallest load information from the virtual machine 3 to which the own device distributes the processing request. Is specified (S22). Hereinafter, specific examples of the VM management information will be described.

[Specific example of VM management information]
25 and 27 are diagrams for explaining a specific example of the VM management information 133. FIG. Note that the content of the VM management information 133 illustrated in FIG. 25 corresponds to the state of each device described with reference to FIG.

  The VM management information 133 illustrated in FIG. 25 and the like includes a “host ID” that identifies the physical machine 2 in which each virtual machine 3 is generated, a “VMID” that identifies each virtual machine 3, and the IP of each virtual machine 3. “IP address” to which an address is set is included as an item. Further, the VM management information 133 illustrated in FIG. 25 and the like is set with a “processing type” for identifying a service provided to the user by the operation of each virtual machine 3 and the current load information of each virtual machine 3. “Load information” is included as an item.

  Specifically, in the VM management information 133 illustrated in FIG. 25, the information “host ID” is “20001” and “VMID” is “30001” has “100.100.100. 101 ”is set,“ A ”is set as“ Processing type ”, and“ 10% / second ”is set as“ Load information ”. Also, in the VM management information 133 shown in FIG. 25, the information “IP ID” is “100.100.100.102” for the information that “host ID” is “20001” and “VMID” is “30002”. Is set, “A” is set as the “processing type”, and “5% / second” is set as the “load information”. Description of other information included in FIG. 25 is omitted.

  In the process of S22, when the load distribution apparatus 31a receives a process request from the management apparatus 1, the process request distribution unit 312 refers to, for example, the VM management information 133 illustrated in FIG. “20001” corresponding to the operating physical machine 2 identifies “30001”, “30002”, and “30003” that are information set in “VMID” of the information set in “Host ID”. Here, in the VM management information 133 illustrated in FIG. 25, “10% / second”, “5” is included in “load information” of information in which “30001”, “30002”, and “30003” are set in “VMID”. % / Second "and" 10% / second "are set. Therefore, the processing request distribution unit 312 sets “5% / second” (the minimum value among “10% / second”, “5% / second”, and “10% / second”) to “load information”. “30002”, which is the information set in “VMID” of the set information, is specified (S22).

  Returning to FIG. 16, the process request distribution unit 312 transmits the process request received in the process of S21 to the virtual machine 3 specified in the process of S22 (S23). Specifically, when the virtual machine 3 whose “VMID” is “30002” is identified in the processing of S22, the processing request distribution unit 312 requests the virtual machine 3 whose “VMID” is “30002”. Send.

  As a result, the load distribution device 3a can level the processing load on each virtual machine 3 that distributes processing requests.

[Information update processing]
Next, processing for updating the distribution management information 131, the host management information 132, and the VM management information 133 stored in the information storage area 130 (hereinafter also referred to as information update processing) will be described. FIG. 18 is a diagram illustrating the information update process.

  As illustrated in FIG. 18, the information acquisition unit 113 of the management device 1 waits until the information acquisition timing (NO in S31). The information acquisition timing is a timing for acquiring information necessary for updating the distribution management information 131, the host management information 132, and the VM management information 133, and may be a periodic timing, for example.

  Then, when the information acquisition timing comes (YES in S31), the information acquisition unit 113 accesses each physical machine 2, and information necessary for updating the distribution management information 131, the host management information 132, and the VM management information 133 Is acquired (S32).

  Specifically, the information acquisition unit 113 acquires information set in “number of VMs” and “number of own service VMs” in the host management information 132, for example. Further, the information acquisition unit 113 acquires information set in “VMID”, “IP address”, “processing type”, and “load information” in the VM management information 133, for example.

  Thereafter, the information management unit 114 of the management apparatus 1 updates the distribution management information 131, the host management information 132, and the VM management information 133 stored in the information storage area 130 based on the information acquired in the process of S32 (S33). ).

  Specifically, the information management unit 114 updates the host management information 132 and the VM management information 133 with the information acquired in the process of S32, for example. Further, the information management unit 114 sets the information set in the “distribution number” and “upper limit” in the distribution management information 131 and the “load information” in the host management information 132 from the information acquired in the process of S32. The distribution management information 131 and the host management information 132 are updated with the calculated information.

  As a result, the management apparatus 1 can perform control processing based on information updated as needed according to a change in the state of the virtual machine 3 or the like.

[Details of control processing]
Next, details of the control process will be described. 19 to 23 are diagrams illustrating details of the control process.

  As shown in FIG. 19, the placement destination specifying unit 115 of the management device 1 waits until the control timing (NO in S41). The control timing may be a periodic timing, for example.

  When the control timing comes (YES in S41), the placement destination specifying unit 115 acquires the load information included in the host management information 132 stored in the information storage area 130 (S42). Specifically, the placement destination specifying unit 115 acquires the load information for each physical machine 2 from the information included in the host management information 132. In this case, the placement destination specifying unit 115 acquires the threshold information 134 stored in the information storage area 130 (S43).

  Thereafter, the placement destination specifying unit 115 specifies the physical machine 2 from the plurality of physical machines 2 in which the load information acquired in the process of S42 exceeds the threshold information 134 acquired in the process of S43 (S44).

  As a result, when there is no physical machine 2 whose load information exceeds the threshold information 134 (NO in S45), the placement destination specifying unit 115 ends the control process. That is, in this case, the placement destination specifying unit 115 determines that it is not necessary to place a new load distribution device 3a in any of the physical machines 2.

  On the other hand, when there is a physical machine 2 whose load information exceeds the threshold information 134 (YES in S45), the placement destination specifying unit 115 assigns a new load balancer 3a to any of the physical machines 2, as shown in FIG. Performs processing to arrange the crab. Hereinafter, a specific example in the case where there is a physical machine 2 whose load information exceeds the threshold information 134 will be described.

[Specific example when there is a physical machine whose load information exceeds the threshold information]
26 and 27 are diagrams illustrating specific examples of the distribution management information 131, the host management information 132, and the VM management information 133 when there is a physical machine 2 whose load information exceeds the threshold information 134. FIG. Specifically, as described in FIG. 11, specific examples of each information when the load information of the physical machine 21 exceeds the threshold information 134 as a result of the VM 31e that is another virtual machine 3 being generated in the physical machine 21. Will be described.

  In this case, the information management unit 114 sets the “number of VMs” of the information whose “group ID” is “0001” and “host ID” is “20001” as indicated by the underlined portion in FIG. Add 1 to the set value. On the other hand, the information management unit 114 does not update the value set in the “number of own service VMs” because the VM 31e generated in the physical machine 21 is another virtual machine 3.

  Further, in this case, the information management unit 114 updates “load information” of information whose “group ID” is “0001” and whose “host ID” is “20001”. Specifically, as illustrated in the underlined portion of FIG. 26B, the information management unit 114, for example, “load” of information whose “group ID” is “0001” and “host ID” is “20001”. Information "is updated to" 45% seconds ".

  Further, as information corresponding to the VM 31e generated in the physical machine 21, the information management unit 114 has “host ID” “20001” and “VMID” “ 30009 ”,“ IP Address ”is“ 100.100.100.109 ”,“ Processing Type ”is“ B ”, and“ Load Information ”is“ 15% / sec ”. .

  For example, when the value indicated by the threshold information 134 stored in the information storage area 130 is “40% / second”, the placement destination specifying unit 115 indicates that the load information of the physical machine 21 exceeds the threshold information 134. And start processing for placing a new load balancer 3a in any of the physical machines 2 (YES in S44 and S45).

  Returning to FIG. 19, when there is a physical machine 2 whose load information exceeds the threshold information 134 (YES in S45), the placement destination specifying unit 115 refers to the host management information 132 as shown in FIG. Among the physical machines 2, one or more physical machines 2 on which one or more virtual machines 3 to which processing requests are distributed by the load balancer 3a operating on the existing physical machines 2 operate are specified (S51).

  Specifically, when the load information of the physical machine 21 exceeds the threshold information 134, the placement destination specifying unit 115 refers to the host management information 132 described with reference to FIG. The physical machine 22 in which the information set in the “group ID” is the same as the physical machine 21 is specified.

  Then, the placement destination specifying unit 115 refers to the host management information 132 and specifies the physical machine 2 whose load information is less than a predetermined threshold from one or more physical machines specified in the process of S51 (S52). The predetermined threshold is, for example, a value obtained by subtracting the value indicated by the load information of the load balancer 3a operating in the physical machine 2 existing in the process of S45 from the value indicated by the load information of each physical machine 2. Good.

  Specifically, when the load information of the physical machine 21 exceeds the threshold information 134, the placement destination specifying unit 115 refers to the distribution management information 131 described with reference to FIG. “10001” corresponding to the load balancer 3a specifies “10% / second” which is information set in “load information” of information set in “device ID”. For example, the placement destination specifying unit 115 refers to the host management information 132 described with reference to FIG. 26B, and “20002” corresponding to the physical machine 22 included in the same group as the physical machine 21 is “host ID”. “20% / second”, which is the information set in the “load information” of the information set in “1”, is specified.

  When the threshold information 134 stored in the information storage area 130 is “40% / second”, the predetermined threshold is “10% / second”, which is the specified value, and the threshold information 134 is “40% / second”. “30% / second” subtracted from “second”, which is larger than the specified value “20% / second”. Therefore, in this case, the placement destination specifying unit 115 specifies the physical machine 22 as the physical machine 2 whose load information is less than a predetermined threshold.

  Thereafter, when there is a physical machine 2 whose load information is less than the predetermined threshold (YES in S53), the placement destination specifying unit 115 designates any of the existing physical machines 2 as the placement destination of the new load balancer 3a. And a new load balancer 3a is arranged for the determined physical machine 2 (S54).

  Further, in this case, the distribution destination determination unit 117 of the management apparatus 1 selects all the virtual machines 3 to which the processing request is distributed among the virtual machines 3 operating in the physical machine 2 that existed in the processing of S53 in S54. The load distribution apparatus 3a arranged in the process determines the virtual machine 3 that distributes the process request (S55).

  Specifically, in the physical machine 22, when the VMs 32b, 32c, and 32d are operating as the virtual machines 3 to which the load distribution apparatus 31a distributes processing requests, the distribution destination determination unit 117 newly sets the VMs 32b, 32c, and 32d. The load balancer 32a, which is a simple load balancer 3a, determines the virtual machine 3 that distributes processing requests.

  That is, when the load information of the physical machine 21 exceeds the threshold information 134, the placement destination specifying unit 115 suppresses the number of virtual machines 3 to which the load distribution apparatus 31a operating on the physical machine 21 distributes processing requests.

  As a result, the management apparatus 1 can reduce the load on the physical machine 21 whose load information exceeds the threshold information 134.

  Subsequently, as illustrated in FIG. 21, the machine number calculation unit 116 of the management apparatus 1 applies to the virtual machines 3 other than the virtual machine 3 determined by the load distribution apparatus 3 a arranged in the process of S <b> 54. The virtual machine generated in the physical machine 2 other than the physical machine 2 determined in the process of S54 among the one or more physical machines 2 specified in the process of S51 as to whether or not the process request can be distributed. A determination is made every three (S61).

  If it is determined that the processing request can be distributed to the virtual machines 3 other than the virtual machine 3 determined in S55 (YES in S62), the distribution destination determination unit 117 performs the process in S62. The virtual machine 3 that has been determined to be able to be distributed at step S54 is determined as the virtual machine 3 to which the new load distribution apparatus 3a arranged in the process of S54 distributes the processing request (S63).

  In other words, the physical machine 2 that has generated the new load balancer 3 a performs processing not only on the virtual machine 3 that operates on the physical machine 2 but also on virtual machines 3 other than the virtual machine 3 that operates on the physical machine 2. It may be possible to distribute requests. Therefore, the number-of-machines calculation unit 116 determines whether or not the load distribution device 3a arranged in the process of S54 can distribute processing requests to more virtual machines 3 in the process of S61. I do.

  Specifically, the number-of-machines calculation unit 116 determines, for example, in the process of S54 within a range where the value set in the “number of distributions” in the distribution management information 131 does not exceed the value set in the “upper limit value”. The number of virtual machines 3 to which processing requests distributed on physical machines 2 other than the physical machine 2 are distributed are assigned to the load balancer 3a arranged in the process of S54 for each virtual machine 3 generated in each physical machine 2. Add to the value set in the corresponding "number of distributions". Then, the number-of-machines calculation unit 116 determines the virtual machine 3 obtained by adding the number of virtual machines 3 to the value set in the “number of distributions” as the virtual machine 3 to which the new load balancer 3a distributes.

  Thereafter, the information management unit 114 determines whether or not the load balancer 3a operating on the physical machine 2 existing in the process of S45 can be deleted (S64). Specifically, the information management unit 114 determines whether or not there is the virtual machine 3 to which the load distribution apparatus 3a operating on the physical machine 2 existing in the process of S45 distributes the processing request. Similarly, when the information management unit 114 determines in the process of S62 that the process request cannot be distributed to the virtual machines 3 other than the virtual machine 3 determined in the process of S55 (NO in S62). , S64 is performed.

  If it is determined that the load balancer 3a operating on the physical machine 2 existing in the process of S45 can be deleted (YES in S64), the information management unit 114 determines that the physical machine 2 existing in the process of S45. The load balancer 3a that has been operating in (1) is deleted (S65). On the other hand, when it is determined that the load balancer 3a operating on the physical machine 2 existing in the process of S45 cannot be deleted (NO in S64), the information management unit 114 does not perform the process of S65.

  Thereby, it is possible to further reduce the load on the physical machine 21 whose load information exceeds the threshold information 134.

[Specific example of distribution management information and host management information (1)]
Next, specific examples of the distribution management information 131 and the host management information 132 when the processes of S63 and S65 are performed will be described. Specifically, specific examples of the distribution management information 131 and the host management information 132 updated in the information update process when each device is in the state described in FIG. 14 will be described.

  In this case, as shown in FIG. 28A, the information management unit 114 sets “00001” in “Group ID” in the distribution management information 131 shown in FIG. Information with “10001” set in is deleted. Then, the information management unit 114 sets “00001” to “Group ID”, “false” to “Flag”, and “10003” to “Device ID”, as indicated by the underlined portion in FIG. “100.100.100.53” is set in the “IP address”, “6” is set in the “number of distributions”, “6” is set in the “upper limit value”, and “load information” is set in the “load information”. Information in which “10% / second” is set and “20002” is set in “host ID” is added.

  That is, the information management unit 114 causes the distribution management information 131 to reflect information indicating that the load balancer 31 a that has been operating in the physical machine 21 has been deleted and the load balancer 32 a has been generated in the physical machine 22. In addition, the information management unit 114 distributes processing requests to the three virtual machines 3 generated in the physical machine 21 as well as the three virtual machines 3 generated in the physical machine 22 to the load balancer 31a. Instead, information indicating that the load distribution device 32 a has started is reflected in the distribution management information 131.

  Further, in this case, as shown in FIG. 28B, the information management unit 114 sets the “number of VMs” and the “group ID” of the information whose “group ID” is “00001” and “host ID” is “20001”. 1 is subtracted from the value set in the “number of VMs of own service”. Then, as shown in FIG. 28B, the information management unit 114 displays “number of VMs” and “own service” of the information whose “group ID” is “00001” and “host ID” is “20002”. 1 is added to the value set in the “number of VMs”.

  That is, the information management unit 114 causes the host management information 132 to reflect information indicating that the load balancer 31 a that has been operating in the physical machine 21 has been deleted and the load balancer 32 a has been generated in the physical machine 22.

  Returning to FIG. 20, in the process of S53, when there is no physical machine 2 whose load information is less than the predetermined threshold (NO in S53), the placement destination specifying unit 115 performs host management as shown in FIG. Referring to the information 132, the load balancer 3a operating in the physical machine 2 other than the one or more physical machines 2 identified in the process of S51 is identified from the physical machines 2 in other groups (S71).

  That is, in this case, the placement destination specifying unit 115 performs processing by the load balancer 3a among the load balancers 3a operating in the physical machines 2 in a different group from the physical machines 2 whose load information exceeds the threshold information 134. The load balancer 3a that can distribute the processing request to the virtual machine 3 that distributes the request is specified.

  Specifically, when the load information of the physical machine 21 exceeds the threshold information 134, the placement destination specifying unit 115 refers to, for example, the host management information 132 described with reference to FIG. The load balancer 33a that operates in the physical machine 23 that is the physical machine 2 other than the machine 22 is specified.

  Then, the number-of-machines calculation unit 116 processes each of the load balancers 3a identified in the process of S71 for the virtual machines 3 to which the load balancer 3a operating in the physical machine 2 existing in S45 distributes the processing request. Whether or not the request can be distributed is determined for each virtual machine 3 generated in one or more physical machines 2 identified in the process of S51 among the plurality of physical machines 2 (S72).

  If it is determined that there is no load balancer 3a that can distribute processing requests (NO in S73), the placement destination specifying unit 115 assigns a new physical machine 2 to the new load balancer 3a. It is determined as the previous physical machine 2 (S74).

  That is, in this case, the placement destination specifying unit 115 determines that a new load distribution device 3a cannot be generated in any of the physical machines 2 that are currently operating. For this reason, the placement destination specifying unit 115 determines that it is necessary to prepare a new physical machine 2 and generate a new load distribution device 3 a in the physical machine 2. And the arrangement | positioning destination specific | specification part 115 performs the output about said content with respect to the provider terminal 5, for example.

  On the other hand, in the process of S73, when it is determined that there is a physical machine 2 that can distribute the process request (YES in S73), the placement destination specifying unit 115, as shown in FIG. 117 determines the virtual machine 3 determined to be able to be distributed in the process of S73 as the virtual machine 3 to which the load distribution apparatus 3a specified in the process of S71 distributes the processing request (S81).

  Specifically, the machine number calculation unit 116, for example, loads the threshold information 134 within a range in which the value set in the “number of distributions” in the distribution management information 131 does not exceed the value set in the “upper limit value”. The number of processing request distribution destination virtual machines 3 that operate in the physical machines 2 that exceed the number corresponding to the load balancer 3a identified in the processing of S71 for each virtual machine 3 generated in each physical machine 2 " Add to the value set in "Number of distributions". Then, the number-of-machines calculation unit 116 loads the virtual machine 3 in which the number of virtual machines 3 is added to the value set in the “number of distributions” operating on the physical machine 2 identified in the process of S71. The virtual machine 3 to be distributed by the distribution device 3a is determined.

  Thereafter, the information management unit 114 determines whether or not the load balancer 3a operating on the physical machine 2 existing in the process of S45 can be deleted (S82). Specifically, the information management unit 114 determines whether or not there is the virtual machine 3 to which the load distribution apparatus 3a operating on the physical machine 2 existing in the process of S45 distributes the processing request.

  If it is determined that the load balancer 3a operating on the physical machine 2 existing in the process of S45 can be deleted (YES in S82), the information management unit 114 determines that the physical machine 2 existing in the process of S45. The load balancer 3a that was operating in step S3 is deleted (S83). On the other hand, when it is determined that the load balancer 3a operating on the physical machine 2 existing in the process of S45 cannot be deleted (NO in S82), the information management unit 114 does not perform the process of S83.

  Thereby, it is possible to further reduce the load on the physical machine 21 whose load information exceeds the threshold information 134.

[Specific example of distribution management information and host management information (2)]
Next, specific examples of the distribution management information 131 and the host management information 132 when the process of S81 is performed and the process of S83 is performed will be described. Specifically, specific examples of the distribution management information 131 and the host management information 132 updated in the information update process when each device is in the state described in FIG. 15 will be described.

  In this case, as indicated by the underlined portion in FIG. 29A, the information management unit 114 is information in which “00001” is set in the “group ID” in the distribution management information 131 illustrated in FIG. Is updated to a value obtained by subtracting “3 (units)”, which is the number of virtual machines 3 operating on the physical machine 22. In addition, as shown in the underlined portion of FIG. 29A, the information management unit 114 includes information of “00002” set in “Group ID” in the distribution management information 131 shown in FIG. The value set in the “number of distributions” is updated to a value obtained by adding “3 (units)” which is the number of virtual machines 3 operating on the physical machine 22.

  That is, the information management unit 114 reflects in the distribution management information 131 information indicating that the load distribution device 33a has started distribution for a part of the virtual machine 3 to which the load distribution device 31a has distributed the processing request. Let

  As described above, the management apparatus 1 according to the present embodiment receives a processing request received in any of the plurality of virtual machines 3 constructed in any of the plurality of physical machines 2 based on the load information of the plurality of physical machines 2. The physical machine 2 that is the placement destination in which the specific load distribution device 3a for sorting is placed is identified. Based on the load information of the identified physical machine 2, the management apparatus 1 calculates the number of virtual machines 3 that can distribute the processing requests received by the specific load distribution apparatus 3 among the plurality of virtual machines 3.

  Furthermore, the management apparatus 1 assigns the number of virtual machines 3 corresponding to the calculated number out of the plurality of virtual machines 3 according to the arrangement of the specific load distribution apparatus 3 a to the specified physical machine 2. A distribution destination virtual machine 3 that distributes a processing request received by the distribution apparatus 3a is assumed.

  That is, when adding a new load balancer 3a, the management device 1 adds a new load balancer 3a to a physical machine 2 having a relatively small load. Then, the management apparatus 1 distributes the virtual machine 3 (part of the virtual machine 3) to which the load distribution apparatus 3a generated in the physical machine 2 having a large load distributes the processing request to the new load distribution apparatus. Let 3a do.

  As a result, the management apparatus 1 can prevent a new load distribution apparatus 3a from being added to the physical machine 2 with a large load. In addition, the management apparatus 1 can suppress the load on the physical machine 2 having a large load by reducing the number of virtual machines 3 to which the load distribution apparatus 3a generated in the physical machine 2 having a large load distributes. Become. Therefore, the management apparatus 1 can level the load on each physical machine 2.

  The above embodiment is summarized as follows.

(Appendix 1)
In a control program for allocating a specific virtual machine that distributes a request received to one of a plurality of virtual machines, each of which is constructed on one of a plurality of physical machines, to one of the plurality of physical machines,
Based on the load information of the plurality of physical machines, identify a physical machine to which the specific virtual machine is to be placed,
Based on the load information of the identified physical machine, out of the plurality of virtual machines, calculate the number of virtual machines that can distribute the request received by the specific virtual machine,
According to the arrangement of the specific virtual machine on the specified physical machine, among the plurality of virtual machines, the number of virtual machines corresponding to the calculated number is distributed to requests received by the specific virtual machine. A virtual machine to distribute to
A control program for causing a computer to execute processing.

(Appendix 2)
In Appendix 1,
The load information of the plurality of physical machines is information indicating a current load for each of the plurality of physical machines,
In the specifying process, the load information of the plurality of physical machines is referred to, and when there are physical machines having a current load equal to or greater than a predetermined threshold in the plurality of physical machines, the physical machine of the placement destination is determined. Do,
A control program characterized by that.

(Appendix 3)
In Appendix 2,
In the process of specifying,
Among the plurality of physical machines, specify one or more physical machines on which one or more virtual machines to which a request is distributed by a virtual machine that operates on a physical machine having a current load equal to or greater than a first threshold,
The physical machine of the placement destination is determined from physical machines whose current load is less than a second threshold among the one or more physical machines.
A control program characterized by that.

(Appendix 4)
In Appendix 3,
In the process of specifying,
When there is no physical machine having a current load less than the second threshold value in the one or more physical machines, the physical location of the placement destination is selected from physical machines other than the one or more physical machines among the plurality of physical machines. Make machine decisions,
A control program characterized by that.

(Appendix 5)
In Appendix 3,
In the process of setting the number of virtual machines according to the number as the distribution destination virtual machine, the distribution destination virtual machine is determined from the one or more virtual machines.
A control program characterized by that.

(Appendix 6)
In Appendix 5,
In the process of assigning a number of virtual machines according to the number to a distribution destination virtual machine, for each virtual machine operating in each physical machine, determine as the distribution destination virtual machine,
A control program characterized by that.

(Appendix 7)
In Appendix 6,
In the process of assigning a number of virtual machines according to the number to a destination virtual machine, of the one or more virtual machines, a virtual machine that operates on a different physical machine from a virtual machine that distributes requests to the one or more virtual machines Priority is given to determining as a virtual machine of the distribution destination,
A control program characterized by that.

(Appendix 8)
In Appendix 5,
If all of the one or more virtual machines are determined as the distribution destination virtual machines, the virtual machines that distribute requests to the one or more virtual machines are deleted.
A control program for causing a computer to execute processing.

(Appendix 9)
In Appendix 1,
In the calculation process,
Refer to the load information of the identified physical machine, calculate the difference between the current load of the identified physical machine and the third threshold,
Calculating the additional number of virtual machines in which the increase in the load of the identified physical machine is equal to or less than the difference, as the number of virtual machines to which requests received by the specific virtual machine can be distributed,
A control program characterized by that.

(Appendix 10)
In Appendix 9,
In the process of calculating, a virtual machine capable of distributing a request accepted by the specific virtual machine, the maximum number of the distribution-destination virtual machines in which the increase in load of the specified physical machine is equal to or less than the difference Calculated as the number of
A control program characterized by that.

(Appendix 11)
In a control apparatus that places a specific virtual machine that distributes a request received in any of a plurality of virtual machines, each of which is constructed in any of a plurality of physical machines, in any of the plurality of physical machines,
Based on load information of the plurality of physical machines, an arrangement destination specifying unit that specifies an arrangement destination physical machine in which the specific virtual machine is arranged;
Based on the identified physical machine load information, among the plurality of virtual machines, a machine number calculation unit that calculates the number of virtual machines that can distribute requests received by the specific virtual machine;
According to the arrangement of the specific virtual machine on the specified physical machine, among the plurality of virtual machines, the number of virtual machines corresponding to the calculated number is distributed to requests received by the specific virtual machine. A distribution destination determination unit as a distribution destination virtual machine,
A control device characterized by that.

(Appendix 12)
In Appendix 11,
The load information of the plurality of physical machines is information indicating a current load for each of the plurality of physical machines,
The placement destination specifying unit refers to the load information of the plurality of physical machines, and when there are physical machines having a current load equal to or greater than a predetermined threshold in the plurality of physical machines, the placement destination specifying unit is determined. I do,
A control device characterized by that.

(Appendix 13)
In Appendix 12,
The placement destination specifying unit is
Among the plurality of physical machines, specify one or more physical machines on which one or more virtual machines to which a request is distributed by a virtual machine that operates on a physical machine having a current load equal to or greater than a first threshold,
The physical machine of the placement destination is determined from physical machines whose current load is less than a second threshold among the one or more physical machines.
A control device characterized by that.

(Appendix 14)
In Appendix 13,
The placement destination specifying unit is
When there is no physical machine having a current load less than the second threshold value in the one or more physical machines, the physical location of the placement destination is selected from physical machines other than the one or more physical machines among the plurality of physical machines. Make machine decisions,
A control device characterized by that.

(Appendix 15)
In the control method of allocating a specific virtual machine that distributes a request received to any of a plurality of virtual machines, each of which is constructed on any of a plurality of physical machines, to any of the plurality of physical machines,
Based on the load information of the plurality of physical machines, identify a physical machine to which the specific virtual machine is to be placed,
Based on the load information of the identified physical machine, out of the plurality of virtual machines, calculate the number of virtual machines that can distribute the request received by the specific virtual machine,
According to the arrangement of the specific virtual machine on the specified physical machine, among the plurality of virtual machines, the number of virtual machines corresponding to the calculated number is distributed to requests received by the specific virtual machine. A virtual machine to distribute to
A control method characterized by that.

(Appendix 16)
In Appendix 15,
The load information of the plurality of physical machines is information indicating a current load for each of the plurality of physical machines,
In the specifying step, the load information of the plurality of physical machines is referred to, and if there are physical machines having a current load equal to or greater than a predetermined threshold in the plurality of physical machines, the physical machine of the placement destination is determined. Do,
A control method characterized by that.

(Appendix 17)
In Appendix 16,
In the identifying step,
Among the plurality of physical machines, specify one or more physical machines on which one or more virtual machines to which a request is distributed by a virtual machine that operates on a physical machine having a current load equal to or greater than a first threshold,
The physical machine of the placement destination is determined from physical machines whose current load is less than a second threshold among the one or more physical machines.
A control method characterized by that.

(Appendix 18)
In Appendix 17,
In the identifying step,
When there is no physical machine having a current load less than the second threshold value in the one or more physical machines, the physical location of the placement destination is selected from physical machines other than the one or more physical machines among the plurality of physical machines. Make machine decisions,
A control method characterized by that.

1: Management device 2: Physical machine 3: Virtual machine 4: Virtualization software 5: Service provider terminal

Claims (12)

In a control program for allocating a specific virtual machine that distributes a request received to one of a plurality of virtual machines, each of which is constructed on one of a plurality of physical machines, to one of the plurality of physical machines,
Based on the load information of the plurality of physical machines, identify a physical machine to which the specific virtual machine is to be placed,
Based on the load information of the identified physical machine, out of the plurality of virtual machines, calculate the number of virtual machines that can distribute the request received by the specific virtual machine,
According to the arrangement of the specific virtual machine on the specified physical machine, among the plurality of virtual machines, the number of virtual machines corresponding to the calculated number is distributed to requests received by the specific virtual machine. A virtual machine to distribute to
A control program for causing a computer to execute processing. In claim 1,
The load information of the plurality of physical machines is information indicating a current load for each of the plurality of physical machines,
In the specifying process, the load information of the plurality of physical machines is referred to, and when there are physical machines having a current load equal to or greater than a predetermined threshold in the plurality of physical machines, the physical machine of the placement destination is determined. Do,
A control program characterized by that. In claim 2,
In the process of specifying,
Among the plurality of physical machines, specify one or more physical machines on which one or more virtual machines to which a request is distributed by a virtual machine that operates on a physical machine having a current load equal to or greater than a first threshold,
The physical machine of the placement destination is determined from physical machines whose current load is less than a second threshold among the one or more physical machines.
A control program characterized by that. In claim 3,
In the process of specifying,
When there is no physical machine having a current load less than the second threshold value in the one or more physical machines, the physical location of the placement destination is selected from physical machines other than the one or more physical machines among the plurality of physical machines. Make machine decisions,
A control program characterized by that. In claim 3,
In the process of setting the number of virtual machines according to the number as the distribution destination virtual machine, the distribution destination virtual machine is determined from the one or more virtual machines.
A control program characterized by that. In claim 5,
In the process of assigning a number of virtual machines according to the number to a distribution destination virtual machine, for each virtual machine operating in each physical machine, determine as the distribution destination virtual machine,
A control program characterized by that. In claim 6,
In the process of assigning a number of virtual machines according to the number to a destination virtual machine, of the one or more virtual machines, a virtual machine that operates on a different physical machine from a virtual machine that distributes requests to the one or more virtual machines Priority is given to determining as a virtual machine of the distribution destination,
A control program characterized by that. In claim 5, further:
If all of the one or more virtual machines are determined as the distribution destination virtual machines, the virtual machines that distribute requests to the one or more virtual machines are deleted.
A control program for causing a computer to execute processing. In claim 1,
In the calculation process,
Refer to the load information of the identified physical machine, calculate the difference between the current load of the identified physical machine and the third threshold,
Calculating the additional number of virtual machines in which the increase in the load of the identified physical machine is equal to or less than the difference, as the number of virtual machines to which requests received by the specific virtual machine can be distributed,
A control program characterized by that. In claim 9,
In the process of calculating, a virtual machine capable of distributing a request accepted by the specific virtual machine, the maximum number of the distribution-destination virtual machines in which the increase in load of the specified physical machine is equal to or less than the difference Calculated as the number of
A control program characterized by that. In a control apparatus that places a specific virtual machine that distributes a request received in any of a plurality of virtual machines, each of which is constructed in any of a plurality of physical machines, in any of the plurality of physical machines,
Based on load information of the plurality of physical machines, an arrangement destination specifying unit that specifies an arrangement destination physical machine in which the specific virtual machine is arranged;
Based on the identified physical machine load information, among the plurality of virtual machines, a machine number calculation unit that calculates the number of virtual machines that can distribute requests received by the specific virtual machine;
According to the arrangement of the specific virtual machine on the specified physical machine, among the plurality of virtual machines, the number of virtual machines corresponding to the calculated number is distributed to requests received by the specific virtual machine. A distribution destination determination unit as a distribution destination virtual machine,
A control device characterized by that. In the control method of allocating a specific virtual machine that distributes a request received to any of a plurality of virtual machines, each of which is constructed on any of a plurality of physical machines, to any of the plurality of physical machines,
Based on the load information of the plurality of physical machines, identify a physical machine to which the specific virtual machine is to be placed,
Based on the load information of the identified physical machine, out of the plurality of virtual machines, calculate the number of virtual machines that can distribute the request received by the specific virtual machine,
According to the arrangement of the specific virtual machine on the specified physical machine, among the plurality of virtual machines, the number of virtual machines corresponding to the calculated number is distributed to requests received by the specific virtual machine. A virtual machine to distribute to
A control method characterized by that.

Images