JP2015046076A - Resource management device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow resources to be moved between resource pools.SOLUTION: A resource pool having no unused resources and a resource pool having unused resources are specified, and a plurality of physical servers having unused resources out of physical servers belonging to the specified resource pool having unused resources are specified, and resource capacities of virtual machines operating on the specified physical servers are compared with capacities of unused resources of physical servers other than the physical servers to which the virtual machines are assigned. If a virtual machine operating on a physical server can be moved to another physical server, the virtual machine operating on the physical server is moved to another physical server to generate a physical server to which a virtual machine is not assigned, and this physical server to which a virtual machine is not assigned is made to belong to the resource pool having no unused resources.

Description

  The present invention relates to a technique for accommodating resources between resource pools.

  In the case of installing a plurality of virtual machines, a service is known in which a virtual machine may be installed in any way as long as it is within a resource allocated in advance as a resource pool. By belonging to a specific resource pool, a virtual machine can use physical resources with the resource set in each resource pool as an upper limit.

  In addition, when resources of a virtual machine are insufficient in the virtual machine world, it is generally performed to make resources available from other virtual machines. For example, Patent Document 1 discloses that an appropriate resource request amount is calculated and resources are rearranged in each virtual machine. As a result, more resources can be allocated to the virtual machine that needs the resources.

JP 2011-170787 A

  While resource pools have the advantage of pre-allocating resources, when processing that requires more resources than resource pools occurs, allocate more resources to the virtual machines in that resource pool. I can't. Thus, it is conceivable to allow resources to be allocated from resource pools with sufficient resources among resource pools to resource pools with no resources so that resources are distributed among virtual machines. However, in that case, there is a characteristic that the resource pool cannot be set so as to straddle physical servers (cannot be set like one and a half physical servers. It is set in units of physical servers like two physical servers). .) Because only half of a physical server cannot be transferred to another resource pool, even if you try to change the resource capacity of a resource pool in a resource pool that straddles multiple physical servers, There may be cases where resources cannot be accommodated. An object of the present invention is to make it possible to accommodate resources even in such a case.

  In order to solve the above problems, the present invention identifies a resource pool having no unused resource and a resource pool having an unused resource, and a physical server belonging to the identified resource pool having an unused resource. Identify multiple physical servers with unused resources, the resource capacity of the virtual machine running on the specified physical server, and unused resources of other physical servers other than the physical server with the virtual machine When the virtual machine operating on the physical server can be moved to the other physical server, the virtual machine operating on the physical server is moved to the other physical server, Create a physical server to which no virtual machine is assigned, and assign a physical server to which the virtual machine is not assigned to the resource pool without the unused resource. Characterized in that to belong.

  Resources can be interchanged between resource pools.

It is a block diagram of the system of embodiment of this invention. It is a figure which shows a virtual machine performance table. It is a figure which shows a resource pool performance table. It is a figure which shows a physical server performance table. It is a figure which shows a virtual machine resource pool corresponding | compatible table. It is a figure which shows a virtual machine physical server corresponding | compatible table. It is a figure which shows a resource pool physical server corresponding | compatible table. It is a flowchart showing the processing flow of rearrangement of a resource pool. It is a continuation of the flowchart of the resource pool rearrangement process.

  In the present embodiment, a system that performs resource pool (CPU) relocation processing will be described. In the present embodiment, the CPU is used. However, the present invention is not limited to this. Any device may be used as long as it represents resources such as a storage device such as a memory and an HDD and a network bandwidth.

  FIG. 1 is a configuration diagram of a system according to an embodiment of the present invention. The processing system of the present embodiment includes a monitoring server 100, a control / storage server 110, and a physical server 120. The monitoring server 100, the control / storage server 110, and the physical server 120 are connected via a network. The physical server 120 includes a server virtualization program 121, and the virtual machine 122 operates on the physical server 120 by using the program. Resource management is performed by the control / storage server 110. The resource management apparatus according to the present invention corresponds to the control / storage server 110, but is not limited to this mode, and the monitoring server 100 and the control / storage server 110 may be a single server as a resource management apparatus. The resource management device is not limited to a server, but refers to a device that manages resources.

  The physical server 120 is associated with the resource pool 130, and the resource pool 130 is set for each physical server, and sets an upper limit value of the resource capacity that can be set by the user. The user can set an arbitrary number of virtual machines having an arbitrary capacity within the resource pool. The control / storage server 110 includes a control unit 111 and a storage unit 112. The control unit 111 controls the physical server 120, the virtual machine 122, and the resource pool 130. The storage unit 112 includes a table necessary for system operation.

  The monitoring server 100 uses the monitoring unit 101 to monitor the physical server 120, the server virtualization program 121, the virtual machine 122, and the resource pool 130. Further, the monitoring server 100 transfers information obtained by monitoring to the control / storage server 110 via the network.

  The control / storage server 110 controls the correspondence between the physical server 120, the virtual machine 122, and the resource pool 130 and manages tables necessary for system operation. A virtual machine performance table 200, a resource pool performance table 300, a physical server performance table 400, a virtual machine resource pool correspondence table 800, a virtual machine physical server correspondence table 900, and a resource pool physical server correspondence table 1000 are stored in a storage unit such as an HDD. To do. In addition, the control unit 111 performs a process of changing resources (CPU, memory, storage device, etc.) from a resource pool with sufficient capacity to a resource pool with insufficient capacity.

  FIG. 2 is a configuration diagram of the virtual machine performance table 200. The virtual machine performance table 200 includes a virtual machine No 201, a CPU allocation value 202, a CPU usage value 203, a CPU allocation unused value 204, a CPU allocation shortage value 205, and a CPU allocation upper limit value 206.

  Here, the CPU value is used, but the value is not limited to this, and any value may be used as long as it is a resource value. In this case, the CPU allocation value 202 is a resource allocation value, the CPU usage value 203 is a resource usage value, the CPU allocation unused value 204 is a resource allocation unused value, and the CPU allocation insufficient value 205 is a resource allocation insufficient value. The CPU allocation upper limit 206 is a resource allocation upper limit. The resource allocation value is the allocated resource capacity, the resource usage value is the used capacity of the allocated resource capacity, the resource allocation unused value is the unused resource capacity, and the resource allocation is insufficient The value is the capacity of a resource that is insufficient, and the resource allocation upper limit value is the upper limit value of the resource capacity that can be allocated.

  For example, when the virtual machine No 201 in the virtual machine performance table 200 of FIG. 2 is virtual machine A, the CPU allocation value 202 is 2 GHz, the CPU usage value 203 is 2 GHz, the CPU allocation unused value 204 is 0 MHz, and the CPU allocation shortage value 205 is The 500 MHz, CPU allocation upper limit 206 is N / A because it is unlimited. The CPU assigned value 202 is the capacity of the CPU assigned to the virtual machine, the CPU used value 203 is the used capacity of the assigned CPU, and the CPU assigned unused value 204 is the assigned CPU. The unused capacity and CPU allocation shortage value 205 are the CPU capacity that is not sufficient for the virtual machine to process, and the CPU allocation upper limit value 206 is the upper limit value to be allocated. In this embodiment, the value of the CPU is used, but the present invention is not limited to this, and any resource allocation value, usage value, unused allocation value, insufficient allocation value, or allocation upper limit value may be used for the memory or storage device. It may be a value such as a capacity.

  FIG. 3 is a configuration diagram of the resource pool performance table 300. The resource pool performance table 300 includes a resource pool No 301, a CPU allocation value 302, a CPU usage value 303, a CPU allocation unused value 304, a CPU allocation shortage value 305, and a CPU allocation upper limit value 306. This table manages resource allocation values, usage values, allocation unused values, allocation shortage values, and allocation upper limit values for each resource pool. In this case, the CPU allocation value 302 is a resource allocation value, the CPU usage value 303 is a resource usage value, the CPU allocation unused value 304 is a resource allocation unused value, and the CPU allocation insufficient value 305 is a resource allocation insufficient value. The CPU allocation upper limit value 306 refers to a resource allocation upper limit value.

  For example, when the resource pool No 301 in the resource pool performance table 300 of FIG. 3 is resource pool A, the CPU allocation value 302 is 2 GHz, the CPU usage value 303 is 2 GHz, the CPU allocation unused value 304 is 0 MHz, and the CPU allocation shortage value 305 is The 500 MHz CPU allocation upper limit 306 is N / A because it is unlimited. In this embodiment, the value of the CPU is used, but the present invention is not limited to this, and any resource allocation value, usage value, unused allocation value, insufficient allocation value, or allocation upper limit value may be used for the memory or storage device. It may be a value such as a capacity.

  FIG. 4 is a configuration diagram of the physical server performance table 400. The physical server performance table 400 includes a physical server No 401, a CPU allocation value 402, a CPU usage value 403, and a CPU unused value 404. This table manages resource allocation values, used values, and unused values for each physical server. In this case, the CPU allocation value 402 is a resource allocation value, the CPU usage value 403 is a resource usage value, and the CPU unused value 404 is a resource unused value. For example, if the physical server No 401 in the physical server performance table 400 of FIG. 4 is physical server A, the CPU allocation value 402 is 2 GHz, the CPU usage value 403 is 2 GHz, and the CPU unused value 404 is 0 MHz. In the present embodiment, the CPU value is used, but the present invention is not limited to this, and may be a value such as the capacity of a memory or a storage device as long as it is a resource allocation value, a used value, or an unused value. .

  FIG. 5 is a configuration diagram of the virtual machine resource pool correspondence table 800. The virtual machine resource pool correspondence table 800 includes a virtual machine No 801 and a resource pool No 802. The virtual machine resource pool correspondence table 800 is a table that stores the correspondence between the virtual machine 122 and the resource pool 130.

  For example, if the virtual machine No. 801 in the virtual machine resource pool correspondence table 800 of FIG. 5 is the virtual machine A, the corresponding resource pool No. 802 becomes the resource pool A. The resource pool A indicates that the virtual machine A belongs to one kitchen.

  FIG. 6 is a configuration diagram of the virtual machine physical server correspondence table 900. The virtual machine physical server correspondence table 900 includes a virtual machine No 901 and a physical server No 902. The virtual machine physical server correspondence table 900 is a table that stores the correspondence between the virtual machine 122 and the physical server 120.

  For example, when the virtual machine No 901 in the virtual machine physical server correspondence table 900 in FIG. 6 is the virtual machine A, the corresponding physical server No 902 is the physical server A. This indicates that the physical server A has one virtual server A.

  FIG. 7 is a configuration diagram of the resource pool physical server correspondence table 1000. The resource pool physical server correspondence table 1000 includes resource pool No 1001 and physical server No 1002. The resource pool physical server correspondence table 1000 is a table that stores the correspondence between the resource pool 130 and the physical server 120.

  For example, in the resource pool A of the resource pool physical server correspondence table 1000 in FIG. 7, the corresponding physical server No 1002 is the physical server A. It means that one physical server A belongs to the resource pool A.

  8 and 9 are flowcharts of resource pool relocation processing. Next, processing will be described according to the flow of FIGS.

  The monitoring server 100 uses the monitoring unit 101 to acquire information on the CPU usage value of the virtual machine. The monitoring unit 101 monitors the usage status of the CPU of the virtual machine, and the acquired information is reflected in the storage unit 112 of the control / storage server 110. In this embodiment, the CPU value is monitored, but the present invention is not limited to this. Any resource value may be used. For example, any resource information such as a capacity of a memory or a storage device or a network bandwidth may be used.

  For example, in the virtual machine performance table 200, for each virtual machine No 201, information on the CPU allocation value 202, the CPU usage value 203, the CPU allocation unused value 204, the CPU allocation shortage value 205, and the CPU allocation upper limit value 206 is updated as needed. (Step 1101).

  The control unit 111 determines whether there is a virtual machine X in which the CPU allocation unused value 204 is 0. The control unit 111 refers to the CPU allocation unused value 204 of the virtual machine performance table 200 included in the storage unit 112 and determines whether there is a virtual machine having a value of 0. When there is no virtual machine X whose CPU allocation unused value 204 is 0, information on the CPU allocation unused value 204 of the virtual machine is acquired. When there is a virtual machine X whose CPU allocation unused value 204 is 0, the process proceeds to the next step (step 1102).

  For example, in the configuration diagram 200 of the virtual machine performance table in FIG. 2, since the CPU allocation unused value 204 of the virtual machine A is 0, the process proceeds to the next step 1103 here. A CPU allocation unused value of 0 means a virtual machine in which the resource (CPU) unused rate is 0, that is, all resources (CPUs) are used and resources are insufficient. Here, a search is made for a CPU allocation unused value (resource allocation unused value) of 0, but a CPU allocation insufficient value (resource allocation insufficient value) other than 0 may be specified.

  In this implementation, when there is a virtual machine that runs out of resources, it is checked whether there is a resource shortage in the resource pool to which the virtual machine belongs, but it is not limited to this, the resource pool performance table In reference to 300, among them, it may be controlled to identify a resource pool No301 having a resource allocation deficiency value (other than 0) and identifying a resource pool having a resource deficiency.

  It is assumed that resources are automatically allocated from the resource pool when a resource shortage occurs when there are no unused resources in the virtual machine. Therefore, when a resource shortage occurs in the virtual machine, this indicates a state where a resource shortage occurs in the resource pool. Even when the automatic distribution is not controlled in this way, as described above, it is possible to directly identify the resource shortage of the resource pool by looking at the resource pool performance table.

  The control unit 111 identifies the virtual machine X whose CPU allocation unused value 204 is 0 (step 1103). When it is determined in step 1102 that the CPU allocation unused value 204 is 0, the control unit 111 sets the virtual machine No 201 of the virtual machine X whose CPU allocation unused value 204 in the virtual machine performance table 200 is 0. Identify.

  For example, in the configuration diagram 200 of the virtual machine performance table in FIG. 2, the virtual machine No 201 whose CPU allocation unused value 204 is 0 is identified as the virtual machine A, and the process proceeds to the next step 1104.

  The control unit 111 identifies the resource pool X corresponding to the virtual machine No 201 of the virtual machine X identified in step 1103 (step 1104). The control unit 111 refers to the virtual machine resource pool correspondence table 800 in the storage unit 112 and identifies the value of the resource pool No 802 corresponding to the virtual machine No 801 of the virtual machine X identified in Step 1103. This identifies a resource pool with insufficient resources.

  For example, when it is identified as the virtual server A in step 1103, the resource pool A is identified as the corresponding resource pool No 802 in the virtual machine resource pool correspondence table 800.

  The control unit 111 checks whether there is a virtual machine whose CPU allocation unused value is not 0 (step 1105). The control unit 111 refers to the CPU allocation unused value 204 in the virtual machine performance table 200 and checks whether there is a virtual machine whose value is not 0 (other than 0). Here, when there is a virtual machine No whose CPU allocation unused value 204 is not 0, the process proceeds to the next step. If there is no virtual machine whose CPU assignment unused value 204 is not 0 in the virtual machine specified in step 1104, an alert is displayed (step 1106). If the assigned unused value is not 0, it means that there is an unused resource.

  For example, in the virtual machine performance table 200 of FIG. 2, since the CPU assignment unused value 204 of the virtual machine B, virtual machine C, virtual machine D, and virtual machine E is not 0, the process proceeds to the next step.

  The control unit 111 identifies a virtual machine Y whose CPU allocation unused value is not 0 (step 1107). The control unit 111 refers to the CPU allocation unused value 204 (resource allocation unused value) in the virtual machine performance table 200 and identifies the virtual machine No 201 of the virtual machine Y whose value is not 0.

  For example, in the virtual machine performance table 200 of FIG. 2, virtual machine B, virtual machine C, virtual machine D, and virtual machine E are identified as virtual machine No 201 whose CPU allocation unused value (resource allocation unused value) is not 0. The Here, the resource allocation unused value is other than 0, that is, a virtual machine having unused resources and having enough resources is identified. By specifying a virtual machine with sufficient resources, it is possible to specify a resource pool with unused resources and resources.

  In this embodiment, the resource unused value of the virtual machine is viewed. However, the present invention is not limited to this. The resource allocation unused value in the resource pool performance table 300 is referred to, and there is an unused resource. A resource pool having a sufficient margin may be specified.

  The control unit 111 identifies the resource pool Y corresponding to the virtual machine Y (step 1108). The control unit 111 refers to the virtual machine resource pool correspondence table 800 and identifies the resource pool No 802 of the resource pool Y corresponding to the virtual machine No 801 of the virtual machine Y.

  For example, when the virtual machine B, the virtual machine C, the virtual machine D, and the virtual machine E are identified in step 1107, the resource pool B, the resource is set as the resource pool No 802 in the virtual machine resource pool correspondence table 800 of FIG. Pool C is identified.

  The control unit 111 identifies the resource pool Z having the largest CPU allocation unused value in the resource pool Y (step 1109). Here, the resource allocation unused value is the largest, that is, the resource that has the most margin and is likely to be able to accommodate the resource to other resource pools, but control is performed so that other resource pools are identified. May be. The control unit 111 refers to the CPU allocation unused value 304 of the resource pool performance table 300, and in the resource pool No 301 corresponding to the resource pool Y, the resource pool Z having the largest CPU allocation unused value 304 of the resource pool performance table 300 is shown. Resource pool No. 301 is identified.

  For example, in the resource pool performance table 300 of FIG. 3, the CPU allocation unused value 1 GHz of the resource pool B is compared with the CPU allocation unused value 2.5 GHz of the resource pool C, and the CPU allocation unused value of the resource pool C is compared. Since 304 is larger, the resource pool C is specified as the resource pool No 301 of the resource pool Z having the larger CPU allocation unused value 304.

  The control unit 111 identifies the physical server Z having the lowest CPU usage value (resource usage value) among the physical servers corresponding to the resource pool Z identified in step 1109 (step 1110). Although the CPU usage value (resource usage value) is low, the CPU unused value (resource unused value) may be large. Here, there is an unused resource among the physical servers that belong to the resource pool that has unused resources and that has sufficient resources, and the physical server that has the most available resources is identified. Here, the resource having the most resources is specified, but the present invention is not limited to this, and a plurality of physical servers having unused resources and having resources may be specified.

  The control unit 111 refers to the resource pool No 1001 of the resource pool physical server correspondence table 1000 of the storage unit 112 and identifies the physical server No 1002 of the physical server corresponding to the resource pool No of the resource pool Z. Then, among the identified physical servers, the physical server No 401 with the lowest CPU usage value 403 (high CPU unused value) is identified among the physical server performance table 400 that has been referenced.

  For example, when the resource pool C is identified in step 1109, the physical server C and the physical server D are identified as the physical server No 1002 corresponding to the resource pool C in the resource pool physical server correspondence table 1000 of FIG. Then, from the physical server performance table 400 of FIG. 4, the physical server C is identified as the physical server No 401 with the lowest CPU usage value (high CPU unused value). In this case, it may be controlled to select a value with a small CPU usage value 403, or may be controlled to select a value with a large CPU unused value, and a physical server with a low CPU usage value is selected. It is not limited as long as it is controlled as described above.

  The control unit 111 identifies a physical server Z ′ having a low CPU usage value (a high CPU unused value) other than the physical server Z among the physical servers corresponding to the resource pool Z (step 1111). As a result, the physical server having the most margin among the physical servers having unused resources in the same resource pool and the physical server having the next margin are specified. When a plurality of physical servers having a margin are specified in step 1110, it is not necessary to identify a physical server having the second margin.

  The control unit 111 refers to the physical server No 1002 in the resource pool physical server correspondence table 1000 of the storage unit 112 and identifies the physical server No 1002 of the physical server corresponding to the resource pool Z. Then, among the identified physical servers, while referring to the physical server performance table 400, the physical server Z ′ other than the physical server Z identified in step 1110 has a low CPU usage value (a high CPU unused value). To do.

  For example, in step 1110, the physical server C is identified as a physical server having a low CPU usage value (a high CPU unused value). The physical server D is identified as a physical server other than the physical server C among the physical servers corresponding to the resource pool C. Among them, the physical server D is specified as the physical server No 404 having a low CPU usage value (a high CPU unused value). In this embodiment, the physical server in the resource pool C other than the physical server C is only the physical server D, and the physical server D is selected. However, if there are other physical servers such as the physical server E, look at the CPU usage rate. The one with the low usage rate is selected. Among the physical servers in the same resource pool, the top two physical servers with low resource usage values (high unused values) are identified.

  The control unit 111 calculates whether the value obtained by subtracting the CPU usage value (resource usage value) of the physical server Z from the CPU unused value (resource usage value) of the physical server Z ′ is greater than zero. (Step 1112). The control unit 111 refers to the physical server performance table 400 of the storage unit 112, refers to the CPU unused value 404 of the physical server Z ′, and determines the CPU used value of the physical server Z from the CPU unused value 404 of the physical server Z ′. Calculate whether the value obtained by subtracting 403 is greater than zero. Here, if there is a calculation result larger than 0, the process proceeds to the next step. If no calculation result is greater than 0, the process returns to step 1101. Here, in order to determine whether or not the resource used in the physical server Z can be moved to the physical server Z ', two magnitude relationships are seen.

  For example, if the physical server C is identified in step 1110, the physical server performance table 400 in FIG. 4 is referred to, and the CPU unused value 404 value of the physical server C is referred to from the CPU unused value 404 value of the physical server D. Since the value obtained by subtracting is larger than 0, the process proceeds to the next step.

  In the process so far, it has been confirmed whether virtual machines can be consolidated to either physical server between two physical servers with low resource usage values. If two physical servers with low resource usage values are not selected and multiple physical servers are specified, the virtual machine is switched between the physical servers and the physical machine that does not operate You will create a server. In other words, the virtual machine operating on the physical server is compared with the resource capacity of the virtual machine operating on the specified physical server and the capacity of unused resources of the specified other physical server. If the virtual machine can be moved to the physical server, the virtual machine is sequentially moved to the physical server, and a physical server to which no virtual machine is allocated is created.

  Specifically, the identified physical server is collated with the virtual machine physical server correspondence table 900 to identify a virtual machine that operates on the physical server, and the resource usage value of the virtual machine performance table 200 determines the virtual machine's Identify resource capacity. The unused resource capacity of the physical server is specified by the resource unused value in the physical server performance table 400 and compared.

  In that case, compare the resource usage value of the physical server with the unused value of the resource of another physical server, and if the values are close or far apart, add multiple resource usage values It is preferable to compare the unused values of other physical servers and control the virtual machines to move preferentially from those with close values. If the unused physical capacity of the identified physical server is larger than the resource capacity of the identified virtual machine, the virtual machine can be moved. By moving the virtual machine, a physical server without virtual machine assignment is created. As described above, when the unused resource capacity of a physical server is larger than the sum of the capacity of multiple virtual machines on the same physical server, it is possible to move multiple virtual machines. It is possible to create a physical server to which no virtual machine is allocated.

  It is also conceivable to create a physical server without virtual machine allocation by moving a plurality of virtual machines on the same physical server to different physical servers. In this case, the virtual machine resource capacity and the unused resource capacity of other physical servers are moved to each virtual machine in order. This is a process due to restrictions because it can only belong to a resource pool in units of physical servers.

  In this step, the CPU usage value (resource usage value) of the physical server Z is compared with the CPU unused value (resource unused value) of a physical server other than the physical server Z corresponding to the resource pool Z. Then, when the CPU usage value of the physical server Z becomes large, the processing is returned to step 1101. However, the processing is not limited to this. For example, the remaining unused resource pool amount is desired to be increased. In this case, when the difference is 100 or less, control may be performed so as to return to step 1101. The difference value may be received from the user in a timely manner, or may be a value stored in the storage device from the beginning.

  If it is greater than 0, the control unit 111 determines that the virtual machine in the physical server Z can be moved to the physical server Z ′, and the corresponding destination of the virtual machine Z belonging to the physical server Z is transferred to the physical server Z ′. Change (step 1113). The control unit 111 changes the correspondence destination of the virtual machine No 901 of the virtual machine Z in the virtual machine physical server correspondence table 900 from the physical server No 902 of the physical server Z to the physical server No 902 of the physical server Z ′. This makes it possible to create a physical server without virtual or machine assignment.

  For example, when the physical server C is specified as the physical server Z in step 1112 and the process proceeds to step 1113, the correspondence destination of the virtual machine D in the virtual machine physical server correspondence table 900 of FIG. 6 is changed from the physical server C to the physical server D. And change.

  Here, since the only virtual machine running on the physical server C is the virtual machine D, the virtual machine D is moved to the physical server D. If there are a plurality of virtual machines, all the virtual machines are moved to the physical server. Move to D. This is because there is no virtual machine running on the physical server C, so that the physical server C can be accommodated in a resource pool with heavy processing.

  In addition, when there are a plurality of virtual machines on the physical server C, not all of them are moved to the same physical server (here, the physical server D), but if there are other physical servers in the resource pool, If the unused value of the physical server is compared with the usage rate of the virtual machine, and the usage rate of the virtual machine is smaller, processing is performed so that the virtual machines in the physical server C are distributed to a plurality of physical servers. Also good.

  Next, the control unit 111 changes the correspondence destination of the physical server Z from the resource pool Z to the resource pool X (step 1114). The control unit 111 changes the correspondence destination of the physical server No 1002 of the physical server Z in the resource pool physical server correspondence table 1000 from the resource pool No 1001 of the resource pool Z to the resource pool No 1001 of the resource pool X.

  For example, when the physical server Z is identified as the physical server C in step 1110 and the process moves to step 1114, the resource pool C corresponds to the resource pool physical server correspondence table 1000 in FIG. Change to pool A.

  The control unit 111 changes the correspondence destination of the virtual machine X to the physical server Z (step 1115). The control unit 111 changes the correspondence destination of the virtual machine No. 901 of the virtual machine X in the virtual machine physical server correspondence table 900 to the physical server X. To the physical server No.

  For example, when the virtual machine A is specified as the virtual machine X in step 1103 and the physical server C is specified as the physical server Z in step 1110, the corresponding destination of the virtual machine A in the virtual machine physical server correspondence table 900 of FIG. , Change to physical server C.

  The monitoring unit 101 checks whether the CPU allocation shortage value of the resource pool X is 0. The monitoring unit 101 refers to the resource pool performance table 300 of the storage unit 111 and checks whether the CPU allocation shortage value 305 corresponding to the resource pool No 301 of the resource pool X is 0. If the CPU allocation shortage value 305 is not 0, the process returns to step 1105. If the CPU allocation shortage value 305 corresponding to the resource pool No. 301 of the resource pool X is 0, the process is terminated (step 1116).

  For example, when the correspondence destination of the resource pool of the physical server C in the resource pool physical server correspondence table 1000 of FIG. 7 is changed from the resource pool C to the resource pool A in step 1114, it corresponds to the resource pool No301 of the resource pool A. The CPU allocation shortage value 305 is 0, and the process ends.

100: monitoring server 101: monitoring unit 110: control / storage server 111: control unit 112: storage unit 120: physical server 121: server virtualization program 122: virtual machine 130: resource pool 200 : Virtual machine performance table, 300: Resource pool performance table, 400: Physical server performance table, 500: Virtual machine memory performance table, 800: Virtual machine resource pool correspondence table, 900: Virtual machine physical server correspondence table, 1000: Resource pool Physical server correspondence table

Claims (5)

Identify resource pools with no unused resources and resource pools with unused resources, and select physical servers with unused resources from among the physical servers that belong to the specified resource pool with unused resources. Identify multiple
The resource capacity of the virtual machine operating on the specified physical server is compared with the capacity of unused resources of another specified physical server other than the physical server on which the virtual machine is located. If the operating virtual machine can be moved to the other physical server, the virtual machine operating on the physical server is moved to the other physical server to create a physical server without virtual machine allocation,
A resource management apparatus, wherein a physical server to which no virtual machine is allocated belongs to a resource pool having no unused resources. The resource management device according to claim 1 is:
Furthermore,
Of the plurality of identified physical servers, the resource usage value of the first physical server with the lowest resource usage value is compared with the resource unused value of the second physical server with the lower resource usage value, If the resource unused value of the second physical server is higher, move the virtual machine running on the first physical server to the second physical server to create a physical server without virtual machine allocation A resource management device characterized by: The resource management device according to claim 1 or 2,
Furthermore,
A resource management apparatus characterized by identifying a resource pool having no unused resources as a resource pool having insufficient resources. The resource management device according to any one of claims 1 to 3,
Furthermore,
A resource pool having the largest unused resource is identified, and a physical server having an unused resource is identified from among the physical servers belonging to the resource pool having the highest identified unused resource. Resource management device. Identifying a resource pool with no unused resources and a resource pool with unused resources;
Identifying a plurality of physical servers with unused resources among physical servers belonging to a resource pool with the specified unused resources;
A virtual machine that operates on the physical server by comparing the resource capacity of the virtual machine that operates on the specified physical server and the capacity of unused resources on other physical servers other than the physical server on which the virtual machine is located Moving the virtual machine to the physical server when the virtual machine can be moved to another physical server;
A resource management method comprising: creating a physical server to which no virtual machine is assigned, and causing the physical server to which the virtual machine is not assigned to belong to a resource pool having no unused resources.

Images