JP5412304B2 - Resource providing system, resource providing method, and virtualization apparatus - Google Patents

Description

  The present invention relates to a resource providing system, a resource providing method, and a virtualization apparatus that provide computer resources in response to a computer resource provision request transmitted from a user terminal.

Research has been conducted on a virtualization technology that virtualizes computer resources such as a processor, a memory, and a disk of a computer device, and starts a plurality of virtual machines on one physical machine. Also, multiple virtualization devices with such a virtualization function are prepared, and virtual machines can be started up by the virtualization devices in response to resource provision requests transmitted from user terminals via the network, and can be used for user terminals The resource provision system to provide is being studied. According to a resource provision service, also called a cloud service provided by such a resource provision system, computer hardware and software can be used on a pay-as-you-go basis rather than purchased or leased, or used as needed. It is possible to do.
Such a resource providing system is generally a so-called public cloud that provides a large number of physical machine computer resources prepared in a data center or the like through the Internet. In addition, in recent years, a so-called private cloud in which such a cloud is applied to a computer system in a closed environment such as an organization such as a specific company has been studied (for example, Non-Patent Document 1).

"From cloud details to basic technologies", Nikkei Business Publications, April 2009

  However, in order to immediately provide computer resources in response to a resource provision request from a user terminal, a provider of the resource providing system prepares computer resources of a scale suitable for when the required amount of computer resources is maximized. It is necessary to keep it. However, the required amount of computer resources is considered to fluctuate according to the time zone, day of the week, season, and the like. For this reason, if a provider providing a service prepares a computer resource of a size that matches the amount of computer resources required to be maximum, a surplus of computer resources will be generated during a period when the amount of requested requests is small. The resource cannot be used effectively. In particular, when constructing a large-scale private cloud, it is conceivable to divide a physical machine in the cloud into a plurality of clusters and use them in different organizations for each cluster. At this time, since a certain number of virtualization devices are prepared in advance for each of a plurality of clusters, compared to the case where a large number of virtualization devices are prepared as a whole like a public cloud, a computer that responds to a provision request It is considered that there is a possibility that excess and deficiency will occur in the amount of resources. Here, even in a plurality of resource providing systems that provide computer resources to a limited range of users, a stable resource that increases the use efficiency of computer resources and immediately provides computer resources in response to user resource provision requests. It is desirable to provide a delivery system.

  The present invention has been made in view of such a situation, and improves the use efficiency of computer resources, and provides a stable resource provision system, a resource provision method, and a resource provision method that immediately provide computer resources in response to a resource provision request from a user. A virtualization device is provided.

In order to solve the above-described problems, the present invention provides a plurality of virtualization devices that provide virtualized computer resources to user terminals connected via a network, and a plurality of virtualizations via a network. A resource providing system including a plurality of resource providing devices connected to the device, wherein the resource providing device identifies a plurality of virtualization devices associated as subordinates of the own device among the plurality of virtualization devices A subordinate virtualization apparatus information storage unit that stores subordinate virtualization apparatus identification information, and a load situation value storage that stores load situation values indicating the load situations of a plurality of virtualization apparatuses identified by the subordinate virtualization apparatus identification information Load status values of the virtualization device collected from the virtualization device identified by the subordinate and the subordinate virtualization device identification information and stored in the load status value storage unit A value acquisition unit, a virtualization device, a virtualization device be an additional request for assigning to the own device, the additional request included the identification information of the virtualization apparatus is transmitted, under virtualized identification information A subordinate virtualization apparatus update unit to be stored in the subordinate virtualization apparatus information storage unit as apparatus identification information, and a subordinate virtual stored in the subordinate virtualization apparatus information storage unit in response to a computer resource provision request transmitted from the user terminal and a virtualization device determining unit for providing computer resources to virtual device identified by the apparatus identification information, the virtualization device, a plurality of the resource providing device, to the load status value storing unit of the resource providing device The second load situation value collection unit that collects the stored load situation value and the load situation value for each of the plurality of resource providing devices collected by the second load situation value collection unit are compared. Based on the comparison result, it is determined that the own resource device is attributed by the attribute resource providing device determination unit and the attribute resource providing device determination unit that determines which of the resource providing devices to which the resource device belongs. The resource providing apparatus includes an additional request transmitting unit that transmits an additional request including identification information of the own apparatus.

  In addition, the present invention provides, for a plurality of virtualization apparatuses connected via a network, a game virtualization apparatus that changes a resource providing apparatus to which the own apparatus belongs by transmitting an addition request to the resource providing apparatus, and an addition A dedicated virtualization device that does not change the resource providing device to which the device belongs without transmitting a request, and the subordinate virtualization device information storage unit of the resource providing device is provided to the user terminal The exclusive subordinate virtualization apparatus identification information for identifying the exclusive virtualization apparatus according to the amount of the computer resource assigned, and the game subordinate virtualization apparatus identification information for identifying the game virtualization apparatus that has transmitted the addition request are stored. It is characterized by.

  Further, according to the present invention, the virtualization apparatus determination unit of the resource providing apparatus responds to the provision request of the computer resource transmitted from the user terminal with the subordinate virtualization apparatus identification information stored in the subordinate virtualization apparatus information storage unit. Among them, the exclusive subordinate virtualization apparatus identification information is preferentially selected over the game subordinate virtualization apparatus identification information, and computer resources are provided to the virtualization apparatus identified by the selected subordinate virtualization apparatus identification information. And

  Furthermore, the present invention is characterized in that the addition request transmission unit of the virtualization apparatus transmits an addition request when a predetermined condition is satisfied.

  Further, according to the present invention, the second load status value collection unit of the virtualization apparatus collects load status values from a plurality of resource providing devices over a plurality of times, and the belonging resource providing device determination unit includes the second load According to the transition status of a plurality of load status values collected by the status value collection unit, it is determined which of the plurality of resource providing devices the own device belongs to.

Further, the present invention is the user terminal connected via a network, a plurality of virtualization apparatus that provides computing resources of its own device virtualized, associated as under the own apparatus among the plurality of virtualization apparatuses A subordinate virtualization apparatus information storage unit storing subordinate virtualization apparatus identification information for identifying a plurality of virtualization apparatuses, and a load status indicating a load situation of the plurality of virtualization apparatuses identified by the subordinate virtualization apparatus identification information A resource providing method of a resource providing system comprising a plurality of resource providing devices connected to a plurality of virtualization devices via a network , the resource providing device comprising a load status value storage unit for storing values The resource providing device collects the load status value of the virtualization device from the virtualization device identified by the subordinate virtualization device identification information and stores the load status value A step of storing, the virtualization apparatus, a plurality of the resource providing device, step a, the collected plurality of resource providing device every collecting the load status values stored load status value in the storage unit of the resource providing device Based on the comparison result of comparing the load status values, a step of determining which of the plurality of resource providing devices the own device belongs to, and a resource providing device that has been determined to attribute the own device by the determination And the step of transmitting an addition request including the identification information of the own device, and the resource providing device virtualizes the identification information when an additional request including the identification information of the virtualization device is transmitted from the virtualization device. In response to a request to provide computer resources transmitted from the user terminal, and a step of storing in the subordinate virtualization apparatus information storage unit as apparatus identification information Characterized by comprising the step of providing a computer resources virtualization device identified by the subordinate virtualization apparatus identification information stored under the virtualization apparatus information storing unit.

In addition, the present invention provides a plurality of virtualization devices that provide virtualized computer resources to user terminals connected via a network, and a plurality of virtualization devices connected to a plurality of virtualization devices via a network. A subordinate virtualization apparatus information storage unit for storing subordinate virtualization apparatus identification information for identifying a plurality of virtualization apparatuses associated as subordinates of the own apparatus among the plurality of virtualization apparatuses; From the load status value storage unit that stores the load status values indicating the load status of the plurality of virtualization devices identified by the subordinate virtualization device identification information, and the virtualization device identified by the subordinate virtualization device identification information, A first load status value collection unit that collects the load status value of the virtualization apparatus and stores the load status value in the load status value storage unit, and an additional request including identification information of the virtualization device is transmitted from the virtualization device Then, the subordinate virtualization apparatus update unit that stores the identification information as the subordinate virtualization apparatus identification information in the subordinate virtualization apparatus information storage unit, and the subordinate virtualization apparatus according to the computer resource provision request transmitted from the user terminal A virtualization apparatus in a resource provision system comprising: a virtualization apparatus determination section that provides a virtualization apparatus identified by subordinate virtualization apparatus identification information stored in an information storage section; there are, collected from multiple resource providing device, a second load status value collecting unit that collects the load status values stored in the load status value storing unit of the resource providing device, the second load status value collecting unit Based on the comparison result obtained by comparing the load status values for each of the plurality of resource providing devices, any resource provision among the plurality of resource providing devices An attribution resource providing device determination unit that determines whether the device belongs to the device and a resource providing device that has been determined to attribute the device by the attribution resource provision device determination unit And an additional request transmission unit for transmission.

  As described above, according to the present invention, a plurality of resource providing apparatuses store subordinate virtualization apparatus identification information for identifying a plurality of virtualization apparatuses associated as subordinates of the own apparatus, and control subordinate virtualization A computer resource that collects and stores a load status value from a device, stores it as subordinate virtualization device identification information when an additional request including the identification information of the virtualization device is transmitted from the virtualization device, and is transmitted from the user terminal In response to the provision request, the virtual device under management provides the computer resources, and the virtualization device collects the load status values of the virtual devices under the resource providing device from a plurality of resource providing devices, and the collected load Based on the status value comparison result, it is determined which of the plurality of resource providing devices the own device belongs to, and an addition request is sent In raising the efficiency of utilization of computer resources, it is possible to provide a stable resource providing system that provides computing resources immediately in response to the resource providing request from the user. For example, when a private cloud is provided by a resource providing system, the virtualization device can determine and change the cluster to which the virtualization device belongs according to the load status, so computer resources between clusters according to the load status Can be accommodated.

It is a block diagram which shows the structural example of the resource provision system by one Embodiment of this invention. It is a figure which shows the item example of the node controller information memorize | stored in the node controller memory | storage part of the cluster controller by one Embodiment of this invention. It is a figure which shows the item example of the instance information memorize | stored in the instance memory | storage part of the cluster controller by one Embodiment of this invention. It is a figure which shows the item example of the resource status information memorize | stored in the resource status memory | storage part of the cluster controller by one Embodiment of this invention. It is a figure which shows the data example of the node controller information memorize | stored in the node controller memory | storage part of the cluster controller by one Embodiment of this invention. It is a figure which shows the structural example of the node controller by one Embodiment of this invention. It is a figure which shows the example of data of the instance information which the node controller by one Embodiment of this invention memorize | stores. It is a figure which shows the data example of the resource status information which the node controller by one Embodiment of this invention memorize | stores. It is a figure which shows the example of data of the cluster controller information which the node controller by one Embodiment of this invention memorize | stores. It is a figure which shows the example of the resource status information which the resource status collection part 423 of the node controller collects for every fixed time by one Embodiment of this invention. It is a figure which shows the data example of the cluster controller information to which the own apparatus which the node controller by one Embodiment of this invention belongs belongs. It is a sequence diagram which shows the operation example of the resource provision system by one Embodiment of this invention.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing a configuration of a resource providing system 1 according to the present embodiment. The resource providing system 1 includes a user terminal 100, a cloud controller 200, a plurality of cluster controllers 300-N (cluster controller 300-1, cluster controller 300-2,...), And a plurality of node controllers 400 ( Exclusive node controller 410-1, exclusive node controller 410-2, exclusive node controller 410-3, exclusive node controller 410-4,..., Game node controller 420-1, game node controller 420-2,. And a computer device. These computer apparatuses are connected via a network. Here, the plurality of node controllers 400 have a function of virtualizing their own computer resources and providing them to the user terminal 100. In particular, it is necessary to distinguish between the dedicated node controller 410-N and the game node controller 420-N. If there is no node controller 400, the description will be made.

  Here, the plurality of cluster controllers 300-N have the same configuration, and will be described as the cluster controller 300 when there is no particular need to distinguish them. The cluster controller 300 manages the node controller 400 belonging to the own device among the plurality of node controllers 400 as the subordinate node controller 400. That is, the node controller 400 belongs to one cluster controller 300 under its own device, and one cluster controller 300 has a plurality of node controllers 400 under its own device. Here, a cluster managed by the cluster controller 300-1 is referred to as cluster A, and a cluster managed by the cluster controller 300-2 is referred to as cluster B. For example, when the node controller 400 provides a computer resource in response to a computer resource provision request (instance activation request) from the cluster controller 300-1 of the cluster A, the node controller 400 uses the cluster controller 300-1 (cluster Belonging to A). Specifically, as will be described later, the node controller 400 stores the ID of the cluster controller 300 to which the node controller 400 belongs in its own storage area, and the cluster controller 300 stores the ID of the node controller 400 to which it belongs to its own storage area. To remember.

  The node controller 400 is a computer device that provides a virtualized computer resource of its own device to the user terminal 100 connected via a network. The plurality of node controllers 400 are divided for each cluster, which is a unit for providing a resource providing service, and belong to one of the cluster controllers 300 among the plurality of cluster controllers 300. Here, to the plurality of node controllers 400 connected via the network, a game node controller 420-N that changes the cluster controller 300 to which the own device belongs by transmitting a belonging information change request to the cluster controller 300; An exclusive node controller 410-N that does not change the cluster controller 300 to which the own device belongs without transmitting an attribution information change request is included. The plurality of dedicated node controllers 410-N have the same configuration, and will be described as the dedicated node controller 410 when there is no need to distinguish them. Similarly, the plurality of game node controllers 420-N have the same configuration, and will be described as game node controllers 420 when there is no need to distinguish them. Here, the exclusive node controller 410-1, the exclusive node controller 410-2, and the game node controller 420-1 belong to the cluster A, and the exclusive node controller 410-3, the exclusive node controller 410-4, and the game A description will be given assuming that the node controller 420-2 belongs to the cluster B.

  The user terminal 100 is a computer terminal that transmits a computer resource provision request to the cloud controller 200 and uses a computer resource provided in response to the transmitted provision request. The user terminal 100 has an Internet browser function, for example, and connects to a website provided by the cloud controller 200 based on a predetermined Uniform Resource Locator (URL). The user terminal 100 transmits a computer resource provision request to the cloud controller 200 and receives from the cloud controller 200 the IP (Internet Protocol) address of the instance activated in the node controller 400 that is the computer resource provider. The user terminal 100 connects to the node controller 400 based on the received IP address and uses computer resources of the node controller 400. Here, an instance is a virtual machine started on a slot which is a virtual computer resource obtained by virtualizing physical hardware resources such as the CPU, memory, and disk of the node controller 400 and cutting them out as a specific unit. is there.

  The cloud controller 200 provides a website that is an interface from the user terminal 100 to the resource providing service, and receives a computer resource provision request transmitted from the user terminal 100. The cloud controller 200 transmits a computer resource provision request to the cluster controller 300 in response to the received provision request. For example, the cloud controller 200 stores in advance a user account corresponding to the user terminal 100 that provides computer resources in its storage area. Further, a cluster controller 300 that provides computer resources for each user account is stored in association with each other. When the cloud controller 200 receives a computer resource provision request from the user terminal 100, the cloud controller 200 transmits a computer resource provision request to the cluster controller 300 that manages the cluster corresponding to the user terminal 100 that has transmitted the provision request. In addition, the cloud controller 200 stores an image file of a virtual machine activated by the node controller 400.

  The cluster controller 300 is a computer device that is connected to a plurality of node controllers 400 via a network and controls the operation of the node controller 400 under its control. The cluster controller 300 includes a node controller storage unit 310, an instance storage unit 320, a resource status storage unit 330, a resource status collection unit 340, a node controller information update unit 350, and a node controller determination unit 360. .

  The node controller storage unit 310 includes node controller information including node controller IDs for identifying a plurality of node controllers 400 associated as subordinates of the own device among the plurality of node controllers 400 connected via the network. Remembered. Here, the node controller information stored in the node controller storage unit 310 includes a node controller ID that identifies the dedicated node controller 410 according to the amount of computer resources that is determined to be provided to the user terminal 100, and A node controller ID for identifying the game node controller 420 belonging to itself is included. The node controller storage unit 310 stores resource status information indicating the resource status of the plurality of node controllers 400 identified by the node controller ID.

  FIG. 2 is a diagram illustrating an example of items of node controller information stored in the node controller storage unit 310. Here, the node controller ID, the IP address of the node controller, the memory capacity of the node controller, the free memory capacity of the node controller, the disk capacity of the node controller, the free disk capacity of the node controller, and the number of CPUs of the node controller And the number of free CPUs of the node controller and a true / false value (flag) indicating whether the node controller is the game node controller 420 (whether it is the exclusive node controller 410) are stored in association with each other. Here, the resource status information of the free memory capacity of the node controller, the free disk capacity of the node controller, and the free CPU count of the node controller is acquired and stored by the resource status collection unit 340. The flag indicating whether or not the node controller 420 is a battle node controller 420 indicates that the corresponding node controller is the exclusive node controller 410 if the value is “0”, for example, and if the value is “1”, the corresponding node controller Indicates that it is a batting node controller 420.

  The instance storage unit 320 stores instance information related to an instance activated in the node controller 400 under its own device. FIG. 3 is a diagram illustrating an example of items of instance information. The instance information includes an instance ID, a node controller ID of a node controller that operates the instance, a memory capacity allocated to the instance, a disk capacity allocated to the instance, and a CPU allocated to the instance. Numbers are stored in association with each other.

  The resource status storage unit 330 stores resource status information including a total value of resource statuses of the plurality of node controllers 400 identified by the node controller ID stored in the node controller storage unit 310. FIG. 4 is a diagram illustrating an example of items of resource status information. The resource status information includes the total memory of the entire cluster, the total amount of free memory of the entire cluster, the total amount of disk of the entire cluster, the total amount of free disk of the entire cluster, the total number of CPUs of the entire cluster, and the total number of free CPUs of the entire cluster. Are stored in association with each other. The resource status information is generated and stored by the resource status collection unit 340 based on the resource status information stored in the node controller storage unit 310. Here, the free space of the computer resource may be stored in units of slots corresponding to a predetermined computer resource amount. For example, the computer resource amount of 1 GB of memory capacity, disk capacity of 20 GB, and the number of CPUs of 1 may be 1 slot, and the free space of computer resources may be stored in slot units.

  The resource status collection unit 340 collects the resource status information of the node controller 400 from the node controller 400 identified by the node controller ID stored in the node controller storage unit 310, and the node controller storage unit 310 and the resource status storage Store in the unit 330. The resource status collection unit 340 periodically acquires resource status information by, for example, SOAP (Simple Object Access Protocol) communication or SNMP (Simple Network Management Protocol) communication. Here, the resource status collection unit 340 is a computer having a memory capacity, a free memory capacity, a disk capacity, a free disk capacity, a CPU count, and a free CPU count from each of the node controllers 400 under its control. The resource status information is received and stored in the node controller storage unit 310. Also, the resource status collection unit 340 collects the computer resource status information from all the node controllers 400 under its control, and then calculates the total value of the values of the computer resource status information stored in the node controller storage unit 310. , It is stored in the resource status storage unit 330.

  The node controller information updating unit 350 updates and stores the node controller information stored in the node controller storage unit 310 in response to the attribution information change request transmitted from the game node controller 420. For example, when an add request including the node controller ID of the game node controller 420 is transmitted from the game node controller 420, the node controller information update unit 350 stores the node controller information corresponding to the received node controller ID in the node controller storage. It is added to the unit 310 and stored. Further, when a delete request including the node controller ID of the game node controller 420 is transmitted from the game node controller 420, the node controller information update unit 350 stores the node controller information corresponding to the received node controller ID in the node controller storage. It deletes from the part 310. As a result, the cluster controller 300 to which the node controller that transmitted the attribution information change request belongs is changed.

  In response to the computer resource provision request transmitted from the user terminal 100, the node controller determination unit 360 transmits the computer resource provision request to the node controller 400 identified by the node controller ID stored in the node controller storage unit 310. The user terminal 100 is provided with computer resources. Here, the node controller determination unit 360 refers to the resource status information of the node controller 400 stored in the node controller storage unit 310, and extracts the node controller 400 that can provide the requested amount of computer resources. Also, the node controller determination unit 360 preferentially selects the node controller ID indicating the dedicated node controller 410 over the node controller ID indicating the game node controller 420 among the node controller IDs stored in the node controller storage unit 310. Thus, the computer resource is provided to the node controller 400 identified by the selected node controller ID.

  For example, it is assumed that the node controller information stored in the node controller storage unit 310 is stored as shown in FIG. Here, it is assumed that a request for providing computer resources having a memory 1G, a disk capacity 20G, and a CPU number 1 (one slot) is transmitted from the user terminal 100. The node controller determination unit 360 compares the requested amount of the computer resource included in the computer resource provision request with the resource status information stored in the node controller storage unit 310 to determine the dedicated node controller 410-2 or the play node controller 420. It is determined that the computer resource corresponding to the provision request amount can be provided by -1. The node controller storage unit 310 refers to the flag and preferentially selects the node controller ID of the exclusive node controller 410-2 that is the exclusive node. In this way, the node controller determination unit 360 preferentially selects the dedicated node controller over the game node controller and provides computer resources, thereby reducing the operation of the game node controller 420 and making the cluster easy to move. Can keep. In other words, when resources in a cluster different from the cluster to which the game node controller 420 belongs become insufficient, the instance operating on its own computer resource is moved to another node controller 400, and the attribution destination is assigned to the different cluster. The possibility of changing can be increased.

  The exclusive node controller 410 is a node controller 400 that belongs to a fixed cluster. Here, functional units that the exclusive node controller 410 and the game node controller 420 have similarly will be described with reference to the node controller 400 shown in FIG. The node controller 400 is a computer device that virtualizes its computer resources and provides it to the user terminal 100. The node controller 400 includes a node management unit 401, a virtualization unit 402, and a computer resource 404. The node management unit 401 communicates with the node management unit 401 included in the cluster controller 300 and other node controllers 400 to control provision of computer resources by the node controller 400. The virtualization unit 402 virtualizes the computer resource 404 and activates the instance 403 (virtual machine) on the virtualized computer resource. Here, the virtualization unit 402 may determine a plurality of slots on the computer resource 404 in accordance with a predetermined amount of computer resources, and activate instances in slot units.

  Here, the node management unit 401 measures the usage status of the computer resource 404 of its own device and stores the resource status information in its own storage area. 7 and 8 are diagrams illustrating examples of items of information that the node management unit 401 stores in its own storage area. As illustrated in FIG. 7, the node management unit 401 stores an instance ID for identifying an activated instance on the computer resource 404 of the own device. Further, as illustrated in FIG. 8, the node management unit 401 stores the resource status information of the computer resource 404 of the own device. The resource status information includes the own device memory capacity, the own device free memory capacity, the own device disk capacity, the own device free disk capacity, the own device CPU number, and the own device CPU number. Are stored in association with each other. Further, the node management unit 401 has a function of moving a plurality of instances 403 activated by the computer resource 404 of the own device to another node controller 400 by communicating with the node management unit 401 of the other node controller 400. (Live migration).

  Returning to FIG. 1, the dedicated node controller 410 includes a node management unit 411 and a virtualization unit 412. The node management unit 411 and the virtualization unit 412 correspond to the above-described functional units of the same name provided in the node controller 400. The exclusive node controller 410 may include the function units of the resource status collection unit 423 to the attribution information change request unit 425 included in the game node controller 420, but operates them to change the cluster controller 300 to which it belongs. There is nothing. On the other hand, the game node controller 420 changes the cluster controller 300 to which it belongs according to the resource status of the cluster controller 300. That is, the provider of the resource providing service by the resource providing system 1 sets the dedicated node controller 410 in an amount corresponding to the amount of computer resources determined based on a contract such as an SLA (Service Level Agreement) with the user. Be attributed to Such a dedicated node controller 410 can always maintain a certain level of computer resources under the control of a specific cluster controller 300.

  The game node controller 420 is a node controller 400 that changes the cluster controller 300 to which the own device belongs by transmitting a belonging information change request to the cluster controller 300. The game node controller 420 includes a node management unit 421, a virtualization unit 422, a resource status collection unit 423, an belonging cluster determination unit 424, and an belonging information change request unit 425. The node management unit 421 and the virtualization unit 422 have the same configuration as the above-described functional unit of the same name provided in the exclusive node controller 410.

  The resource status collection unit 423 collects the resource status information stored in the resource status storage unit 330 of the cluster controller 300 from the plurality of cluster controllers 300 and stores it in its own storage area. Here, the resource status collection unit 423 stores in advance a list of cluster controllers 300 that can become parents to which the resource status collection unit 423 belongs, as shown in FIG. The resource status collection unit 423 collects the resource status information stored in the resource status storage unit 330 of the cluster controller 300 included in such a list and stores it in its own storage area. Further, the resource status collection unit 423 may collect and store the resource status information from the cluster controller 300 over a plurality of times. For example, the resource status collection unit 423 collects the resource status information from the cluster controller 300 at predetermined time intervals, and stores the resource status information at fixed time intervals as a history. Thus, the belonging cluster determination unit 424 can determine the belonging cluster in accordance with the transition of the resource status.

  The belonging cluster determination unit 424 compares the resource status information for each of the plurality of cluster controllers 300 collected by the resource status collection unit 423, and automatically determines which cluster controller 300 of the plurality of cluster controllers 300 is based on the comparison result. It is determined whether to attach the device. For example, the belonging cluster determination unit 424 compares the resource status information of a plurality of cluster controllers 300 and determines that the own device belongs to the cluster controller 300 with less free capacity.

  Here, as long as the resource status collection unit 423 has collected a plurality of resource status information, the belonging cluster determination unit 424 determines which cluster of the plurality of cluster controllers 300 according to the transition status of the plurality of resource status information. You may make it determine whether the own apparatus belongs to the controller 300. FIG. FIG. 10 is a diagram illustrating examples of resource availability collected by the resource status collection unit 423 at the times t1, t2, t3, and t4. While the resource availability of cluster A is not substantially changed, the resource availability of cluster B is decreasing. The belonging cluster determination unit 424 calculates the slope of such a change in the resource status, and determines that the resource availability of the cluster B is lower than the resource availability of the cluster A at the next measurement time point. It is determined to be attributed. In addition, the belonging cluster determination unit 424 may determine the belonging cluster according to the type of memory, disk, and CPU in the computer resource of the own device and the type of computer resource that is insufficient for each cluster. good. For example, when the CPU of the own device has a CPU having a higher performance (higher frequency) than the CPUs of other node controllers, the cluster having the insufficient CPU is preferentially selected and assigned. It may be determined that this is a cluster.

  The attribution information change request unit 425 transmits to the cluster controller 300 an attribution information change request for requesting a change of the attribution destination of the own device. The attribute information change request includes, for example, an addition request and a deletion request including the node controller ID of the own device. When changing the cluster controller 300 to which the own device belongs, the belonging information change request unit 425 first transmits a deletion request for deleting its own belonging to the cluster controller 300 to which the own device belongs. On the other hand, the belonging information change request unit 425 transmits an additional request for requesting the own device to belong to the cluster controller 300 that has been determined by the belonging cluster determining unit 424 to belong to the own device. Here, the belonging information change request unit 425 stores the ID of the cluster controller 300 to which it belongs as shown in FIG. 11 in its own storage area.

  Here, the attribution information change request unit 425 transmits an attribution destination change request according to the determination result determined by the attribution cluster determination unit 424 based on the resource status information collected by the resource status collection unit 423 at regular intervals. Alternatively, the attribution change request may be transmitted when other predetermined conditions are satisfied. For example, referring to the resource status information of the own device stored in the node management unit 421, when a certain amount or more of free space is generated in the computer resource of the own device, an attribution change request is transmitted to change the attribution destination. You can change it.

Next, an operation example of the resource providing system 1 according to the present embodiment will be described. FIG. 12 is a sequence diagram showing an operation example in which the game node controller 420 in the resource providing system 1 changes its own destination.
Here, the game node controller 420-1 belongs to the cluster controller 300-1 (cluster A) in the initial stage. The resource status collection unit 340 of the cluster controller 300-1 receives the resource status information from the dedicated node controller 410-1 and (exclusive node controller 410-2) that are subordinate to the cluster controller 300-1. S1). On the other hand, the resource status collection unit 340 of the cluster controller 300-2 receives the resource status information from the dedicated node controller 410-3 (the dedicated node controller 410-4, the game node controller 420-2) that is under its control. (Step S2).

  The resource status collection unit 423 of the game node controller 420-1 receives the resource status information from the cluster controller 300-1 and the cluster controller 300-2 (steps S3 and S4). The belonging cluster determination unit 424 of the game node controller 420 compares the resource status information of the cluster controller 300-1 with the resource status information of the cluster controller 300-2 (step S5), and based on the comparison result, belongs to the own device. Is determined (step S6). When the belonging cluster determination unit 424 determines that the own device belongs to the cluster B, the belonging information change request unit 425 transmits a request for deleting the belonging destination to the cluster controller 300-1 (step S7).

  The cluster controller 300-1 refers to the instance storage unit 320 included in the cluster controller 300-1, and determines whether an instance exists on the game node controller 420-1. When the cluster controller 300-1 determines that an instance exists on the game node controller 420-1, the cluster controller 300-1 determines the destination of the instance operating on the computer resource of the game node controller 420-1. When the cluster controller 300-1 determines to move the instance operating on the computer resource of the game node controller 420-1 to the dedicated node controller 410-1, the cluster controller 300-1 An instance move request to the dedicated node controller 410-1 is transmitted to (step S8). In response to the instance movement request to the dedicated node controller 410-1 transmitted from the cluster controller 300-1, the node management unit 421 of the game node controller 420-1 exclusively allocates the instance operating on the computer resource of the own device. The node controller 410-1 is moved (step S9). The dedicated node controller 410-1 operates the instance moved from the game node controller 420-1 on the computer resource of the own device.

  The node controller information update unit 350 of the cluster controller 300-1 deletes the node controller information of the game node controller 420-1 stored in the node controller storage unit 310 (step S10). In addition, the cluster controller 300-1 deletes the instance information corresponding to the instance activated on the game node controller 420-1 from the instance storage unit 320. In addition, the amount of computer resources that has been reduced as a result of the game node controller 420-1 being deleted from its subordinates is subtracted from the resource status information stored in the resource status storage unit 330 and updated. The attribution information change request unit 425 of the game node controller 420-1 transmits an addition request to attribute itself to the cluster controller 300-2 (step S11). When the node controller information update unit 350 of the cluster controller 300-2 receives the addition request transmitted from the game node controller 420-1, the node controller information update unit 350 receives the node controller information of the game node controller 420-1 that is the transmission source of the received addition request. It is additionally stored in the node controller storage unit 310 (step S12). In addition, the cluster controller 300-2 adds the instance information activated on the game node controller 420-1 to the instance storage unit 320. In addition, the amount of computer resources increased by belonging to the game node controller 420-1 is increased and updated to the resource status information stored in the resource status storage unit 330. As a result, the game node controller 420-1 is stored as a node controller under the cluster controller 300-2, and the attribution destination is changed from cluster A to cluster B.

  Thereafter, when the cluster controller 300-2 receives a computer resource provision request from the user terminal 100, if there are not enough free resources of the dedicated node controller 410 under its own control, the game node controller 420-1 under its own control. The instance activation request is transmitted to (step S14). The game node controller 420-1 reads an image file corresponding to the computer resource provision request from the cloud controller 200, and activates the instance on its own computer resource.

  As described above, according to the resource providing system 1 of the present embodiment, the dedicated node controller 410 belonging to a specific cluster always secures a certain amount or more of computer resources according to the SLA and the like, while the game node controller 420 is Depending on the situation, it can be attributed to any of a plurality of clusters. As a result, it is possible to improve the utilization efficiency of computer resources and provide a stable resource providing service that immediately provides computer resources in response to a resource provision request from a user.

  For example, when the computer resources belonging to a specific cluster are insufficient for the provision request, in the conventional technology, the provider of the resource provision service needs to purchase a new physical machine and connect it to the specific cluster. . However, in such a case, it is conceivable that surplus computer resources are generated in other clusters. At this time, if the computer resource attribution can be changed from a cluster that has a surplus of computer resources to another cluster that lacks computer resources, the limited computer resources can be used more efficiently. It becomes. Therefore, according to the resource providing system 1 of the present embodiment, since the cluster to which the game node controller 420 belongs can be changed, it is possible to use limited computer resources more efficiently. Further, the node controller 400 provides the computer resources that are actually provided, and the node controller 400 itself determines the destination to which the node controller 400 belongs. For this reason, compared with a case where the attribution destination is changed by the cluster controller 300, for example, a large-scale system change is unnecessary, and the system maintainability of the entire private cloud can be improved.

  It is to be noted that a computer resource is provided by recording a program for realizing the function of the processing unit in the present invention on a computer-readable recording medium, causing the computer system to read and execute the program recorded on the recording medium. May be performed. Here, the “computer system” includes an OS and hardware such as peripheral devices. The “computer system” includes a WWW system having a homepage providing environment (or display environment). The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Further, the “computer-readable recording medium” refers to a volatile memory (RAM) in a computer system that becomes a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. In addition, those holding programs for a certain period of time are also included.

  The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

DESCRIPTION OF SYMBOLS 1 Resource provision system 100 User terminal 200 Cloud controller 300 Cluster controller 310 Node controller memory | storage part 320 Instance memory | storage part 330 Resource status memory | storage part 340 Resource status collection part 350 Node controller information update part 360 Node controller determination part 400 Node controller 401 Node management part 402 Virtualization unit 403 Instance 404 Computer resource 410 Exclusive node controller 411 Node management unit 412 Virtualization unit 420 Play node controller 421 Node management unit 422 Virtualization unit 423 Resource status collection unit 424 Attribution cluster determination unit 425 Attribution information change request unit

Claims (7)

A plurality of virtualization apparatuses that provide virtualized computer resources of the user apparatus to user terminals connected via a network; and a plurality of resource provision apparatuses connected to the plurality of virtualization apparatuses via the network A resource providing system comprising:
The resource providing device includes:
A subordinate virtualization apparatus information storage unit in which subordinate virtualization apparatus identification information for identifying a plurality of virtualization apparatuses associated as subordinates of the own apparatus among the plurality of virtualization apparatuses is stored;
A load status value storage unit that stores load status values indicating the load status of the plurality of virtualization devices identified by the subordinate virtualization device identification information;
A first load status value collection unit that collects a load status value of the virtualization device from the virtualization device identified by the subordinate virtualization device identification information and stores the load status value in the load status value storage unit;
When the virtualization device sends an additional request for causing the virtualization device to belong to the own device and includes the identification information of the virtualization device, the virtualization device subordinates the identification information. A subordinate virtualization apparatus update unit that stores the apparatus identification information in the subordinate virtualization apparatus information storage unit;
In response to a computer resource provision request transmitted from the user terminal, the virtual device identified by the subordinate virtualization apparatus identification information stored in the subordinate virtualization apparatus information storage unit is configured to provide a computer resource. A determination device determining unit,
The virtualization device is
A plurality of the resource providing device, a second load status value collecting unit for collecting the load status value stored before Kimake load status value storing unit of the resource providing device,
Based on a comparison result obtained by comparing the load status values of the plurality of resource providing devices collected by the second load status value collecting unit, any resource providing device of the plurality of resource providing devices Belonging resource providing device determining unit for determining whether to attach
The resource providing apparatus comprising: an additional request transmitting unit configured to transmit an additional request including identification information of the own device to the resource providing device determined to be attributed to the own device by the belonging resource providing device determining unit. system. To the plurality of virtualization devices connected via the network, a game virtualization device that changes the resource providing device to which the device belongs by transmitting an addition request to the resource providing device, and the addition request And a dedicated virtualization device that does not change the resource providing device to which the device belongs without transmitting
In the subordinate virtualization apparatus information storage unit of the resource providing apparatus, an exclusive subordinate virtualization apparatus identification that identifies the exclusive virtualization apparatus according to the amount of computer resources determined to be provided to the user terminal The resource providing system according to claim 1, wherein information and game-controlled virtual device identification information for identifying the game virtualization device that transmitted the addition request are stored. The virtualization apparatus determination unit of the resource providing apparatus is
Of the subordinate virtualization apparatus identification information stored in the subordinate virtualization apparatus information storage unit in response to the computer resource provision request transmitted from the user terminal, the subordinate virtual apparatus is more exclusive than the game subordinate virtualization apparatus identification information. The resource providing system according to claim 2, wherein the virtualization apparatus identification information is preferentially selected, and the virtualization apparatus identified by the selected subordinate virtualization apparatus identification information is provided with a computer resource. The additional request transmission unit of the virtualization apparatus includes:
The resource providing system according to any one of claims 1 to 3, wherein the addition request is transmitted when a predetermined condition is satisfied. Of the virtualization device,
The second load status value collection unit collects the load status values from the plurality of resource providing devices over a plurality of times,
The belonging resource providing device determination unit
Determining which of the plurality of resource providing devices the own device belongs to according to the transition status of the plurality of load status values collected by the second load status value collecting unit. The resource providing system according to any one of claims 1 to 4, characterized by: A user terminal connected via a network, a plurality of virtualization apparatus that provides computing resources of its own device virtualized, a plurality of the virtual associated as under the own apparatus among the plurality of virtualization apparatuses A subordinate virtualization apparatus information storage unit for storing subordinate virtualization apparatus identification information for identifying the subordinate virtualization apparatus, and a load status value indicating a load situation of the plurality of virtualization apparatuses identified by the subordinate virtualization apparatus identification information A resource providing method comprising a plurality of resource providing devices connected to the plurality of virtualization devices via the network, the resource providing device comprising a stored load status value storage unit There,
The resource providing device is
Collecting the load status value of the virtualization device from the virtualization device identified by the subordinate virtualization device identification information and storing it in the load status value storage unit;
The virtualization device is
A plurality of the resource providing device, a step of collecting the load status value stored before Kimake load status value storing unit of the resource providing device,
Based on the comparison result of comparing the load status values for each of the collected resource providing devices, a step of determining which resource providing device among the plurality of resource providing devices should belong to the device;
The resource providing device is determined to be assigned to its own device by the determination, and transmitting the addition request identification information of the own device is included,
The resource providing device is
When an addition request including identification information of the virtualization apparatus is transmitted from the virtualization apparatus, the identification information is stored in the subordinate virtualization apparatus information storage unit as the subordinate virtualization apparatus identification information;
A step of causing the virtualization apparatus identified by the subordinate virtualization apparatus identification information stored in the subordinate virtualization apparatus information storage unit to provide the computer resource in response to a computer resource provision request transmitted from the user terminal. When,
A resource providing method comprising: A plurality of virtualization devices that provide virtualized computer resources of a user's own device to user terminals connected via a network, and a plurality of resource providing devices connected to the plurality of virtualization devices via the network A subordinate virtualization apparatus information storage unit for storing subordinate virtualization apparatus identification information for identifying a plurality of virtualization apparatuses associated as subordinates of the own apparatus among the plurality of virtualization apparatuses; A load status value storage unit that stores load status values indicating the load status of the plurality of virtualization devices identified by the virtualization device identification information, and the virtualization device identified by the subordinate virtualization device identification information A first load status value collection unit that collects the load status value of the virtualization apparatus and stores the load status value in the load status value storage unit, and the virtualization apparatus identifies the virtualization apparatus. When an addition request including information is transmitted, the subordinate virtualization apparatus update unit that stores the identification information in the subordinate virtualization apparatus information storage unit as the subordinate virtualization apparatus identification information is transmitted from the user terminal. A virtualization apparatus determining unit that provides the virtualization apparatus identified by the subordinate virtualization apparatus identification information stored in the subordinate virtualization apparatus information storage unit in response to a computer resource provision request; The virtualization apparatus in a resource providing system comprising the resource providing apparatus comprising:
A plurality of the resource providing device, a second load status value collecting unit for collecting the load status value stored before Kimake load status value storing unit of the resource providing device,
Based on a comparison result obtained by comparing the load status values of the plurality of resource providing devices collected by the second load status value collecting unit, any resource providing device of the plurality of resource providing devices Belonging resource providing device determining unit for determining whether to attach
An additional request transmitting unit that transmits an additional request including identification information of the own device to the resource providing device determined to be attributed to the own device by the belonging resource providing device determining unit. apparatus.

Images