JP5218530B2 - Importance calculation method and apparatus for resources - Google Patents

Description

  The present invention relates to a technique for calculating importance for resources to be allocated or allocated to a business service.

  In a distributed computing environment in which resources are shared by a plurality of business services, it is required to optimize resource allocation for each business service. There are already known examples of methods for adjusting resource allocation based on statically defined business service priorities. Further, a method has been proposed in which the resource manager of a corresponding business service determines the importance of the resource allocated to each business service, and the overall resource allocation is switched based on the importance.

  For example, as shown in FIG. 1, the administrator of the business service 1 sets a BV calculation policy for the business service 1 for a BV (Business Value) calculation device of a local optimization mechanism for the business service 1. The BV calculation device allocates 20 to 60 BVs (business service importance) to the resource nodes for the business service 1 according to the set BV calculation policy. Specifically, two resource nodes with BV = 20, one resource node with BV = 50, and one resource node with BV = 60 are secured.

  Similarly, the administrator of the business service 2 sets the BV calculation policy of the business service 2 with respect to the BV (Business Value) calculation device of the local optimization mechanism for the business service 2. The BV calculation apparatus allocates 50 to 100 BVs (business service importance) to the resource nodes for the business service 2 according to the set BV calculation policy. Specifically, one resource node with BV = 60, one resource node with BV = 80, and one resource node with BV = 90 are secured.

  When the local optimization mechanism of the business service 1 and the local optimization mechanism of the business service 2 are actually assigned resource nodes, the BV is notified to the overall optimization mechanism. The overall optimization mechanism allocates resource nodes according to the settings of the operation manager. In the example of FIG. 1, resource nodes are allocated using the BV from the local optimization mechanism as it is. In the overall optimization mechanism, BV is given in the range of 0 to 100. Here, there are seven resource nodes, four resource nodes are assigned to the business service 1, and three resource nodes are assigned to the business service 2.

  Here, as shown in FIG. 2, when the business service 2 requests a resource node with BV = 50, since there is no space in the resource node, the resource node with the lowest BV (here, BV = assigned to the business service 1 = 20 resource nodes) are transferred to the business service related to the current request.

  In the prior art, the manager of each business service and the operation manager of the overall optimization mechanism make an arrangement separately so that an appropriate BV is output in each local optimization mechanism. That is, it is impossible to perform complete optimization using only the overall optimization mechanism.

  The system as described above is disclosed in the following documents.

Haruyasu Ueda, Nobutaka Imamura, Hiromichi Kobashi, Junichi Shigeta, Miho Murata, Akira Asato, Yoshimasa Kadooka, “Prototype Grid Service Platform for Business Resource Sharing”, Proceedings of Computational Engineering, Vol.11, No.2 , pp.531-534 (2006.Jul)

  In such conventional technology, not only the criteria and strategy for calculating the importance of business services, but also the range of importance used by the business services (the importance of the business services and the importance of other business services) Since the manager of the business service also sets the relative size relationship, there are the following problems. That is, (1) The relative importance of business services cannot be directly changed by the operation manager of the overall optimization mechanism, and the relative importance of each business service can be changed in order to respond to changes in the operational status. If you want to change it, you can't change it quickly. (2) When setting the BV calculation policy in the BV calculation device of the business service, the business service manager agrees in advance with the operation manager (and other business service managers) about the importance of the business service Need to take. That is, there is much labor involved in setting the BV calculation policy of the local optimization mechanism. (3) Because business service managers and management hosts are geographically distributed, it may be necessary or difficult to understand and manage the relative importance of business services as a whole. . (4) It is impossible to prevent an inappropriate importance from being used by setting an inappropriate BV calculation policy in the BV calculation device of the business service due to an error or intention of the business service administrator. That is, it cannot be prevented that the importance of the business service is relatively unreasonably higher or lower than the importance of other business services.

  Further, (5) since the BV calculation policy set by the business service administrator in the BV calculation device for the business service is inappropriate, the distribution of BV allocated to the resources allocated to the business service is biased. However, it cannot be corrected.

  Accordingly, an object of the present invention is to provide a technology that enables flexible and appropriate setting of the importance of a resource node to be allocated or allocated to a business service in a distributed computing environment in which a plurality of business services share resources. It is to be.

  Another object of the present invention is to provide a technique for appropriately allocating resource nodes even when the importance assigned to the resource nodes in business services is inappropriate. is there.

  The importance calculation method for a resource according to the present invention includes a step of receiving importance assigned to a specific business service or assigned to an assigned resource node based on a unique policy in the specific business service; In accordance with the data related to the importance recalculation policy that is set for the specific business service regardless of the above unique policy and is stored in the policy storage unit, the received importance is converted into the importance for the entire system and stored in the storage device. Storing.

  Since the importance of the entire system will be recalculated according to the importance recalculation policy, the importance recalculation policy will be appropriate and flexible by the system-wide operation manager so that the resource allocation of the entire system is optimized. Can be set. Here, optimization refers to maximizing effects and gains obtained from the entire business services when a plurality of business services are executed using a finite number of resource nodes existing in the system.

  Note that a step of switching the importance recalculation policy for a specific business service may be further included. For example, it may be necessary to apply different importance calculation policies between day and night. In this case, flexible processing is possible by performing such processing.

  Further, the importance recalculation policy may be at least one of a method for performing interval compression and interval shift, a method for performing mapping to a discontinuous predetermined value, and a method using a predetermined function value. is there.

  Further, the importance recalculation policy may include a rule regarding the range of importance after conversion. In this way, even if an inappropriate importance level is received, this step corrects it.

  Further, the step of determining whether the distribution of the received importance is predetermined and deviates from the reference stored in the distribution adjustment data storage unit; and the received importance distribution is predetermined and stored in the distribution adjustment data storage unit If it is determined that the stored standard is deviated, a correction step for correcting the received distribution of importance may be included.

  Further, the correction step described above includes a step of invalidating the importance recalculation policy and a step of correcting the received importance so as to be distributed according to the distribution stored in the distribution adjustment data storage unit. You may make it. This is because if the received importance level is biased, problems such as improperly occupying resources or indefinite service may occur.

  Furthermore, when the difference between the average value of the received importance and the reference stored in the distribution adjustment data storage unit is equal to or greater than a predetermined value, the correction step described above depends on the difference from the reference. Then, a step of correcting the received importance may be included. By performing such correction, the importance can be kept within an appropriate range.

  A program for causing a computer to execute the importance calculation method for resources according to the present invention can be created, and the program includes, for example, a flexible disk, an optical disk such as a CD-ROM, a magneto-optical disk, a semiconductor memory, and a hard disk. It is stored in a storage medium or a storage device. In some cases, digital signals are distributed over a network. Note that data being processed is temporarily stored in a storage device such as a computer memory.

  According to the present invention, in a distributed computing environment in which a plurality of business services share resources, it is possible to flexibly and appropriately set the importance for resource nodes to be allocated or allocated to the business services.

  In addition, according to another aspect of the present invention, even when the importance assigned to the resource node in the business service is inappropriate, the resource node is appropriately allocated.

It is a figure for demonstrating a prior art. It is a figure for demonstrating a prior art. It is a system outline figure in an embodiment of the invention. It is a figure which shows the relationship between a local optimization mechanism and a whole optimization mechanism. It is a functional block diagram of a whole optimization mechanism. It is a figure which shows a local BV distribution monitoring table. It is a figure which shows a distribution adjustment policy table. It is a figure which shows a BV recalculation policy table. It is a figure which shows a resource management table. It is a figure which shows a resource request management table. It is a figure which shows a distribution change management table. It is a sequence diagram at the time of switching of resource allocation. It is a sequence diagram at the time of switching of resource allocation. It is a sequence diagram at the time of update of local BV. It is a figure which shows the processing flow of the process which converts local BV into global BV. It is a figure which shows an example which calculates local BV from the state of a process. It is a figure which shows the other example which calculates local BV from the state of a process. It is a figure showing the relationship between a process state, local BV, and global BV in a section compression + section shift system. It is a figure showing the relationship between a process state, local BV, and global BV in a section compression + section shift system. It is a figure which shows an example of the mapping system to a discontinuous value. It is a figure which shows an example of a function definition system. It is a figure which shows the other example of a function definition system. (A) is a figure showing the system which converts the local BV in the daytime of a weekday into a global BV, (b) is a figure showing the system which converts the local BV in a night and a holiday into a global BV. It is a figure which shows an example of the resource allocation in embodiment of this invention. It is a figure which shows an example of the resource allocation in embodiment of this invention. It is a figure which shows an example of the resource allocation in embodiment of this invention. It is a figure which shows an example of the resource allocation in embodiment of this invention. It is a figure which shows an example of the resource allocation in embodiment of this invention. (A) thru | or (c) is a figure for demonstrating the inappropriate distribution of local BV. (A) And (b) is a figure for demonstrating correction | amendment of distribution of local BV. It is a figure which shows the processing flow of the correction process of local BV distribution. It is a functional block diagram of a computer.

  FIG. 3 shows the entire system according to the embodiment of the present invention. As shown in FIG. 3, a distributed computing environment is assumed in which a plurality of business services (business services 1 to N) share a plurality of resource nodes included in a plurality of sites. Such a distributed computing environment is divided into a service layer, a grid middleware layer, and a resource layer. The service layer includes business services used by a plurality of users, and the resource layer includes resource nodes. The grid middleware layer includes a mechanism for effectively allocating resource nodes to each business service. Specifically, the grid middleware layer includes a grid service subsystem including a local optimization mechanism corresponding to each business service, a resource brokering subsystem 100 that determines resource allocation for the entire system, and switching at each site. Execution mechanism. The resource brokering subsystem 100 includes an overall optimization mechanism 110, a physical resource management mechanism 120, and a switching command mechanism 130.

  For example, when acquiring a resource based on a request from a user in the business service 1, a BV (referred to as a local BV) for a resource to be acquired according to a predetermined policy in the local optimization mechanism 1 of the grid service subsystem 1 is used. The resource request including the local BV is calculated and output to the overall optimization mechanism 110 of the resource brokering subsystem 100. The overall optimization mechanism 110 converts a local BV to a global BV according to a preset importance recalculation policy, and determines resource allocation according to the global BV. If the resource has been allocated, the fact that the resource has been allocated is returned to the requesting local optimization mechanism 1. In addition, the overall optimization mechanism 110 outputs the resource allocation result to the physical resource management mechanism 120. When a resource switch or the like occurs, the physical resource management mechanism 120 specifies which resource node is physically assigned the business service and outputs an instruction to the switch command mechanism 130. The switching command mechanism 130 outputs an instruction to the requested switching execution mechanism to assign the requested business service to the specific physical resource node. The corresponding switching execution mechanism causes the resource node to switch processing to be executed in accordance with a command from the switching command mechanism 130.

  Since the present embodiment mainly relates to the local optimization mechanism of the grid service subsystem corresponding to each business service and the overall optimization mechanism 110 of the resource brokering subsystem 100, the other parts are more than this. Shall not be mentioned.

  Next, the configuration of the local optimization mechanism will be mainly described with reference to FIG. The local optimization mechanism of each business service receives a policy setting unit 202 that receives and stores an input of a local BV calculation policy by the business service administrator, and the business service based on the local BV calculation policy set in the policy setting unit 202. A BV calculation unit 203 that calculates a local BV for a resource node to be allocated or allocated to a service, and a resource node used in its own business service are managed, and the local BV calculated by the BV calculation unit 203 is And a resource management unit 201 that transmits the request to the overall optimization mechanism 110 as a local BV notification or as a resource request (may be a resource return).

  In the policy setting unit 202 of the local optimization mechanism, a unique local BV calculation policy is set according to the business service, and the BV calculation unit 203 calculates a local BV according to the unique local BV calculation policy. Conventionally, a calculation policy (distribution) of a local BV in the local optimization mechanism is determined and set in consultation with the operation manager of the overall optimization mechanism 110 and the manager of each business service. However, in this embodiment, the overall optimization mechanism 110 adjusts the entire system in accordance with the setting of the operation manager, so that the local optimization mechanism also calculates the local BV according to its own local BV calculation policy. For example, local BVs from 1 to 100 are calculated. However, as described below, when the local BV is calculated with a distribution that is too biased, the global optimization mechanism 110 also performs a forced correction.

  Next, a specific configuration of the overall optimization mechanism 110 will be described with reference to FIG. 5A. The overall optimization mechanism 110 includes a local BV distribution monitoring / adjustment unit 111 having a local BV distribution monitoring table 1111 (FIG. 5B), a distribution adjustment policy setting unit 112 having a distribution adjustment policy table 1121 (FIG. 5C), and a BV Recalculation unit 113, BV recalculation policy setting unit 114 having BV recalculation policy table 1141 (FIG. 5D), resource management unit 115 having resource management table 1151 (FIG. 5E), and resource request / return receiving unit 116, and a distribution determination unit 117 having a resource request management table 1171 (FIG. 5F) and a distribution change management table 1172 (FIG. 5G).

  The local BV distribution monitoring / adjusting unit 111 receives a local BV update notification (including service ID, local BV, and resource ID) from the local optimization mechanism of each business service and performs processing. At the time of processing, the distribution adjustment policy table 1121 is referred to. In addition, the BV recalculation policy setting unit 114 is set as necessary. Furthermore, when the local BV is adjusted, the adjustment result is output to the BV recalculation unit 113.

  The local BV distribution monitoring table 1111 stores the reception count, count start time, average, variance, average threshold flag, variance threshold flag, and status flag for each service ID. Since the count start time is used to calculate the time from the count start, the elapsed time itself may be maintained while being updated independently.

  As shown in FIG. 5A, the distribution adjustment policy setting unit 112 receives a setting file from the operation manager of the overall optimization mechanism 110 or displays a setting screen (GUI: Graphical User Interface) on the display device for input. Stored in the distribution adjustment policy table 1121. As shown in FIG. 5C, the distribution adjustment policy table 1121 includes, for each service ID, a distribution monitoring flag indicating whether distribution monitoring and adjustment are performed, a grace period (time / number of times), a preferable distribution function, and a preferable distribution function. Parameters (average, variance), average difference threshold, and variance difference threshold are stored.

  The BV recalculation policy setting unit 114 receives a setting file from the operation manager of the overall optimization mechanism 110 or displays a setting screen on the display device for input and stores it in the BV recalculation policy table 1141. . As shown in FIG. 5D, the BV recalculation policy table 1141 includes, for each service ID, a recalculation flag indicating whether recalculation is performed, a parameter value (for example, a section compression rate and a section shift value, or a graph / Function) and BV range (lower limit value and upper limit value).

  The BV recalculation unit 113 receives a local BV update notification from the local optimization mechanism of each business service, and converts the local BV into a global BV based on the data stored in the BV recalculation policy table 1141. carry out. Note that processing may be performed on the local BV included in the resource request. When the local BV update notification is received, the calculated global BV is output to the resource management unit 115.

  The resource management unit 115 updates the resource management table 1151 in response to an instruction from the BV recalculation unit 113 or the allocation determination unit 117. As shown in FIG. 5E, the resource management table 1151 registers a global BV and a distribution destination service ID for each resource ID.

  When the resource request / return receiving unit 116 receives a request for a resource including a local BV from the local optimization mechanism of each business service, the resource request / return receiving unit 116 requests the BV recalculation unit 113 to calculate the global BV, The request for the resource to be included is output to the allocation determination unit 117. When a resource return request including the resource ID is received from the local optimization mechanism of each business service, the request is output to the distribution determination unit 117.

  The distribution determining unit 117 receives a request for a resource including the global BV via the resource request / return receiving unit 116, registers it in the resource request management table 1171, and refers to the resource management table 1151 managed by the resource management unit 115. To determine whether or not to change the resource allocation. When switching the resource allocation, such as when the global BV related to the resource request is larger than the lowest global BV of the currently allocated resource, the allocation change management table 1172 is referred to and the resource is not being switched. After confirming this, data related to switching is registered in the distribution change management table 1172, and when switching of resource distribution is completed, the resource management unit 115 is instructed to change the resource management table 1151. If switching is performed, the corresponding record registered in the resource request management table 1171 and the distribution change management table 1172 is invalidated. On the other hand, when the return of the resource including the resource ID is received via the resource request / return receiving unit 116, the allocation determining unit 117 registers the release of the resource related to the return in the resource management table 1151 in the resource management unit 115. Let Further, the resource management unit 115 is caused to register in the asset management table 1151 the allocation of vacant resources for the business service having the largest global BV value in the resource request management table 1171.

  As shown in FIG. 5F, the resource request management table 1171 stores a request source service ID, the number of requested resources, and a global BV at the time of request. Further, as shown in FIG. 5G, the distribution change management table 1172 stores a resource ID, a global BV, a picking source service ID, and an allocation destination service ID.

  In the present embodiment, the resource management unit 115, the resource request / return receiving unit 116, and the distribution determining unit 117 are not main units, and the specific operation will be described with reference to FIGS. 6A and 6B. Keep it.

  FIG. 6A illustrates a case where resources allocated to the business service 2 are transferred in response to a resource request from the business service 1. First, the resource management unit 201 of the local optimization mechanism of the business service 1 outputs a BV calculation request to the BV calculation unit 203 before transmitting a resource request to the overall optimization mechanism 110 (step (1)). The BV calculation unit 203 receives a BV calculation request from the resource management unit 201 and calculates a local BV according to a policy set in the policy setting unit 202 (for example, a function of a predefined number of sessions and a local BV). (Step (2)) and output to the resource management unit 201 (Step (3)). When receiving the local BV, the resource management unit 201 transmits a resource request including the local BV to the resource request / return receiving unit 116 of the overall optimization mechanism 110 (step (4)).

  The resource request / return receiving unit 116 of the overall optimization mechanism 110 receives a resource request including a local BV from the resource management unit 201 of the local optimization mechanism of the business service 1, and the BV recalculation unit 113 receives a global BV for the local BV. The calculation request is output (step (5)). The BV recalculation unit 113 refers to the BV recalculation policy table 1141 and calculates a global BV for the received local BV (step (6)). This process will be described in detail below. The BV recalculation unit 113 outputs the global BV to the resource request / return receiving unit 116 (step (7)), and the resource request / return receiving unit 116 outputs the resource request including the global BV to the allocation determining unit 117. . The allocation determining unit 117 registers data related to the resource request (service ID of the requesting business service (in this case, the business service 1), the number of requested resources, and the global BV (when requested)) in the resource request management table 1171. At the same time, the resource management table of the resource management unit 115 is referred to and it is determined whether or not to allocate resources to the resource request. Here, the resource assigned to the business service 2 is taken up and determined to be assigned to the request source business service of this resource request (step (8)). In this case, in the distribution change management table 1172, the resource ID of the resource to which the business service is switched, the global BV included in the resource request, the picking-up service ID that is the service ID of the business service picking up the resource (in this case, the business service 2) An allocation destination service ID (in this case, business service 1), which is a service ID of the resource request source business service, is registered.

  In addition, the allocation determination unit 117 transmits a resource release request including the resource ID of the picked up resource to the resource management unit 201 in the local optimization mechanism of the business service (business service 2 in this case) specified by the picking source service ID. (Step (9)). The resource management unit 201 in the local optimization mechanism of the business service 2 receives the resource release request including the resource ID of the resource picked up from the overall optimization mechanism 110, and registers the resource release of the resource (step (10)). . Further, the resource management unit 201 outputs a BV calculation request to the BV calculation unit 203 (step (11)). Upon receiving the BV calculation request from the resource management unit 201, the BV calculation unit 203 performs BV calculation (step (12)). Then, the local BV of the calculation result is output to the resource management unit 201 (step (13)). The resource management unit 201 receives the local BV from the BV calculation unit 203, and transmits a local BV update notification including the local BV, the resource ID, and the service ID to the BV recalculation unit 113 of the overall optimization mechanism 110 (step) (14)).

  The BV recalculation unit 113 of the overall optimization mechanism 110 receives the local BV update notification from the local optimization mechanism of the business service 2 and performs the calculation of the global BV as described below (step (15)). Then, the BV recalculation unit 113 outputs the calculated global BV together with the service ID and resource ID to the resource management unit 115 (step (16)). The resource management unit 115 updates the global BV of the corresponding record in the resource management table 1151 (step (17)).

  Further, the resource management unit 201 in the local optimization mechanism of the business service 2 transmits a resource return notification including the resource ID to the distribution determination unit 117 via the resource request / return reception unit 116 in the overall optimization mechanism 110 ( Step (18)). Then, the allocation determination unit 117 causes the resource management unit 115 to register the allocation destination service ID (in this case, the business service 1) and the requested BV corresponding to the resource ID related to the switching (FIG. 6B: Step ( 19)). Further, the corresponding record is invalidated in the resource request management table 1171 and the distribution change management table 1172. Then, the distribution determination unit 117 transmits a resource allocation notification to the resource management unit 201 of the business service (in this case, the business service 1) related to the service ID of the allocation destination (step (20)).

  The resource management unit 201 in the local optimization mechanism of the business service 1 receives the resource allocation notification from the overall optimization mechanism 110, updates its own resource management table, and outputs a BV calculation request to the BV calculation unit 203 ( Step (21)). Upon receiving the BV calculation request from the resource management unit 201, the BV calculation unit 203 performs BV calculation (step (22)). Then, the local BV of the calculation result is output to the resource management unit 201 (step (23)). The resource management unit 201 receives the local BV from the BV calculation unit 203, and transmits a local BV update notification including the local BV, the resource ID, and the service ID to the BV recalculation unit 113 of the overall optimization mechanism 110 (step) (24)).

  The BV recalculation unit 113 of the overall optimization mechanism 110 receives the local BV update notification from the local optimization mechanism of the business service 1 and performs the calculation of the global BV as described below (step (25)). Then, the BV recalculation unit 113 outputs the calculated global BV together with the service ID and resource ID to the resource management unit 115 (step (26)). The resource management unit 115 updates the global BV of the corresponding record in the resource management table 1151 (step (27)). By performing such processing, the resource is switched.

  6A and 6B can be modified in various ways as long as the same function can be realized. For example, steps (11) to (13) can be omitted. That is, without recalculating the local BV of the resource to be released, the resource entry in step (10) may be simply deleted from the resource management table of the local optimization mechanism. In this case, step (14) is changed to a step of notifying that the local BV has been updated to a predetermined value that means “release of resources”. Further, step (14) may be absorbed into step (18). That is, when a resource return notification (corresponding to step (18)) is received, the global BV of the corresponding resource is recalculated (or a predetermined global BV to be given to a resource not assigned to any service is output. In this case, steps (15) and (16) are also omitted.) When updating the table of the resource management unit 115 of the overall optimization mechanism, a local BV update notification (step ( 14)) may be omitted and only the resource return notification (step (18)) may be transmitted.

  In the processing described above, the scene where the global BV is calculated from the local BV is shown, but the global BV is also calculated in the scene as shown in FIG.

  The service master at the business service master node that aggregates the notifications from the monitor agents of individual resource nodes and the monitor agents performs status notifications (including the number of sessions and elapsed execution time, for example) periodically or when the status changes. The data is transmitted to the resource management unit 201 in the service local optimization mechanism (step (31)). The resource management unit 201 receives the status notification from the service master in the master node of the business service and the monitor agent of the allocated resource node, and outputs a local BV calculation request including the status notification to the BV calculation unit 203 (step (32) )). The BV calculation unit 203 receives the local BV calculation request including the status notification, and performs the calculation so as to update the local BV (step (33)). The BV calculation unit 203 notifies the resource management unit 201 of the recalculated local BV (step (34)). The resource management unit 201 updates its own management information (step (35)) and sends a local BV update notification including the resource ID and service ID related to the local BV to the BV recalculation unit of the overall optimization mechanism 110. It transmits to 113 (step (36)).

  When receiving the local BV update notification including the local BV, the resource ID, and the service ID from the resource management unit 201 in the business service local optimization mechanism, the BV recalculation unit 113 of the overall optimization mechanism 110 receives the following from the local BV. The global BV calculation described in detail in (4) is performed (step (37)). Then, the BV recalculation unit 113 notifies the resource management unit 115 of the global BV, the resource ID, and the service ID (step (38)). The resource management unit 115 receives the global BV, resource ID, and service ID from the BV recalculation unit 113, and updates the corresponding record in the resource management table 1151 (step (39)).

  Such processing is repeated and the global BV is updated.

  Next, calculation of global BV by the BV recalculation unit 113, which is omitted in FIGS. 6A and 6B and FIG. 7, will be described in detail with reference to FIGS. First, the BV recalculation unit 113 receives a local BV update notification including a local BV, a resource ID, and a service ID from one of the business service local optimization mechanisms, or receives a local BV update notification from the resource request / return receiving unit 116. A global BV calculation request including the BV and service ID is received and stored in a storage device such as a main memory (step S1).

  Thereafter, with reference to the BV recalculation policy table 1141 of the BV recalculation policy setting unit 114, it is determined whether a BV recalculation policy for the business service is set (step S3). That is, it is determined whether a record for the received service ID exists. Even if a record exists, the recalculation flag may be turned off. In this case, as will be described in detail below, the input from the local BV distribution monitoring / adjusting unit 111 is followed. If a record for the received service ID does not exist, an error process determined in advance because of an error is performed (step S19). Then, the process returns to step S1.

  On the other hand, if there is a record for the received service ID, the BV recalculation unit 113 determines whether the “section compression + section shift” method is designated as the BV recalculation policy (step S5). This is determined using data such as a parameter value (or graph / function) in the BV recalculation policy table 1141. When the “section compression + section shift” method is adopted, for example, global BV = local BV × (section compression ratio) + section shift value is calculated and stored in a storage device such as a main memory. (Step S7). However, the present invention is not limited to such linear compression and interval shift. Then, the process proceeds to step S17.

  An example of the “section compression + section shift” method is shown in FIGS. 9A and 9B. FIG. 9A shows calculations performed in the BV calculation unit 203 of the local optimization mechanism. The horizontal axis represents the elapsed execution time, and the vertical axis represents the local BV. For example, if the elapsed execution time is 0, the local BV is 1; if the elapsed execution time is less than 30, the local BV is 5, and if the elapsed execution time is 30 or more and less than 60, the local BV is If the elapsed execution time is 60 or more and less than 90, the local BV is 40. If the elapsed execution time is 90 or more and less than 120, the local BV is 60 and the elapsed execution time is 120 or more and less than 150. If so, the local BV is 80, and if the elapsed execution time is 150 or more, the local BV is 100.

  On the other hand, in FIG. 9B, the first axis represents the elapsed execution time, the second axis represents the local BV, and the third axis represents the global BV. The plane stretched by the first axis and the second axis represents the calculation (FIG. 9A) performed in the BV calculation unit 203 of the local optimization mechanism. On the other hand, the BV recalculation unit 113 of the overall optimization mechanism 110 has local BV values of 5, 20, 40, 60, like a straight line in a plane stretched between the second axis and the third axis. 80 and 100 are converted to a straight line passing through the origin with a slope of 1 / 2.5. That is, in this example, a local BV section having a width of about 100 is compressed into a global BV section with a section compression ratio of 1 / 2.5, and the section shift is set to zero. Note that straight lines z1 to z6 represented by bold lines represent the relationship between the global BV, the local BV, and the elapsed execution time in the three-dimensional space.

  An example of the “section compression + section shift” method is shown in FIGS. 10A and 10B. FIG. 10A shows the calculation performed in the BV calculation unit 203 of the local optimization mechanism. The horizontal axis represents the number of sessions, and the vertical axis represents the local BV. That is, the local BV is output with a width of 50 to 100 according to the number of sessions according to the curve a. On the other hand, in FIG. 10B, the first axis represents the number of sessions, the second axis represents the local BV, and the third axis represents the global BV. A plane stretched between the first axis and the second axis represents a calculation (FIG. 10A) performed in the BV calculation unit 203 of the local optimization mechanism. On the other hand, the BV recalculation unit 113 of the overall optimization mechanism 110 converts the value of the local BV to the value of the global BV 40 to 60 as in the straight line b in the plane stretched between the second axis and the third axis. The section is compressed to the section and the section is shifted. That is, the section compression rate is 1/5 and the section shift is 40. A curve b 'is a curve in a three-dimensional space that directly indicates the relationship between the number of sessions, the local BV, and the global BV. In the case of the straight line c, the local BV value is section-compressed into the section of the global BV value 50 to 100 and shifted. That is, the section compression rate is 1/2 and the section shift is 50. A straight line c 'is a curve in a three-dimensional space that directly represents the relationship between the number of sessions, the local BV, and the global BV.

  On the other hand, if it is determined in step S5 that the “interval compression + interval shift” method is not used, it is determined whether the “mapping to discontinuous values” method is instructed (step S9). This is determined using data such as a parameter value (or graph / function) in the BV recalculation policy table 1141. When the “mapping to discontinuous values” method is instructed, the mapping is performed to the global BV corresponding to the local BV section including the notified value (step S11). Then, the process proceeds to step S17.

  An example of the “mapping to discontinuous values” method will be described with reference to FIG. In FIG. 11, the first axis represents the number of sessions, the second axis represents the local BV, and the third axis represents the global BV. A plane stretched between the first axis and the second axis represents a calculation performed in the BV calculation unit 113 of the local optimization mechanism, and a curve d substantially the same as that in FIG. 10A is used. On the other hand, the plane stretched between the second axis and the third axis represents the calculation performed by the BV recalculation unit 113 of the overall optimization mechanism 110. Specifically, the straight line e1 in which the global BV is 40 in the section 1 (the section where the local BV is about 50 to about 75) and the global BV in the section 2 (the section where the local BV is about 75 to about 85) are 45 straight line e2 and global BV in section 3 (section where local BV is about 85 to about 95) and global BV in section 4 (section where local BV is about 95 to less than 100) Is represented by a straight line e4 with 75 and a point e5 with a global BV of 80 in the section 5 (local BV is 100). Curves e1 'to e5' are curves representing the relationship between the local BV, the global BV, and the number of sessions in the three-dimensional space.

  On the other hand, if it is determined in step S9 that it is not the “mapping method to discontinuous values”, it is determined whether the “function definition” method is designated (step S13). This is determined using data such as a parameter value (or graph / function) in the BV recalculation policy table 1141. If the “function definition” method is instructed, the global BV is calculated by global BV = f (local BV) when f (x) is a designated function (step S15). Then, the process proceeds to step S17.

  An example of the “function definition” method will be described with reference to FIG. In FIG. 12, the first axis represents the number of sessions, the second axis represents the local BV, and the third axis represents the global BV. A plane stretched between the first axis and the second axis represents a calculation performed in the BV calculation unit 203 of the local optimization mechanism, and a curve f substantially the same as that in FIG. 10A is used. On the other hand, the plane stretched between the second axis and the third axis represents the calculation performed by the BV recalculation unit 113 of the overall optimization mechanism 110. Specifically, the curve g corresponds to the function f (x). In the three-dimensional space, a curve g ′ represents the relationship between the local BV, the global BV, and the number of sessions.

  Another example of the “function definition” method will be described with reference to FIG. In FIG. 13, the first axis represents the elapsed execution time, the second axis represents the local BV, and the third axis represents the global BV. A plane stretched between the first axis and the second axis represents a calculation performed in the BV calculation unit 203 of the local optimization mechanism, and shows mapping to almost the same discontinuous values as in FIG. 9A. On the other hand, the plane stretched by the second axis and the third axis indicates the calculation performed by the BV recalculation unit 113 of the overall optimization mechanism 110. Specifically, a polygonal curve k corresponds to the function f (x). In the three-dimensional space, the straight line groups j1 to j6 represent the relationship between the local BV, the global BV, and the elapsed execution time.

  If the “function definition” method is not specified in step S13, the process proceeds to step S19 in the present embodiment. That is, error processing is executed. After the error processing, the process returns to step S1.

  On the other hand, after executing Step S7, S11 or S15, the global BV is output to the resource management unit 115 to update the resource management table 1151 (Step S17). If the notification is from the resource request / return receiving unit 116, the notification is output to the resource request / return receiving unit 116. Then, the process returns to step S1.

  By executing such processing, the global BV can be selected by any one of “interval compression + interval shift” method, “mapping to discontinuous values” method, and “function definition” method according to the business service of the notification source. Can be calculated. That is, independent of the calculation of the local BV in the BV calculation unit 203 of the local optimization mechanism, the global BV can be calculated and resources can be switched after considering the entire system.

  Note that the global BV calculation method is not always constant and can be dynamically changed. In this case, the BV recalculation policy setting unit 114 may register two or more records in the BV recalculation policy table 1141 and set the recalculation flag on for valid records.

  An example of dynamically changing the calculation method will be described with reference to FIGS. 14 (a) and 14 (b). 14A and 14B, the horizontal axis represents the local BV, and the vertical axis represents the global BV. For example, during the daytime on weekdays, as shown in FIG. 14A, the first business service such as a videophone is converted as represented by the straight line m, and the global BV is distributed in the range M. On the other hand, for the second business service such as a batch job, the conversion represented by the straight line n is performed, and the global BV is distributed in the range N. Since the first business service has a relatively large global BV value, resources are preferentially allocated. On the other hand, for night and holidays, as shown in FIG. 14B, the first business service is converted as represented by the straight line m ′, and the global BV is distributed in the range M ′. The global BV is made smaller than the straight line m, and the distribution range is also narrowed. On the other hand, for the second business service, the conversion represented by the straight line n ′ is performed, and the global BV is distributed in the range N ′. The global BV is made larger than the straight line n, and the distribution range is expanded. For some resources, a larger global BV is assigned to the second business service than to the first business service.

  In this way, the first business service that consumes a lot of resources is given priority during the day, and conversely, the time when batch processing is more frequently used than business services such as video conferencing, such as weekday nights and holidays. By changing the value of global BV, it becomes possible to utilize system resources more flexibly and effectively.

  Note that switching of the BV recalculation policy is not limited to two stages, and switching may be performed in more stages. Further, switching may be performed not only according to the time-related standard but also according to another standard. For example, it may be weather or network congestion.

  In summary, resource allocation as shown in FIGS. 15 to 19 is performed. The local optimization mechanisms 1 to 3 output local BV from 1 to 100. In the initial state, as shown in the upper part of FIG. 15, the local optimization mechanism 1 of the business service 1 has two resources with 10 local BVs, 3 with 25 resources, 1 with 30 resources, and 75. 2 resources and 80 resources are requested. Further, the local optimization mechanism 2 of the business service 2 requests four resources with 50 local BVs. The local optimization mechanism 3 of the business service 3 requests 30 resources, 1 resource of 40, 3 resources of 60, 1 resource of 80, and 2 resources of 100. ing. At this time, the BV recalculation unit 113 calculates a global BV in accordance with the BV recalculation policy of the business service 1 stored in the BV recalculation policy table 1141 (section compression ratio 1 and section shift value 0). In addition, the BV recalculation unit 113 calculates a global BV according to the BV recalculation policy for the business service 2 stored in the BV recalculation policy table 1141 (the section shift value is 1.8 and the section shift value is 0). Furthermore, the BV recalculation unit 113 stores the BV recalculation policy for the business service 3 stored in the BV recalculation policy table 1141 (if the local BV is 30 or less, the interval compression value is 1/3 and the interval shift value is 0, If it exceeds 30, the global BV is calculated according to the interval compression value 1/2 and the interval shift value 0). Then, as shown in the lower part of FIG. 15, there are three resources with 10 global BVs, one with 20 resources, three with 25 resources, four with 30 resources, one with 40 resources, 50 2 resources, 75 resources 2, 80 resources 1, and 90 resources 4. Note that the global BV is 0, that is, two unused resources remain.

  Next, in the state of FIG. 15, considering that the local optimization mechanism 1 of the business service 1 requests two resources whose local BV is 5, the global BV becomes 5, and as shown in FIG. Two resources with a global BV of 5 are reserved, and there are no unused resources. In FIG. 16, a resource that is drawn with a dotted line when the local BV is 5 in the local optimization mechanism 1 is assigned, and a business service is switched for a resource that is drawn with a dotted line when the global BV is 0 in the overall optimization mechanism 110. It is done.

  Further, in the state of FIG. 16, when the local optimization mechanism 3 of the business service 3 requests two resources with 30 local BVs, the global BV becomes 10. Then, as shown in FIG. 17, two resources assigned to the business service 1 with a global BV of 5 are switched to 10 resources with a global BV of the business service 3. In FIG. 17, the allocation is made for the resource drawn by the dotted line with the local BV 30 in the local optimization mechanism 3, and the allocation is canceled for the resource drawn by the dotted line with the local BV 5 in the local optimization mechanism 1. Thus, the business service is switched for the resource drawn with a dotted line when the global BV is 5 in the overall optimization mechanism 110.

  Thereafter, in the state of FIG. 17, when the local optimization mechanism 2 of the business service 2 requests two resources with 50 local BVs, the global BV becomes 90. For example, in the case of the same global BV value, assuming that the setting is made so that the business service 1 is preferentially assigned resources, the global BV assigned to the business service 3 is 10 as shown in FIG. These two resources are switched as 90 resources of the global BV of the business service 2. In FIG. 18, the allocation is made for the resource drawn by the dotted line with the local BV of 50 in the local optimization mechanism 2, and the allocation is canceled for the resource drawn by the dotted line with the local BV of 30 in the local optimization mechanism 3. As a result, the business service is switched for the resource indicated by the dotted line when the global BV is 10 in the overall optimization mechanism 110.

  Furthermore, in the state of FIG. 18, when the local optimization mechanism 1 of the business service 1 requests one resource whose local BV is 60, the global BV is 60. For example, in the case of the same global BV value, if the setting is made such that the business service 1 is preferentially assigned resources, the global BV assigned to the business service 3 is 10 as shown in FIG. These resources are switched as 60 resources of the global BV of the business service 1. In FIG. 19, the allocation is made for the resource drawn by the dotted line with the local BV of 60 in the local optimization mechanism 1, and the allocation is canceled for the resource drawn by the dotted line with the local BV of 30 in the local optimization mechanism 3. As a result, the business service is switched for the resource indicated by the dotted line when the global BV is 10 in the overall optimization mechanism 110.

  Next, processing of the local BV distribution monitoring / adjusting unit 111 will be described. For example, let us consider a case where the business service administrator sets a policy as shown in FIG. That is, the local BV is set to 90 until the elapsed execution time is 180, and the local BV is set to 100 if the elapsed execution time is 180 or more. In such a case, even if the global BV is calculated according to the policy in the BV recalculation policy table 1141 of the overall optimization mechanism 110 (for example, the section shift value is 1/5 and the section shift value is 30), it is shown in FIG. Thus, only two values (eg 48 and 50) will occur. Naturally, as shown in FIG. 20C, the distribution of global BV values is biased. As described above, if the local optimization mechanism is set so that the global BV is calculated too deliberately or carelessly, resources may be preferentially allocated to the business service. Conversely, resources may not be allocated at all. Accordingly, it is preferable that the global BV values are dispersed as much as possible within a predetermined range in the overall optimization mechanism 110. For example, as shown in FIG. 21A, it is preferable that the global BV appears in the form of a normal distribution from the lower limit 30 to the upper limit 50. In this case, the relationship between the local BV and the global BV is not a point as shown in FIG. 20B but a straight line (or a curve) as shown in FIG.

  In order to perform such distribution adjustment, the local BV distribution monitoring / adjusting unit 111 performs the following processing. First, the local BV distribution monitoring / adjusting unit 111 receives a local BV update notification including a local BV, a service ID, and a resource ID from the business service local optimization mechanism, and stores it in a storage device such as a main memory. (FIG. 22: Step S21). Then, it is determined whether or not the distribution monitoring flag for the business service is turned on with reference to the distribution adjustment policy table 1121 (step S23). The distribution adjustment policy table 1121 is searched with the service ID to identify whether the distribution monitoring flag is on or off. If the distribution monitoring flag for the business service is off, the process returns to step S21.

  On the other hand, if the distribution monitoring flag for the business service is ON, it is determined whether a local BV monitoring / adjustment policy for the business service is set (step S25). The local BV monitoring / adjustment policy is specified by the distribution function, distribution function parameters (average and variance), average difference threshold, and variance difference threshold in the distribution adjustment policy table 1121. If the local BV monitoring / adjustment policy for the business service is not set, predetermined error processing is performed (step S27). Then, the process returns to step S21.

  On the other hand, when the local BV monitoring / adjustment policy for the business service is set, statistical processing for the business service, that is, calculation of average and variance is performed and registered in the local BV distribution monitoring table 1111 ( Step S29). Furthermore, it is determined whether the period for grace for correction of the local BV of the business service has elapsed from the grace period (time / count) in the distribution adjustment policy table 1121 and the reception count or count start time in the local BV distribution monitoring table 1111. (Step S31). If the grace period is defined by the number of times, it is determined whether the number of receptions in the local BV distribution monitoring table 1111 exceeds that number. If the grace period is defined by time, it is determined whether the grace period has been exceeded based on the difference between the count start time and the current time in the local BV distribution monitoring table 1111.

  If it is determined that the period for delaying correction of the local BV of the business service has not elapsed, the process returns to step S21. On the other hand, when it is determined that the period for grace correction of the local BV of the business service has elapsed, the average in the local BV distribution monitoring table 1111 and the monitoring / regulation specified in the distribution adjustment policy table 1121 It is determined whether the absolute value of the difference from the parameter (average) of the distribution function in the adjustment policy exceeds the average difference threshold in the distribution adjustment policy table 1121 (step S33). If it is determined that the absolute value of the difference between the average in the local BV distribution monitoring table 1111 and the average set in the monitoring / adjustment policy does not exceed the threshold of the average difference, the process proceeds to step S37. When the threshold value includes the upper limit value and the lower limit value, it is determined whether or not the threshold value deviates from the range specified by the upper limit value and the lower limit value, not the absolute value.

  On the other hand, if it is determined that the absolute value of the difference between the average in the local BV distribution monitoring table 1111 and the average set in the monitoring / adjustment policy exceeds the threshold of the average difference, the local BV distribution A parameter value (for example, section shift) for the corresponding business service in the BV recalculation policy table 1141 of the BV recalculation policy setting unit 114 by the difference between the average in the monitoring table 1111 and the average set in the monitoring / adjustment policy Value) is corrected (step S35). {Section shift value− (Difference between average in local BV distribution monitoring table 1111 and average set in monitoring / adjustment policy)}. Other types may be corrected in the same manner.

  The absolute value of the difference between the variance in the local BV distribution monitoring table 1111 and the parameter (variance) of the distribution function in the monitoring / adjustment policy defined in the distribution adjustment policy table 1121 after the No route in step S33 or after step S35. Is over the threshold value of the variance difference in the distribution adjustment policy table 1121 (step S37). If it is determined that the absolute value of the difference between the variance in the local BV distribution monitoring table 1111 and the variance set in the monitoring / adjustment policy does not exceed the variance difference threshold, the process proceeds to step S45. When the threshold value includes the upper limit value and the lower limit value, it is determined whether or not the threshold value deviates from the range specified by the upper limit value and the lower limit value, not the absolute value.

On the other hand, if it is determined that the absolute value of the difference between the variance in the local BV distribution monitoring table 1111 and the variance set in the monitoring / adjustment policy exceeds the threshold of the variance difference, the local BV distribution In the monitoring table 1111, the “status” for the corresponding business service is changed to “adjusting in progress” (step S 39). Further, the BV recalculation policy setting unit 114 is set to turn off the recalculation flag of the corresponding business service in the BV recalculation policy table 1141 (step S41). Then, a local BV of the distribution (distribution function and distribution function parameter) set in the distribution adjustment policy table 1121 is generated and output to the BV recalculation unit 113 (step S43). For example, a random number within the range (lower limit value and upper limit value) of BV defined in the BV recalculation policy table is generated. For example, if a normal distribution of N (μ, σ ² ) is designated, a normal random number is generated by the box = Muller method or the like, and a random number within the range of BV is generated. Another distribution may be designated.

  Note that the section compression rate in the parameter value may be set to 1 without setting the recalculation flag to OFF as in step S41.

  Then, the process returns to step S21.

  On the other hand, if it is determined that the absolute value of the difference between the variance in the local BV distribution monitoring table 1111 and the variance set in the monitoring / adjustment policy does not exceed the variance difference threshold, It is determined whether the “status” is “adjustment in progress” (step S45). If it is not determined that the “status” of the business service is “adjustment in progress”, the process returns to step S21. If it is determined that the “status” of the relevant business service is “adjustment in progress”, the “status” of the relevant business service in the local BV distribution monitoring table is changed to “waiting for adjustment” (step S47). Further, the BV recalculation policy setting unit 114 is set to turn on the recalculation flag of the corresponding business service in the BV recalculation policy table 1141 (step S49). Then, the process returns to step S21.

  By performing the processing as described above, even when the local BV distribution is inappropriate, the distribution can be corrected to an appropriate distribution. Accordingly, appropriate resource allocation can be performed.

  Although the embodiment of the present invention has been described above, the present invention is not limited to this. For example, the functional blocks shown in FIGS. 4 and 5A are examples, and do not necessarily indicate actual program modules.

  Further, the flowcharts and sequence diagrams are not necessarily the same flow and sequence as long as the same processing result can be obtained. You may make it change the order of a step or to perform in parallel.

  Further, the local optimization mechanism and the overall optimization mechanism 110 are, for example, computer devices. As shown in FIG. 23, the computer device includes a memory 2501 (storage unit), a CPU 2503 (processing unit), a hard disk drive (HDD). 2505, a display control unit 2507 connected to the display device 2509, a drive device 2513 for the removable disk 2511, an input device 2515, and a communication control unit 2517 for connecting to a network are connected by a bus 2519. Application programs including an operating system (OS) and a Web browser are stored in the HDD 2505, and are read from the HDD 2505 to the memory 2501 when executed by the CPU 2503. If necessary, the CPU 2503 controls the display control unit 2507, the communication control unit 2517, and the drive device 2513 to perform necessary operations. Further, data in the middle of processing is stored in the memory 2501 and stored in the HDD 2505 if necessary. Such a computer realizes various functions as described above by organically cooperating hardware such as the CPU 2503 and the memory 2501 described above with the OS and necessary application programs.

  The local optimization mechanism and the overall optimization mechanism 110 may be implemented on separate computers or on the same computer.

(Appendix 1)
Receiving an importance given to a resource node to be assigned or assigned to a specific business service based on a unique policy in the specific business service;
According to the data relating to the importance recalculation policy set independently of the unique policy for the specific business service and stored in the policy storage unit, the received importance is converted into the importance in the entire system and stored. Storing in the device;
A method for calculating the importance of a resource, which is executed by a computer.

(Appendix 2)
The importance calculation method for resources according to claim 1, further comprising the step of switching the importance recalculation policy for the specific business service.

(Appendix 3)
The supplementary note 1 wherein the importance recalculation policy is at least one of a method for performing interval compression and interval shift, a method for performing mapping to a discontinuous predetermined value, and a method using a predetermined function value. Of calculating the importance of a resource.

(Appendix 4)
The importance calculation method for a resource according to appendix 1, wherein the importance recalculation policy includes a rule regarding a range of importance after conversion.

(Appendix 5)
Determining whether the received distribution of importance deviates from a predetermined standard;
A correction step for correcting the received distribution of importance when it is determined that the distribution of the received importance is determined in advance and deviates from a standard stored in the distribution adjustment data storage unit;
The importance calculation method for the resource described in Appendix 1 including:

(Appendix 6)
The correction step includes
Disabling the importance recalculation policy;
Correcting the received importance so as to be distributed according to the distribution stored in the distribution adjustment data storage unit;
The importance calculation method for the resource according to appendix 5, including:

(Appendix 7)
The correction step includes
When the difference between the average value of the received importance and the reference stored in the distribution adjustment data storage unit is a predetermined value or more, the received importance is determined according to the difference from the reference. The importance calculation method for resources according to appendix 5, including a step of correcting.

(Appendix 8)
Receiving an importance given to a resource node to be assigned or assigned to a specific business service based on a unique policy in the specific business service;
According to the data relating to the importance recalculation policy set independently of the unique policy for the specific business service and stored in the policy storage unit, the received importance is converted into the importance in the entire system and stored. Storing in the device;
A program that causes a computer to execute.

(Appendix 9)
Means for receiving an importance assigned to a resource node to be assigned or assigned to a specific business service based on a unique policy in the specific business service;
According to the data relating to the importance recalculation policy set independently of the unique policy for the specific business service and stored in the policy storage unit, the received importance is converted into the importance in the entire system and stored. Means for storing in the device;
A device for calculating the importance of a resource.

110 Global Optimization Mechanism 111 Local BV Distribution Monitoring / Adjustment Unit 112 Distribution Adjustment Policy Setting Unit 113 BV Recalculation Unit 114 BV Recalculation Policy Setting Unit 115 Resource Management Unit 116 Resource Request / Return Reception Unit 117 Distribution Determination Unit 1111 Local BV Distribution Monitoring table 1121 Distribution adjustment policy table 1141 BV recalculation policy table 1151 Resource management table 1171 Resource request management table 1172 Distribution change management table

Claims (3)

A process of receiving importance about the specific business service, which is assigned to the resource node to be allocated to the specific business service or assigned based on a policy corresponding to the specific business service ;
When it is detected that the characteristics of the importance distribution for the specific business service stored in the first storage unit deviate from a reference beyond a predetermined range, the specific business service A process of converting the importance of the received specific business service into the importance of the entire system , so that the distribution of importance for is corrected,
The importance calculation method for resources, which is executed by a computer. The conversion process is,
The importance for the resource according to claim 1, further comprising: converting the importance of the received specific business service into the importance of the entire system so as to be distributed according to the distribution stored in the second storage unit. Degree calculation method. The characteristics of the distribution of importance for a particular business service, is the average value of importance the conversion processing for the particular business service,
The importance of the received specific business service according to the difference between the average value of the importance of the specific business service stored in the first storage unit and a preset reference value. The importance calculation method for the resource according to claim 1, further comprising: converting a degree into an importance in the entire system .

Images