WO2017017775A1 - Computer system and volume allocation control method - Google Patents

Abstract

A management computer connected to a storage device that has a plurality of types of media having different performance levels and that provides, to a computer, a volume in which regions of at least one of the media have been allocated to said computer, with said management computer allocating the volume to the computer, wherein said management computer: manages a hierarchy that is formed by the plurality of types of media such that higher performance media are assigned to higher tiers of the hierarchy; allocates at least one tier of the hierarchy to the volume in accordance with a required performance preset for the computer; calculates an allocation modification scheme for the volume by assuming that the uppermost tier of the hierarchy has been removed; and calculates a predicted value of the performance of the volume assuming that the allocation modification scheme has been applied.

Description

Computer system and volume allocation control method

The present invention relates to a storage apparatus that allocates pages to virtual volumes.

A virtual volume (hereinafter referred to as volume) is composed of a plurality of areas (hereinafter referred to as pages) having different performances depending on storage virtualization technology. The pool provided by the storage device is composed of a plurality of pages.

The performance of the virtual volume depends on the type of media (SSD: Solid State Drive, SAS: Serial Attached SCSI, SATA: Serial ATA) etc. The media performance can be indexed by the read / write speed, the number of I / O processes per unit time, and the like. Then, the hierarchy is set according to the performance of the media. For example, a medium with a fast read / write speed is set as the upper hierarchy, and a medium with a low read / write speed is set as the lower hierarchy.

As for the performance of a volume, for example, if a higher tier (SSD) is assigned more, the average performance of the volume is higher than if a lower tier (SATA) is assigned more.

Service providers such as data centers establish SLA (Service Level Agreement) with customers. In the SLA, required performance including I / O performance and response performance of the storage apparatus is defined. A service provider can improve the cost performance of a data center or the like by providing a service with a minimum configuration that satisfies the required performance.

However, even if the required performance of the SLA is constant, the workload of services provided by the data center changes every moment, and it is difficult for service providers to predict in advance.

In the case of a volume, depending on the change in workload, there may be a situation where an excessively high level media that significantly exceeds the required performance is allocated. For this reason, as a service provider, it is desirable to eliminate the excessive allocation of upper layers by providing the minimum resources that satisfy the required performance.

In the prior art, a technique for solving the above problem by migration processing between a plurality of pools is known (for example, Patent Document 1). In this prior art, media is migrated from a pool with sufficient resources to a pool with insufficient performance and capacity in a storage apparatus having a plurality of pools. Thereby, the ratio of each layer of media allocated to the volume is changed.

International Publication No. 2011/092738 Specification

However, since the media (or volume) is moved between pools in Patent Document 1 described above, a large amount of I / O occurs in the migration source pool and the migration destination pool, and the performance of the volume mapped to both pools. There was a problem that would affect.

Accordingly, the present invention has been made in view of the above problems, and an object of the present invention is to eliminate an excessive allocation of upper-layer media to a volume of a storage device by suppressing an influence on other volumes. .

A management system that includes a plurality of types of media having different performances, and that provides the computer with a volume in which one or more areas of the media are allocated to the computer, a processor, and a memory, and allocates the volume to the computer The management computer manages a hierarchy in which the plurality of types of media are set in advance from an upper hierarchy to a lower hierarchy according to a performance order, and is preset in the computer. Depending on the required performance, a storage device management unit that assigns one or more of the tiers to the volume, a computer management unit that assigns the volume to the computer, and the highest tier of the tiers for the volume Calculate the allocation change plan that was assumed to be deleted, and apply the volume to which the allocation change plan was applied. With an excess resource control unit for calculating a predicted value of the performance of arm, the.

According to the present invention, it is possible to change the configuration of a volume that has been over-allocated with computer resources, to reduce the volume range that affects performance, and to reduce the impact on an operating volume. it can.

1 is a block diagram illustrating an example of a computer system according to a first embodiment of this invention. FIG. It is a block diagram which shows a 1st Example of this invention and shows an example of the functional element of a computer system. It is a block diagram which shows a 1st Example of this invention and shows an example of the functional element of a computer management server. It is a flowchart which shows a 1st Example of this invention and shows an example of the process performed by the excess resource control part of a management server. It is a flowchart which shows a 1st Example of this invention and shows an example of the trial calculation process performed with a management server. It is a flowchart which shows a 1st Example of this invention and shows an example of the execution determination process performed by the management server. It is a flowchart which shows a 1st Example of this invention and shows an example of the allocation change process performed with a management server. It is a flowchart which shows a 1st Example of this invention and shows an example of the volume production | generation process performed with a management server. It is a figure which shows 1st Example of this invention and shows an example of a start condition definition table. It is a figure which shows 1st Example of this invention and shows an example of a pool management table. It is a figure which shows the 1st Example of this invention and shows an example of a volume management table. It is a figure which shows the 1st Example of this invention and shows an example of a trial calculation result management table. It is a figure which shows 1st Example of this invention and shows an example of a threshold value management table. It is a figure which shows 1st Example of this invention and shows an example of an instance classification management table. It is a figure which shows the 1st Example of this invention and shows an example of a hierarchy allocation initial value management table. It is a figure which shows the 1st Example of this invention and shows an example of a hierarchy allocation management table. It is a flowchart which shows a 2nd Example of this invention and shows an example of the execution determination process performed by the management server. It is a figure which shows the 2nd Example of this invention and shows an example of an instance management table. It is a figure which shows the 2nd Example of this invention and shows an example of a classification management table. It is a flowchart which shows a 3rd Example of this invention and shows an example of the allocation change process performed with a management server. It is a figure which shows the 3rd Example of this invention and shows an example of a volume management table.

Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

FIG. 1 is a block diagram showing an example of a computer system according to the first embodiment of this invention. The computer system to which the present invention is applied shows an example in which a virtual machine, a communication network 7, and a management network 8 on the hosts 3-1 to 3-m are provided to a customer as IaaS (Infrastructure as a Service).

The computer resources of the data center used by the service user terminal 6 operated by the service user are the hosts 3-1 to 3-m, the storage devices 2-1 to 2-n, the WEB server 5, the communication network 7, and the management network. 8 and SAN (Storage Area Network) 9. The whole of the hosts 3-1 to 3-m is indicated by a symbol “3” without “−”. The same applies to the reference numerals of other components.

The host 3 and the storage device 2 are connected via the SAN 9. The host 3 and the storage device 2 are connected to the management server 1 and the service providing side management terminal 4 via the management network 8.

The management server 1 controls the host 3 and the storage device 2 based on a command from the service providing side management terminal 4. For example, the management server 1 controls the host 3, the storage device 2, the management network 8, the communication network 7, or the SAN 9 so as to satisfy the SLA concluded with the service user.

The service user accesses the WEB server 5 through the service user terminal 6, and requests a computer resource allocation request or a change request through the self-service portal 51 provided by the WEB server 5. The self service portal 51 transmits the request received from the service user terminal 6 to the management server 1. The management server 1 controls the allocation of computer resources according to the received request, and responds to the self-service portal 51 of the WEB server 5 with the allocation result. The self-service portal 51 responds to the service user terminal 6 with the computer resource allocation result in response to the request.

The management server 1 includes a processor 11, a memory 12, a nonvolatile storage device 13, an I / O interface 14 connected to the management network 8, and an I / O interface 15 connected to the communication network 7. The processor 11 executes a program described later to manage computer resources such as the host 3 and the storage device 2.

The host 3 includes a processor 31, a memory 32, an I / O interface 34 connected to the storage apparatus 2, and an I / O interface 33 connected to the management network 8. The processor 31 provides a virtual machine generated by the hypervisor as will be described later.

The storage device 2 includes a processor 21, a memory 22, an I / O interface 24 connected to the SAN, an I / O interface 25 connected to the management network 8, and an I / O interface 23 connected to the medium 26.

The media 26 includes, for example, an SSD (Solid State Drive) 27, an SAS (Serial Attached Scsi) 28, and a SATA (Serial State ATA) 29. In this embodiment, the performance of the media 26 is classified hierarchically according to the difference in performance based on the read / write speed, the number of I / O processes per unit time (IOPS), and the like. The performance of the medium 26 indicates processing performance and can be expressed by I / O processing speed (IOPS), data transfer speed (sequential read / write or random read / write), and the like.

For example, the SSD 27 is the highest layer with the highest performance, the SAS 28 is the middle layer, and the SATA 29 is the lower layer. Regarding the performance of the medium 26, a non-volatile semiconductor storage device such as the SSD 27 may be the upper layer, and the SAS 28 and SATA 29 including the disk may be the lower layer. Alternatively, it may be classified into different layers according to the rotation speed of the disk, and the SAS28 or SATA29 media having a disk whose rotation speed is equal to or higher than a predetermined threshold is set as the upper hierarchy and the disk whose rotation speed is lower than the predetermined threshold is included. The media of SAS 28 or SATA 29 may be a lower layer.

Alternatively, the hierarchy may be set according to the type (configuration) of the medium 26, and the nonvolatile semiconductor memory device (SSD 27) may be the upper hierarchy, and the SAS 28 or SATA 29 HDD may be the lower hierarchy. Further, hierarchization may be performed according to the difference in configuration of the SSD 27. For example, the SSD 27 of SLC (Single Level Cell) may be an upper layer, and the SSD 27 of MLC (Multi Level Cell) may be a middle layer or a lower layer. Good.

In this embodiment, the performance difference of the media 26 is classified as a hierarchy. However, the present invention is not limited to this, and it is sufficient that the elements (devices) of the media 26 can be classified into groups according to the performance difference. For example, the SSD 27 may be classified as a high speed device, the SAS 28 as a medium speed device, and the SATA 29 as a low speed device.

The storage device 2 provides one or more pools, and the pool 220 is composed of a plurality of areas (hereinafter, pages) of the media 26 having different performance. The management server 1 holds the required performance based on the SLA. Upon receiving an allocation request from the service user terminal 6, the management server 1 instructs the storage device 2 to generate a virtual volume with the performance based on the SLA (required performance). To do. The storage apparatus 2 generates a virtual volume by allocating a plurality of pages so as to satisfy the required performance, and provides the virtual volume to the host 3 used by the service user terminal 6.

In the first embodiment, the pool in the storage apparatus 2 shows an example in which three types of media 26 (storage device) areas of SSD 27, SAS 28, and SATA 29 are configured. However, the configuration is not limited to this, and any combination of media 26 having different performances may be used. For example, a combination of two types of SAS (for example, 15000 rpm and 10000 rpm) having different performance from the SSD 27 may be used. Further, the number of combinations of regions need not be limited to three, and may be any number.

Further, for the management network 8 and the communication network 7, dedicated communication lines are constructed between the management server 1, the host 3, and the WEB server 5, but the same physical line may be shared.

FIG. 2A is a block diagram showing an example of functional elements of the computer system. The functional elements of the management server 1 are shown in FIG. 2B.

In the host 3-1, the hypervisor 300 is executed by the processor 31, and the virtual machines 310-1 to 310-q generated in response to a command from the management server 1 operate. In each virtual machine 310, applications 320-1 to 320-P are executed on a guest OS (not shown), and each application 320 is provided to the service user terminal 6. The same applies to the other hosts 3, and one or more virtual machines operate on the hypervisor to execute the application.

In the storage device 2-1, the storage management unit 230 including the storage control unit 232 and the access frequency statistical information 233 operates.

The storage control unit 232 is loaded into the memory 22 as a program. The processor 21 operates as a functional unit that provides a predetermined function by performing processing according to a program of each functional unit. For example, the processor 21 functions as the storage control unit 232 by performing processing according to the storage control program. The same applies to other programs. Furthermore, the processor 21 also operates as a function unit that provides each function of a plurality of processes executed by each program. A computer and a computer system are an apparatus and a system including these functional units.

The storage control unit 232 counts the access frequency (number of times) for each page or data of the SSD 27, SAS 28, and SATA 29 constituting the medium 26, and stores it in the access frequency statistical information 233.

The storage management unit 230 generates one or more pools 220-1 to 220-i from each page of the SSD 27, SAS 28, and SATA 29 constituting the medium 26. Then, the storage management unit 230 allocates pages in each pool 220 and generates virtual volumes 210-1 to 210-k. The storage management unit 230 provides the virtual volume 210 to the virtual computer 310 based on a command from the management server 1. In the first embodiment, an example of handling a set of pages as a virtual volume 210 is shown, but it can be handled as a volume.

FIG. 2B is a block diagram illustrating an example of functional elements of the management server 1. In the memory 12 of the management server 1, an IaaS management unit 110 that provides computer resources based on an SLA defined with a service user, a storage device management unit 130 that manages the storage device 2, and a virtual host 3 A virtual machine management unit (computer management unit) 140 that manages the computer 310, an excess resource control unit 430 that selects the virtual volume 210 to be processed, and an appropriate allocation trial calculation unit 440 that performs a trial calculation of tier allocation to the virtual volume 210. And an execution determination unit 450 that determines whether or not to change the allocation of the virtual volume 210, an allocation change execution unit 460 that changes the floor layer allocation to the virtual volume 210, and a new virtual volume 210. Each function unit of the volume generation unit 470 is stored in the processor 11. It is executed me.

In the memory 12 of the management server 1, the instance type management table 120, the start condition definition table 150, the pool management table 160, the volume management table 170, and the trial calculation result management table 180 are used in order to operate the above functional units. And a threshold management table 190, a tier allocation initial value management table 410, and a tier allocation management table 420 at the time of volume generation are stored.

IaaS management unit 110, storage device management unit 130, virtual machine management unit 140, excess resource control unit 430, appropriate allocation trial calculation unit 440, execution determination unit 450, allocation change execution unit 460, and volume generation unit 470 Is loaded into the memory 12 as a program.

The processor 11 operates as a functional unit that provides a predetermined function by processing according to a program of each functional unit. For example, the processor 11 functions as the IaaS management unit 110 by performing processing according to the IaaS management program. The same applies to other programs. Furthermore, the processor 11 also operates as a functional unit that provides each function of a plurality of processes executed by each program. A computer and a computer system are an apparatus and a system including these functional units.

Upon receiving a request from the service user terminal 6 via the WEB server 5, the IaaS management unit 110 identifies an instance based on a preset SLA, determines computer resources to be allocated to the instance, The management unit 130 and the virtual machine management unit 140 are notified. The IaaS management unit 110 holds an SLA (not shown) preset for each service user in the memory 12 or the storage device 13.

The virtual machine management unit (computer management unit) 140 generates the virtual computer 310 notified from the IaaS management unit 110. The storage device management unit 130 causes the storage device 2 to generate the virtual volume 210 requested from the IaaS management unit 110 and allocates it to the virtual computer 310. Further, the IaaS management unit 110 sets the management network 8 and the communication network 7 according to the specified instance.

In the first embodiment, an example is shown in which the virtual machine management unit 140 manages the virtual volume 210 to be allocated to the virtual computer 310. However, the present invention is not limited to this, and the IaaS management unit 110 may manage the virtual volume 210 to be allocated to the virtual computer 310. Further, the virtual machine management unit 140 has allocation management information (not shown) and manages the virtual volume 210 to be allocated to the virtual computer 310.

In addition, in the said Example 1, although the excess resource control part 430, the appropriate allocation trial calculation part 440, the execution determination part 450, and the allocation change execution part 460 were shown as an independent functional part, it is not limited to this. For example, the excessive resource control unit 430 may include a proper allocation trial calculation unit 440, an execution determination unit 450, and an allocation change execution unit 460.

In the first embodiment, an example in which a tier allocation change plan is calculated for a virtual volume 210 that satisfies the trial calculation start condition is shown, but the present invention is not limited to this, and for any pool 220 or virtual volume 210 You may implement the said process.

<Overview>
The outline of the present invention is to calculate the configuration of the media 26 allocated to the virtual volume 210 in an environment where there are a plurality of types of media 26 having different performances for providing the pages constituting the virtual volume 210 as described later. The virtual volume 210 in which the computer resources of the device 2 are excessively allocated is extracted. That is, when the average performance indicating the average value of IOPS, which is the performance of the virtual volume 210, exceeds the threshold value as compared with the required performance, the management server 1 allocates excessive computer resources to the virtual volume 210 to be processed. Determine that it has occurred.

The management server 1 calculates the performance of the virtual volume 210 when a volume is configured in a lower hierarchy, assuming that the allocation of the highest hierarchy is deleted for the virtual volume 210 to be processed. As a result of the trial calculation, the management server 1 calculates a configuration that satisfies the required performance based on the SLA and has the minimum number of layers as a layer (page) allocation change plan. Then, the management server 1 changes the allocation of the tier (page) and changes the allocation of the tier of the virtual volume 210 of the storage device 2 when a predetermined condition is satisfied, such as when the load of the pool 220 is not high. To do.

According to the present invention, it is not necessary to perform migration between pools as in the conventional example, the range of the virtual volume 210 that affects performance can be reduced, and the influence on the operating virtual volume 210 is reduced. can do. Then, the allocation of the performance can be performed by allocating the page of the upper hierarchy whose allocation has been canceled to the other virtual volume 210 in the same pool 220. In addition, the management server 1 collects surplus computer resources from the storage device 2, whereby the volume aggregation rate in the pool 220 (the number of volumes that can satisfy the SLA) can be improved.

Hereinafter, each table managed by the management server 1 will be described.

<Instance type management table>
FIG. 13 is a diagram illustrating an example of the instance type management table 120 managed by the IaaS management unit 110.

The instance type management table 120 includes an instance type 1201, a CPU 1202 that stores the number of processors 31 (or the number of cores) allocated to the instance, a memory 1203 that stores the capacity of the memory 32 allocated to the instance, and an instance One record includes a capacity 1204 for storing the capacity of the virtual volume 210 to be allocated, an IOPS 1205 for storing the performance of the virtual volume 210 to be allocated to the instance, and a network 1206 for storing the performance (transfer speed) of the network to be allocated to the instance. Including. The instance type management table 120 may be set for each service user.

<Start condition definition table>
FIG. 8 is a diagram illustrating an example of the start condition definition table 150. The start condition definition table 150 is preset with a trigger for executing the excess resource control unit 430 and is managed by the storage device management unit 130.

The start condition definition table 150 includes a condition ID 1501 for storing condition identifiers, an execution condition 1502 for storing triggers for executing the excess resource control unit 430, and a volume for storing conditions for generating a list of virtual volumes 210 to be processed. One record includes a listing condition 1503 and a periodicity 1504 for storing whether or not the execution condition is periodically satisfied.

The execution condition 1502 stores the execution start time and the performance conditions of the virtual volume 210. For example, the condition ID 1501 = “2” indicates that the excess resource control unit 430 is executed when the volume listing condition 1503 “volume whose average performance exceeds three times the required performance” is detected. That is, the volume listing condition 1503 holds a condition (trial calculation start condition) for calculating the allocation state of computer resources.

The average performance and the required performance of the virtual volume 210 are compared by the management server 1 acquiring the average performance from each storage device 2 and comparing it with the required performance 1705 set in the volume management table 170 in advance based on the SLA. be able to.

Here, the required performance is a performance value of the virtual volume 210 determined by the SLA as in the volume management table 170 described later, and in the first embodiment, an example is shown in which IOPS is used as an index of the performance of the volume. However, the performance of the virtual volume 210 is not limited to this, and read / write speed, response time, and the like may be used. In addition, for the average performance of the volume, the management server 1 may obtain the value measured for each volume 210 by the storage management unit 230 of the storage device 2.

The volume listing condition 1503 stores the condition of the volume to be selected as a processing target. Condition IDs 1501 = 1 and 2 indicate that the average performance of the virtual volume 210 is to be processed when it exceeds the required value by a predetermined value.

In the periodicity 1504, “Yes” is set if the execution condition 1502 periodically executes the excess resource control unit 430, and “No” is set otherwise. An execution condition of “Yes” for periodicity 1504 can be executed as an event by a timer (not shown) of the management server 1. On the other hand, when the periodicity 1504 is an execution condition of “No”, the management server 1 can execute it at a predetermined cycle (for example, 30 minutes).

<Pool management table>
FIG. 9 is a diagram illustrating an example of a pool management table 160 that manages the pool 220 generated by the storage apparatus 2. The storage device management unit 130 of the management server 1 acquires information about the pool 220 from the storage device 2 at a predetermined cycle (for example, 10 minutes), and updates the pool management table 160. When there are a plurality of storage apparatuses 2, the storage apparatus management unit 130 of the management server 1 generates a pool management table 160 for each of the storage apparatuses 2-1 to 2-n.

The pool management table 160 includes a pool ID 1601 that stores an identifier of the pool 220, a page ID 1602 that stores an identifier of a page that configures the pool 220, a hierarchy 1603 that stores the type of the media 26 that configures the page, A page size 1604 for storing the capacity (MB) of the page, an average performance 1605 for storing the average value of IOPS (or access performance) of the page, and an allocation destination for storing the identifier of the virtual volume 210 to which the page is allocated A volume (VOL in the figure) ID 1606 is included in one record.

The management server 1 can monitor the average performance for each page allocated to the virtual volume 210 by using the pool management table 160.

<Volume management table>
FIG. 10 is a diagram showing an example of a volume management table 170 that manages the virtual volume 210 generated by the storage apparatus 2. The storage device management unit 130 of the management server 1 acquires information on the virtual volume 210 from the storage device 2 at a predetermined cycle (for example, 10 minutes), and updates the volume management table 170. When there are a plurality of storage apparatuses 2, the storage apparatus management unit 130 generates a volume management table 170 for each of the storage apparatuses 2-1 to 2-n.

The volume management table 170 includes a volume (VOL in the figure) ID 1701 that stores the identifier of the virtual volume 210, a capacity 1702 that stores the capacity (GB) of the virtual volume 210, and the types of media 26 that make up the virtual volume 210. Stores an allocation hierarchy 1703 for storing the ID, a pool ID 1704 for storing the identifier of the pool 220 to which the medium 26 constituting the virtual volume 210 belongs, and an IOPS request value set for the virtual volume 210 based on the SLA. The required performance 1705 and the average performance 1706 for storing the average value of IOPS of the virtual volume 210 are included in one record.

The average performance 1706 is a value acquired by the management server 1 for the average performance for each volume 210 measured by the storage device 2. Note that “N / A” is set to the virtual volume 210 in which the required performance 1705 is not set.

<Trial calculation result management table>
FIG. 11 is a diagram illustrating an example of a trial calculation result management table 180 that stores the results calculated by the proper allocation trial calculation unit 440 of the management server 1. The trial calculation result management table 180 is updated each time the proper allocation trial calculation unit 440 executes a calculation.

The trial calculation result management table 180 stores a trial calculation result ID 1801 for storing a trial calculation identifier, a volume ID 1802 for storing the identifier of the virtual volume 210 subject to the trial calculation, and a type of the medium 26 constituting the virtual volume 210. The virtual volume 210 is configured by the allocation hierarchy 1803, the trial allocation hierarchy 1804 that stores the configuration of the media 26 that has performed the trial calculation, the average performance 1805 that stores the average value of IOPS of the virtual volume 210, and the trial allocation hierarchy 1804. One record includes a predicted performance 1806 for storing a predicted value of performance in the case where it is assumed.

The allocation hierarchy 1803 indicates the type of page that is actually allocated to the virtual volume 210. On the other hand, the trial allocation layer 1804 stores the types of pages that can satisfy the SLA.

The average performance 1805 is the same as the average performance 1706 of the volume management table 170, and is the current IOPS average value of the virtual volume 210. On the other hand, the predicted performance 1806 is a predicted value of average performance when it is assumed that the virtual volume 210 is configured with the media 26 of the trial allocation layer 1804.

<Threshold management table>
FIG. 12 is a diagram illustrating an example of the threshold management table 190 of the management server 1. The threshold management table 190 holds thresholds for determination in allocation change processing described later, and is managed by the excess resource control unit 430. The threshold management table 190 is information preset by a service provider or the like. The comparison item 1901, the threshold value 1902, and the valid flag 1903 are included in one record.

In the comparison item 1901, an item to be compared with the target whose hierarchy assignment is changed is set. In the first embodiment, an example of using either “pool capacity” or “volume capacity” as the comparison item 1901 is shown.

In the threshold value 1902, a value for executing the allocation change process when the threshold value is not exceeded is set. For example, when the comparison item 1901 is “volume capacity”, if the capacity in the virtual volume 210 that is the target of tier allocation change does not exceed “50%”, this indicates that the allocation change process is executed. The threshold value 1902 is a value that can be set by a service provider or a computer system administrator.

The valid flag 1903 sets whether to use the threshold value. If the valid flag 1903 = “YES”, the threshold value 1902 of the record is used. On the other hand, if the valid flag 1903 = “NO”, the threshold value 1902 of the record is not used. The validity flag 1903 is set to “YES” for any one of the plurality of records in the threshold management table 190. In other words, there is one threshold value determined in the assignment change process. The valid flag 1903 is a value that can be set by a service provider or a computer system administrator.

<Hierarchy allocation initial value management table>
FIG. 14 is a diagram illustrating an example of the tier allocation initial value management table 410 used when the management server 1 generates the virtual volume 210. The tier allocation initial value management table 410 is information set in advance by a service provider or the like, and is managed by the storage device management unit 130.

The hierarchical allocation initial value management table 410 includes the required performance 4101 and the hierarchical allocation order 4102 in one record. The required performance 4101 stores the IOPS (I / O processing speed) of the virtual volume 210 to be generated. The hierarchy allocation order 4102 stores the order in which the constituent elements of the media 26 are assigned according to the required performance 4101.

<Hierarchy allocation management table>
FIG. 15 is a diagram showing an example of the tier allocation management table 420 used when the management server 1 creates the virtual volume 210. The tier allocation management table 420 is information preset for each storage device 2 by a service provider or the like, and is managed by the storage device management unit 130.

The hierarchy allocation management table 420 includes a hierarchy allocation order 4201 and an allocation hierarchy 4202 in one record. The tier allocation order 4201 stores the order corresponding to the tier assignment order 4102 of the tier assignment initial value management table 410 shown in FIG.

The assignment hierarchy 4202 stores the types of components of the media 26 assigned to the virtual volume 210. In the first embodiment, the SSD 27 that is the upper layer medium 26 having the highest I / O processing speed is set in the layer allocation order 4102 = “1”. In the hierarchy allocation order 4102 = “2”, a combination of the SSD 27 of the medium 26 of the upper hierarchy and the SAS 28 of the middle hierarchy is set. In the hierarchy assignment order 4102 = “3”, a combination of the SSD 27 of the medium 26 of the upper hierarchy, the SAS 28 of the middle hierarchy, and the SATA 29 of the lower hierarchy is set. In the hierarchy allocation order 4102 = “4”, a combination of the SAS 28 in the middle hierarchy and the SATA 29 in the lower hierarchy is set. Only the lower layer SATA 29 is set in the layer allocation order 4102 = “5”.

<Overall processing>
FIG. 3 is a flowchart illustrating an example of processing performed by the excess resource control unit (resource control unit) 430 of the management server 1. The excess resource control unit 430 of the management server 1 executes the process according to the execution condition 1502 of the start condition definition table 150, and sets the virtual volume 210 corresponding to the volume listing condition 1503 as a virtual volume to be processed in a list (list) (not shown). 210 is extracted (S1). The list includes a volume ID 1701 and a pool ID 1704 of the volume management table 170.

The excess resource control unit 430 performs a trial calculation of tier allocation for the processing target virtual volume 210 of the storage apparatus 2 corresponding to the execution condition 1502. This trial calculation of tier allocation is performed under a preset condition such as a comparison between required performance and average performance, or a specified pool ID.

The excess resource control unit 430 determines whether there are one or more virtual volumes 210 extracted in the list (S2). If there are one or more virtual volumes 210 in the list, the excess resource control unit 430 proceeds to step S3. On the other hand, if there is no virtual volume 210 in the list, the process ends.

In step S3, the excess resource control unit 430 selects one virtual volume 210 from the list, and calculates the hierarchy of pages to be allocated to the virtual volume 210 using the virtual volume 210 as a processing target. As will be described later, the excess resource control unit 430 notifies the appropriate allocation trial calculation unit 440 of the management server 1 of the volume ID 1701 and the pool ID 1704 and executes the trial calculation. As a result of the trial calculation, the appropriate allocation trial calculation unit 440 responds to the excess resource control unit 430 with a plan for changing the allocation of a tier (page) that satisfies the required performance based on the SLA.

The excess resource control unit 430 notifies the execution determination unit 450 of the hierarchy allocation change plan received from the appropriate allocation calculation unit 440 and determines whether or not the execution determination unit 450 satisfies a condition for executing the allocation change. (S4). In the first embodiment, the excessive resource control unit 430 notifies the execution determination unit 450 of the capacity of the virtual volume 210 to be changed from the current tier assignment in the tier assignment change plan. As will be described later, the execution determination unit 450 determines whether or not a condition for executing the tier allocation change plan is satisfied from the capacity threshold and the capacity of the virtual volume 210 to be changed, and notifies the excess resource control unit 430 of the condition. .

If the condition for executing the hierarchy allocation change plan is satisfied, the process proceeds to step S7. If the condition is not satisfied, the process proceeds to step S5.

In step S7, as will be described later, the excessive resource control unit 430 notifies the allocation change execution unit 460 of the allocation change plan and causes the allocation change plan to be executed. The allocation change execution unit 46 changes the configuration of the media 26 allocated to the virtual volume 210 based on the trial calculation result management table 180, and notifies the excess resource control unit 430 when the processing is completed.

The excess resource control unit 430 deletes the virtual volume 210 for which the allocation change plan has been implemented from the list, and then returns to step S2 to repeat the above processing.

On the other hand, in step S5 in the case where the condition for executing the tier allocation change plan is not satisfied, the excess resource control unit 430 refers to the pool management table 160 to determine whether or not it has been updated. The excess resource control unit 430 determines whether or not the record having the pool ID to which the virtual volume 210 for which the allocation change plan is calculated in step S3 belongs has been updated.

If the content of the record having the pool ID to which the processing target virtual volume 210 belongs is updated in the pool management table 160, it can be assumed that the load of the pool 220 has increased. Repeat the determination.

On the other hand, if the content of the pool ID to which the processing target virtual volume 210 belongs has not been updated in the pool management table 160, after waiting for a predetermined time (N minutes) in step S6, the process returns to step S3, and the next A trial calculation of an appropriate hierarchy to be allocated to the virtual volume 210 is executed. The predetermined time (N minutes) can be set from the service provider side management terminal 4.

Through the above processing, the excess resource control unit 430 acquires the virtual volume 210 corresponding to the volume listing condition 1503 as the processing target virtual volume 210 that satisfies the trial calculation start condition and adds it to the list. Then, the proper allocation trial calculation unit 440 calculates a proper tier to be assigned to the virtual volume 210 in the list and calculates it as a tier assignment change plan. Then, the excessive resource control unit 430 performs the tier allocation change for the processing target that can execute the tier allocation change plan.

In the process of step S6, the excessive resource control unit 430 can delete the virtual volume 210 to be processed that does not satisfy the execution condition of the allocation change process from the list while waiting for a predetermined time. This prevents the hierarchical allocation change plan that cannot execute the allocation change process from being repeatedly processed.

<Proper allocation calculation process>
FIG. 4 is a flowchart illustrating an example of a trial calculation process performed by the proper allocation trial calculation unit 440 of the management server 1. This process is a process performed in step S3 of FIG.

The proper allocation trial calculation unit 440 first acquires information of the pool management table 160 and the volume management table 170 (S10).

The proper allocation trial calculation unit 440 acquires a record that matches the volume ID 1701 allocated to the processing target virtual volume 210 from the volume management table 170. The acquired record includes a volume ID 1701 and a pool ID 1704. Further, the proper allocation trial calculation unit 440 acquires a record with the matching pool ID 1601 from the pool management table 160. As a result, the page information of the pool 220 related to the processing target virtual volume 210 can be acquired.

Next, the proper allocation trial calculation unit 440 acquires the access frequency statistical information 233 from the storage management unit 230 of the storage device 2 having the processing target virtual volume 210 (S11).

The proper allocation trial calculation unit 440 refers to the allocation hierarchy 1703 of the processing target virtual volume 210 from the information of the volume management table 170, and determines whether or not the medium 26 constituting the virtual volume 210 is only the SATA 29 (S12). ). That is, it is determined whether or not the processing target virtual volume 210 is composed of the media 26 in the lowest hierarchy. In the first embodiment, the example in which the lowest hierarchy is the SATA 29 is shown, but the present invention is not limited to this.

The proper allocation trial calculation unit 440 proceeds to step S13 if the virtual volume 210 to be processed is configured only by the SATA 29, and uses the current configuration as the trial calculation result.

On the other hand, if it is determined in step S12 that the medium 26 other than the SATA 29 is included in the virtual volume 210 to be processed, the process proceeds to step S15.

In step S15, the proper allocation trial calculation unit 440 generates an allocation plan in which the highest tier of the virtual volume 210 to be processed is removed. That is, the appropriate allocation trial calculation unit 440 acquires the highest hierarchy and the second highest hierarchy from the data of the allocation hierarchy 1703 of the records acquired from the volume management table 170. Next, the proper allocation trial calculation unit 440 deletes the highest hierarchy from the field of the allocation hierarchy 1703 of the volume management table 170, and assumes that the second hierarchy is the highest hierarchy. The proper allocation trial calculation unit 440 sets this configuration as an allocation change plan.

In step S16, the appropriate allocation trial calculation unit 440 calculates the change in the I / O density distribution in the pool 220. The I / O density distribution is a statistical information of the number of I / Os generated within a certain period for a page. That is, the proper allocation trial calculation unit 440 selects a page in the pool management table 160 that is not allocated to the virtual volume 210 and that is in the second tier. Based on the access frequency statistical information 233, the appropriate allocation trial calculation unit 440 performs a trial calculation to allocate the second tier page from data with high access frequency. However, the pool management table 160 is not updated at this point.

Further, the proper allocation trial calculation unit 440 allocates the record whose allocation destination volume ID 1701 of the volume management table 170 matches the ID of the virtual volume 210 to be processed from the second tier from the top without assigning it to the top tier. .

For example, in the volume management table 170, if “SSD, SAS, SATA” is stored in the allocation hierarchy 1703 of the processing target virtual volume 210, the SSD 27 is allocated without being allocated, and “SAS, SATA” is allocated. Is a hierarchy assignment change plan. That is, the proper allocation trial calculation unit 440 performs trial calculation to determine a page ID 1603 for allocating frequently accessed data to the SAS 28 instead of the SSD 27. The proper allocation trial calculation unit 440 allocates data to a page in the third hierarchy when there is a shortage of free pages in the SAS 28 to which data is allocated.

As described above, the proper allocation trial calculation unit 440 generates a hierarchy allocation change plan in which the destination of the data allocated to the highest hierarchy page is set to the second and subsequent hierarchy pages.

In step S17, the proper allocation trial calculation unit 440 calculates the predicted performance of the virtual volume 210 of the generated tier allocation change plan. First, the appropriate allocation trial calculation unit 440 refers to the pool management table 160 and identifies the page ID 1602 of the second and subsequent layers that are the data migration destination determined in step S16.

At this time, the data destination of the highest tier is specified by the page ID 1603 because the allocation destination volume ID 1606 of the record in the pool management table 160 is blank. Furthermore, the appropriate allocation trial calculation unit 440 specifies a page ID 1603 other than the highest hierarchy in the ID 1606 of the processing target virtual volume.

The proper allocation trial calculation unit 440 acquires the average performance 1605 corresponding to the page ID 1603 included in the allocation change plan of the hierarchy. Next, the appropriate allocation trial calculation unit 440 acquires the access frequency (number of times) to each page of the migration source data and the virtual volume 210 to be processed from the access frequency statistical information 233 acquired in step S11.

The appropriate allocation trial calculation unit 440 calculates the product of the access frequency and the average performance 1605 for each identified page. The proper allocation trial calculation unit 440 calculates the sum of the products of all the pages of the processing target virtual volume 210 and sets the value obtained by dividing the sum by the number of pages as the predicted performance of the processing target virtual volume 210. The number of pages is the sum of the number of pages to which the data of the highest tier is moved and the number of pages other than the highest tier in the ID 1606 of the processing target virtual volume.

Next, the proper allocation trial calculation unit 440 refers to the volume management table 170 and determines whether or not the predicted performance calculated in step S17 is the required performance 1705 or more. If the predicted performance is greater than or equal to the required performance 1705, the current hierarchy allocation change plan is stored, and then the process returns to step S12 to remove the highest hierarchy and repeat the trial calculation.

On the other hand, if the predicted performance is less than the required performance 1705, the appropriate allocation trial calculation unit 440 discards the current hierarchy allocation change plan, and uses the previous hierarchy allocation change plan as the appropriate allocation trial calculation result.

In step S14, the appropriate allocation trial calculation unit 440 stores the trial calculation result in the trial calculation result management table 180. The proper allocation trial calculation unit 440 assigns a unique value to the trial calculation result ID 1801, stores the tier of the processing target virtual volume 210 in the allocation tier 1803, and sets the tier calculated based on the trial calculation result of the proper allocation to the trial calculation tier 1804. The average performance 1705 of the virtual volume 210 to be processed is stored in the average performance 1805, the calculated prediction performance is stored in the prediction performance 1806, and a new record is added to the trial calculation result management table 180.

Through the above processing, the appropriate allocation calculation unit 440 deletes the highest-layer medium 26 from the virtual volume 210 to be processed, and generates a layer allocation change plan composed of the lower-layer medium 26. The appropriate allocation trial calculation unit 440 calculates the prediction performance when it is assumed that the allocation change plan of the hierarchy in which the highest hierarchy is deleted is executed.

Then, the appropriate allocation trial calculation unit 440 repeats trial calculation to delete the media 26 of the highest tier while the prediction performance satisfies the required performance 1705, so that the number of tiers is the minimum and the prediction performance is required. It is possible to calculate a tier allocation change plan of the virtual volume 210 that satisfies the performance.

<Execution determination processing>
FIG. 5 is a flowchart illustrating an example of an execution determination process performed by the execution determination unit 450 of the management server 1. This process is a process performed in step S4 of FIG.

In step S21, the execution determination unit 450 refers to the trial calculation result management table 180 and acquires the content of the record in which the ID of the virtual volume 210 to be processed matches the volume ID 1802.

In step S22, the execution determination unit 450 refers to the threshold management table 190, selects a record with a valid flag 1903 of “YES”, and acquires a comparison item 1901 and a threshold 1902.

In step S23, the execution determining unit 450 calculates the capacity of the volume that is the target of the tier assignment change. The execution determination unit 450 extracts a difference between the record allocation hierarchy 1803 and the trial allocation hierarchy 1804 acquired in the trial calculation result management table 180, and specifies that the difference is a hierarchy that is removed as an upper hierarchy.

For example, in the example of FIG. 11, the allocation hierarchy 1803 of ID = 1 of the current virtual volume 210 is “SSD, SAS, SATA”. On the other hand, the trial allocation layer 1804 of the trial calculation result management table 180 is “SAS, SATA”. Therefore, the execution determination unit 450 specifies that the difference “SSD” is the medium 26 to be removed.

Next, the execution determination unit 450 acquires all records (pages) whose allocation destination volume ID 1606 matches the ID of the virtual volume 210 to be processed from the pool management table 160. These records include a hierarchy 1603 and a page size 1604.

Next, the execution determination unit 450 acquires the value of the page size 1604 for the record in which the hierarchy of the record acquired from the pool management table 160 matches the hierarchy specified as the media 26 to be removed, and calculates the sum. . The execution determining unit 450 uses the calculated sum as the capacity in the virtual volume 210 that is the target of the tier assignment change.

In step S24, the execution determination unit 450 determines whether or not the calculated capacity exceeds the threshold value 1902 for the comparison item 1901. In the threshold management table 190 of FIG. 12, the comparison item 1901 = “volume content” and the threshold 1902 = “50%” are valid.

Therefore, the execution determination unit 450 calculates the ratio of the capacity of the tier assignment change target in the capacity of the virtual volume 210. The execution determination unit 450 acquires the capacity 1702 of the processing target virtual volume 210 from the volume management table 170. Then, the execution determination unit 450
Ratio = capacity to be changed / capacity 1702
To calculate the ratio of the capacity of the virtual volume 210 to be changed. If the calculated ratio exceeds the threshold value 1902, the execution determination unit 450 proceeds to step S 25, outputs “NO” as the determination result, and prohibits the hierarchy assignment change process.

On the other hand, if the calculated ratio is equal to or less than the threshold value 1902, the execution determination unit 450 proceeds to step S 26 to output “YES” as the determination result and permits the hierarchy assignment change process.

As a result of the above processing, it is determined whether or not to change the hierarchy assignment by the comparison item 1901 and the threshold value 1902 set in the threshold value management table 190.

<Allocation change process>
FIG. 6 is a flowchart illustrating an example of an assignment change process performed by the assignment change execution unit 460 of the management server 1. This process is a process performed in step S7 of FIG.

In step S51, the allocation change execution unit 460 transmits a command to execute the tier allocation change plan to the storage management unit 230 of the storage apparatus 2. First, the allocation change execution unit 460 acquires a record in which the volume ID 1802 matches the ID of the virtual volume 210 to be processed from the trial calculation result management table 180. Then, the allocation change execution unit 460 acquires the volume ID 1802 and the trial allocation hierarchy 1804 from the acquired record.

Next, the allocation change execution unit 460 refers to the pool management table 160 and acquires a record in which the ID of the processing target virtual volume 210 matches the allocation destination volume ID 1606. If there is a record in which the record hierarchy 1603 of the acquired pool management table 160 does not match the trial allocation hierarchy 1804 of the trial calculation result management table 180, the allocation change execution unit 460 transfers the page of the record to the virtual volume 210 to be processed. Identify as unassigned pages.

Then, the allocation change execution unit 460 allocates the data of the identified page from the highest layer of the trial allocation layer 1804. Similar to step S16 of FIG. 4, this allocation is performed on the access frequency statistical information on the page that is not allocated to the virtual volume 210 in the record of the pool management table 160 and the page of the highest hierarchy of the trial allocation hierarchy 1804. Based on H.233, the page of the highest hierarchy is allocated from data with high access frequency.

Then, the allocation change execution unit 460 instructs the storage management unit 230 to copy the data of the page that is not allocated to the virtual volume 210 to be processed to the page of the highest hierarchy of the trial allocation hierarchy 1804.

The storage management unit 230 notifies the management server 1 when the copying of the data of the page that is not allocated to the virtual volume 210 to be processed is completed to the top layer of the trial allocation layer 1804. Alternatively, when an error occurs during copying, the storage management unit 230 notifies the management server 1 of the occurrence of the error.

In step S52, the allocation change execution unit 460 determines whether the allocation change is successful based on the notification received from the storage management unit 230 of the storage apparatus 2 in step S52. If the allocation change execution unit 460 receives a copy completion notification, the allocation change execution unit 460 determines that the allocation change has succeeded and proceeds to step S53. On the other hand, if the allocation change execution unit 460 receives an error notification, the allocation change execution unit 460 determines that the allocation change has failed and proceeds to step S54.

In step S54, the allocation change execution unit 460 discards the record of the trial calculation result management table 180 corresponding to the ID of the virtual volume 210 to be processed. In step S55, the allocation change execution unit 460 notifies the service providing side management terminal 4 that an error has occurred in the storage apparatus 2.

In step S53, the allocation change execution unit 460 updates various management tables in response to the completion of the execution of the allocation change plan. First, the allocation change execution unit 460 clears (or deletes) the value of the allocation destination volume ID 1606 of the pool management table 160 for the page that is not allocated to the processing target virtual volume 210 in step S51. Also, the allocation change execution unit 460 stores the ID of the processing target virtual volume 210 in the allocation destination volume ID 1606 of the pool management table 160 for an unallocated page obtained by copying data of a page that is not allocated to the processing target virtual volume 210. To do.

Then, the allocation change execution unit 460 acquires the trial allocation hierarchy 1804 and the predicted performance 1806 from the record of the volume ID 1802 of the trial calculation result management table 180 that matches the ID of the virtual volume 210 to be processed. The assignment change execution unit 460 deletes the record in the trial calculation result management table 180.

Then, the allocation change execution unit 460 sets the content of the trial allocation hierarchy 1804 in the allocation hierarchy 1703 in the volume management table 170 in which the ID of the virtual volume 210 to be processed and the volume ID 1701 match, and sets the average performance 1706 as the predicted performance 1806. Set the value of to complete the update.

As a result of the above processing, the virtual volume 210 that satisfies the trial calculation start condition is set as the virtual volume 210 to be processed, and the allocation change execution unit 460 executes the allocation change plan of the hierarchy generated by the appropriate allocation trial calculation unit 440, to the virtual volume 210. It is possible to release the media 26 of the highest hierarchy that has been excessively allocated. The virtual volume 210 with excessive allocation of computer resources can be configured with a minimum hierarchy that satisfies the required performance.

As described above, the configuration of the virtual volume 210 in which the computer resources are excessively allocated can be changed, the range of the virtual volume 210 that affects the performance can be reduced to the same pool 220, and The influence on the virtual volume 210 can be reduced.

Also, by assigning the computer resources of the higher tier that have been allocated excessively to other virtual volumes 210 in the same pool, performance can be accommodated in the pool 220. Furthermore, by generating surplus computer resources, it is possible to improve the volume aggregation rate in the pool (the number of volumes that can satisfy the SLA).

<Volume generation processing>
FIG. 7 is a flowchart illustrating an example of a volume generation process performed by the volume generation unit 470 of the management server 1. This process is executed when the storage device management unit 130 of the management server 1 receives a request to create a new virtual volume 210. That is, the storage apparatus management unit 130 commands the volume generation unit 470 to generate a new virtual volume 210. The creation of a new virtual volume 210 is based on a command from the IaaS management unit 110 or a command from the service providing side management terminal 4.

Immediately after the virtual volume 210 is generated, media 26 in all layers can be allocated. That is, all the hierarchies may be stored in the field of the assignment hierarchy 1703 of the volume management table 170. Here, an example of setting a tier to be allocated based on the requirements of the virtual volume 210 when the virtual volume 210 is generated is shown.

The volume generation unit 470 receives the volume generation request and starts processing (S61). The requirements of the virtual volume 210 to be generated are input from the IaaS management unit 110 or the service providing side management terminal 4. The volume generation unit 470 receives requirements for the virtual volume 210 to be generated (S62).

The volume generation unit 470 acquires the required performance and capacity of the virtual volume 210 to be generated (S63). As the required performance and capacity, values received from the IaaS management unit 110 or the like are used. Next, the volume generation unit 470 generates a new virtual volume 210 based on the accepted requirements (S64).
The volume generation unit 470 acquires a record that satisfies the acquired required performance value and satisfies the required performance 4101 from the tier allocation initial value management table 410. When there are a plurality of records satisfying the required performance 4101, the volume generation unit 470 selects the one having the minimum required performance 4101 value. The volume generation unit 470 acquires the hierarchy allocation order 4102 from the acquired record.

The volume generation unit 470 refers to the hierarchy allocation management table 420 and acquires the allocation hierarchy 4202 from the record that matches the acquired hierarchy allocation order 4102. Next, the volume generation unit 470 instructs the storage management unit 230 to generate a new virtual volume 210 as a tier to which the acquired allocation tier 4202 is assigned to the virtual volume 210.

The storage management unit 230 of the storage device 2 generates a new virtual volume 210 using the allocation hierarchy 4202 based on a command from the management server 1. When the generation of the virtual volume 210 is completed, the storage management unit 230 transmits a completion notification to the management server 1. Further, the storage management unit 230 transmits an error when the creation of the virtual volume 210 fails.

If the notification received from the storage apparatus 2 is a completion notification, the volume generation unit 470 of the management server 1 proceeds to step S66. On the other hand, if the notification received from the storage apparatus 2 is an error, the volume generation unit 470 of the management server 1 proceeds to step S68 and notifies the IaaS management unit 110 or the like of the error.

In step S 66, the volume generation unit 470 adds information on the new virtual volume 210 to the volume management table 170 based on the completion notification from the storage management unit 230. The storage management unit 230 assigns a volume ID and a pool ID to the generated virtual volume 210 and transmits it to the management server 1 in a completion notification.

The volume generation unit 470 sets the volume ID and pool ID to the volume ID 1701 and pool ID 1704 of the volume management table 170 from the completion notification. The volume generation unit 470 sets the values received from the IaaS management unit 110 and the like for the capacity and required performance of the virtual volume 210 in the required performance 1705 and capacity 1702 of the volume management table 170. Then, the volume generation unit 470 sets the value of the allocation hierarchy 4202 of the hierarchy allocation management table 420 to the allocation hierarchy 1703 of the volume management table 170. Further, the volume generation unit 470 sets “0” to the average performance 1706 and adds a new record to the volume management table 170.

Note that the volume generation unit 470 also adds a new record to the pool management table 160 based on the information received from the storage management unit 230 in the same manner. Then, the volume generation unit 470 performs a volume generation completion notification to the IaaS management unit 110 and ends the process.

<Summary>
In the first embodiment, the management server 1 classifies in advance the performance of the computer resources (media 26) constituting the pool 220 and the virtual volume 210 of the storage apparatus 2, and sets the order according to the performance. In this embodiment, this classification is set as a hierarchy according to performance. Then, the classification corresponding to the first rank with the highest performance is the highest hierarchy, and the SSD 27 is included in this embodiment. Next, the classification corresponding to the second highest ranking is the second hierarchy, and includes SAS 28 in this embodiment. The classification corresponding to the third rank with the lowest performance is the lowest hierarchy, and includes SATA 29 in this embodiment. As described above, the classification of the hierarchy may be performed based on the structure of the medium 26.

The excess resource control unit 430 sets the processing target virtual volume 210 for the volume 210 that satisfies the trial calculation start condition set in the volume listing condition 1503 of the start condition definition table 150. In the example of the average performance with respect to the required performance in the trial calculation start condition, a virtual volume 210 that exceeds a predetermined value (or a predetermined ratio) and whose average performance 1706 exceeds the required performance 1705 is defined as a virtual volume 210 to be processed. Alternatively, when the trial calculation start condition is a predetermined pool ID 1704, the virtual volume 210 belonging to the pool ID 1704 is set as a processing target.

The appropriate allocation trial calculation unit 440 calculates the predicted performance 1806 by deleting the upper hierarchy, and generates an allocation change plan for the hierarchy by sequentially deleting the upper hierarchy while the predicted performance 1806 satisfies the required performance 1705. The calculation of the predicted performance 1806 is repeated.

Then, the appropriate allocation trial calculation unit 440 generates a tier allocation change plan in which the predicted performance 1806 satisfies the required performance and the number of tiers is minimized. Furthermore, the execution determination unit 450 calculates the appropriate allocation when the ratio of the capacity of the page whose allocation is changed does not exceed a predetermined threshold with respect to the entire capacity of the virtual volume 210 to be processed. The unit 440 changes the hierarchy of computer resources allocated to the virtual volume 210.

Thus, if the capacity of the page that is the target of the tier allocation change is equal to or greater than a predetermined ratio (threshold) with respect to the entire capacity of the processing target virtual volume 210, the allocation change is prohibited. As a result, it is possible to change the tier assignment while suppressing the pages where I / O occurs, and thus it is possible to prevent the performance related to reading and writing of the pool 220 from being deteriorated.

In the first embodiment, an example is shown in which the required performance of each virtual volume 210 is set based on the SLA. However, the present invention is not limited to this. For example, the service providing management terminal 4 or the IaaS management unit 110 may set the required performance of each virtual volume 210.

In the first embodiment, the example in which the assignment destination of the page to be moved is determined according to the access frequency of the page is shown, but the present invention is not limited to this. For example, data may be sequentially copied from the first page at the movement source to the last page at the movement destination.

In the first embodiment, an example in which the virtual volume 210 is allocated to the virtual computer 310 has been described, but the allocation destination of the virtual volume 210 may be a physical computer.

16 to 18 show a second embodiment of the present invention. In the first embodiment, the example in which the tier allocation change plan is executed when the ratio of the capacity to be changed with respect to the entire capacity of the virtual volume 210 does not exceed the threshold is shown. In the second embodiment, an example of executing the allocation change plan of the tier of the virtual volume 210 to be processed according to the type of instance managed by the IaaS management unit 110 is shown.

In the second embodiment, the instance management table 510 and the type management table 520 are stored in the memory 12 of the management server 1, and a part of the execution determination process is changed. Other configurations are the same as those of the first embodiment.

FIG. 17 is a diagram illustrating an example of the instance management table 510. The instance management table 510 is information that defines the relationship between the virtual volume 210 managed by the storage device management unit 130 and the instance managed by the IaaS management unit 110.

The IaaS management unit 110 sets the relationship between the instance and the virtual volume 210 based on the notification from the storage device 2 or the storage device management unit 130. The instance management table 510 includes an instance ID 511 for storing the identification of the instance managed by the IaaS management unit 110, an instance type 512 for storing the instance type managed by the IaaS management unit 110, and a virtual volume assigned to the instance. The volume ID 513 for storing the identifier 210 is included in one record.

FIG. 18 is a diagram illustrating an example of the type management table 520. The type management table 520 is a table in which whether or not to permit the configuration change of the virtual volume 210 for each instance type managed by the IaaS management unit 110 is set.

The type management table 520 includes a type 521 that stores the type of instance managed by the IaaS management unit 110, and an application flag 522 that sets whether or not the configuration of the virtual volume 210 can be changed for each type 512.

The type 521 corresponds to the type 1201 of the instance type management table 120. If the application flag 522 is “Yes”, the configuration change of the virtual volume 210 is permitted, and if “No”, the configuration change of the virtual volume 210 is prohibited.

FIG. 16 is a flowchart illustrating an example of execution determination processing according to the second embodiment of the present invention. This process is executed in place of the process of FIG. 5 of the first embodiment.

Step S31 is the same as step S21 in FIG. 5, and the execution determination unit 450 refers to the trial calculation result management table 180 to determine the content of the record in which the ID of the processing target virtual volume 210 matches the volume ID 1802. Obtained as a trial calculation result.

In step S32, the execution determination unit 450 acquires the requirements set by the IaaS management unit 110. In other words, the execution determination unit 450 selects a record corresponding to the ID of the processing target virtual volume 210 with reference to the instance management table 510, and the instance type 512 in which the processing target virtual volume 210 is mounted from the record. To get. Then, the execution determination unit 450 reads the application flag 522 from the acquired instance type 512 with reference to the type management table 520.

In step S33, if the read application flag 522 is “Yes”, the execution determination unit 450 determines that execution of the hierarchy allocation change plan is permitted, and the process proceeds to step S34 to “YES” to the excess resource control unit 430. Respond. On the other hand, if the application flag 522 is “No”, it is determined that the execution of the tier allocation change plan is prohibited, and the process proceeds to step S35, where “NO” is returned to the excess resource control unit 430.

As described above, according to the second embodiment, in addition to the effects of the first embodiment, whether or not to execute the tier allocation change plan for the virtual volume 210 to be processed depends on the type of the instance of the IaaS management unit 110. Can be controlled. That is, when the IaaS management unit 110 or the virtual machine management unit 140 permits a tier assignment change plan in the virtual computer 310 to which the virtual volume 210 is assigned, the tier assignment change plan can be implemented.

19 and 20 show a third embodiment of the present invention. In the third embodiment, an example in which it is determined whether or not to execute the allocation change plan for the tier of the virtual volume 210 to be processed according to the status of the pool 220 will be described.

In the third embodiment, a part of the execution determination process is changed using the volume management table 170 shown in FIG. 20 instead of FIG. 10 of the first embodiment. Other configurations are the same as those of the first embodiment.

FIG. 20 is a diagram illustrating an example of the volume management table 170 according to the third embodiment. The volume management table 170 of the third embodiment is obtained by adding a performance improvement flag 1707 to the volume management table 170 shown in FIG. 10 of the first embodiment, and other configurations are the same as those of the first embodiment.

FIG. 19 is a flowchart showing an example of execution determination processing according to the third embodiment of the present invention. This process is executed in place of the process of FIG. 5 of the first embodiment.

Step S41 is the same as step S21 shown in FIG. 5 of the first embodiment, and the execution determination unit 450 refers to the trial calculation result management table 180 and the ID of the virtual volume 210 to be processed is the volume ID 1802. Get the contents of matching records as trial calculation results.

In step S42, the execution determination unit 450 acquires the pool ID 1704 of the record that matches the ID of the virtual volume 210 to be processed from the volume management table 170. Next, the execution determination unit 450 acquires a record including the acquired pool ID 1704.

In step S 43, if the performance determination flag 1707 is “Yes” in the record acquired by the execution determination unit 450, “Yes” is excessive in order to release the top-level tier page from the processing target virtual volume 210. In response to the resource control unit 430, the execution of the hierarchy allocation change plan is permitted. On the other hand, if the performance improvement flag 1707 is not “Yes” in the acquired record, it is not necessary to release the upper tier from the virtual volume 210 to be processed, so “NO” is returned to the excess resource control unit 430. Prohibit execution of hierarchy assignment change plan.

As described above, in the third embodiment, in addition to the effects of the first embodiment, there is a virtual volume 210 whose performance improvement flag 1707 is “Yes” in the same pool 220 as the virtual volume 210 to be processed. In this case, it is possible to release a higher-level page by executing a hierarchy allocation change plan. As a result, it becomes possible to accommodate the upper tier pages that become redundant in the virtual volume 210 whose performance improvement flag 1707 is “Yes”.

<Supplement>
In addition, this invention is not limited to each above-mentioned Example, Various modifications are included. For example, the above-described embodiments are described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. In addition, any of the additions, deletions, or substitutions of other configurations can be applied to a part of the configuration of each embodiment, either alone or in combination.

In addition, each of the above-described configurations, functions, processing units, processing means, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. In addition, each of the above-described configurations, functions, and the like may be realized by software by the processor interpreting and executing a program that realizes each function. Information such as programs, tables, and files that realize each function can be stored in a memory, a hard disk, a recording device such as an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.

Also, the control lines and information lines indicate what is considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. Actually, it may be considered that almost all the components are connected to each other.

Claims (15)

A storage device that has a plurality of types of media having different performances and provides the computer with a volume in which one or more areas of the media are allocated to the computer;
A computer system including a processor and a memory and allocating the volume to the computer;
The management computer is
Managing a plurality of types of media according to a rank of performance from a higher layer to a lower layer, and managing one or more layers among the layers according to a required performance set in the computer A storage device manager assigned to the volume;
A computer management unit for allocating the volume to the computer;
For the volume, an excess resource control unit that calculates an allocation change plan assuming that the highest tier of the tiers is deleted, and calculates a predicted value of the performance of the volume to which the allocation change plan is applied;
A computer system characterized by comprising: The computer system according to claim 1,
The excess resource control unit includes:
The calculated predicted performance value of the volume is compared with the required performance preset for the volume, and if the predicted performance value satisfies the required performance, an allocation change plan is executed. A featured computer system. The computer system according to claim 1,
The excess resource control unit includes:
A computer system, wherein a predicted value of the performance of the volume is calculated from an average performance for each area of the media. The computer system according to claim 1,
The excess resource control unit includes:
A computer system, wherein a predicted value of the performance of the volume is calculated from an I / O density distribution of the volume. The computer system according to claim 1,
The excess resource control unit includes:
The calculated predicted performance value of the volume is compared with the required performance set in advance for the volume. If the predicted performance value satisfies the required performance, A computer system characterized by further calculating an allocation change plan on the assumption that the highest hierarchy is deleted. The computer system according to claim 2,
The excess resource control unit includes:
In the allocation change plan, the allocation change plan is executed when a ratio between the capacity of the medium to be deleted from the highest hierarchy and the capacity of the volume does not exceed a preset threshold value. . The computer system according to claim 2,
The excess resource control unit includes:
The computer system, wherein the computer management unit executes the allocation change plan when the computer to which the volume is allocated permits the allocation change plan. The computer system according to claim 2,
The volume is
Belongs to a pool containing one or more volumes,
The excess resource control unit includes:
A computer system, wherein the allocation change plan is executed when a pool to which the volume belongs permits the allocation change plan. The computer system according to claim 2,
The volume is
Belongs to a pool containing one or more volumes,
The excess resource control unit includes:
In the pool to which the volume belongs, a computer system characterized by executing the allocation change plan when other volumes are set to improve performance. The computer system according to claim 1,
The excess resource control unit includes:
Selecting a volume satisfying a preset trial calculation start condition as a volume to be processed, and calculating an allocation change plan on the assumption that the highest hierarchy among the hierarchies is deleted for the volume to be processed Computer system to do. A management computer that has a plurality of types of media having different performances, is connected to a storage device that provides the computer with one or more areas of the media allocated to the computer, and allocates the volume to the computer. A volume allocation control method for controlling allocation of a volume,
The management computer manages a hierarchy in which the plurality of types of media are set in advance from an upper hierarchy to a lower hierarchy according to a performance order, and one of the hierarchies is set in accordance with a required performance preset in the computer. A first step of assigning the above hierarchy to the volume;
A second step in which the management computer assigns the volume to the computer;
A third step in which the management computer calculates an allocation change plan on the assumption that the highest tier of the tiers has been deleted for the volume, and calculates a predicted value of the performance of the volume to which the allocation change plan is applied When,
A volume allocation control method comprising: The volume allocation control method according to claim 11,
The third step includes
The calculated predicted performance value of the volume is compared with the required performance preset for the volume, and if the predicted performance value satisfies the required performance, an allocation change plan is executed. A characteristic volume allocation control method. The volume allocation control method according to claim 11,
The third step includes
A volume allocation control method, wherein a predicted value of the volume performance is calculated from an average performance for each area of the media. The volume allocation control method according to claim 11,
The third step includes
A volume allocation control method, wherein a predicted value of the volume performance is calculated from an I / O density distribution of the volume. The volume allocation control method according to claim 11,
The third step includes
The calculated predicted performance value of the volume is compared with the required performance set in advance for the volume. If the predicted performance value satisfies the required performance, A volume allocation control method characterized by further calculating an allocation change plan assuming that the highest hierarchy is deleted.

Images