WO2017163322A1

WO2017163322A1 - Management computer, and management method for computer system

Info

Publication number: WO2017163322A1
Application number: PCT/JP2016/059095
Authority: WO
Inventors: 金子　聡
Original assignee: 株式会社日立製作所
Priority date: 2016-03-23
Filing date: 2016-03-23
Publication date: 2017-09-28

Abstract

The present invention makes it possible to prevent degradation of the I/O performance of a storage device due to a management operation that is performed on a logical storage area and that is associated with an object management operation. According to the present invention, a processor is configured to: store, in a memory, policy information indicating association of logical storage area performance information, which indicates the performance of logical storage areas, with computer resources; acquire object management operation information indicating an object management operation, which is a management operation on a target object; and, on the basis of configuration information, the policy information, and the object management operation information, generate logical storage area management operation information indicating a logical storage area management operation, which is a management operation on a target logical storage area for storing the target object.

Description

Management computer and computer system management method

The present invention relates to a management computer.

In the operation management of computer systems, it is important to consider the characteristics and usage of storage media when allocating storage areas of computer systems to host computers so that resources can be used effectively.

Japanese Patent Application Laid-Open No. 2004-133867 discloses a technique in which a storage apparatus allocates data to an appropriate storage medium based on an access status to data in order to support management of a storage area of a storage apparatus to be allocated to a host computer. .

JP 2007-066259 A

In a computer system, data of an application or a virtual host computer (hereinafter referred to as a VM) is generally stored in a logical storage area called a volume of a storage device. For this reason, an operation has been performed in which a volume is created when a VM is created, and the volume is deleted when the VM is deleted. Therefore, in recent years, a technique for changing the configuration of a storage apparatus directly from a host computer or a host management computer has appeared in order to simplify operation management such as construction or destruction of an application or VM. With this technique, a volume is automatically created in conjunction with a host management operation such as VM creation by the host management computer. Thereby, the storage management work when managing the host computer is greatly reduced.

However, on the other hand, as the amount of data increases, the number of VMs in a computer system has also increased rapidly, and there are cases where a large number of VMs are operated simultaneously. As a result, the I / O load on the storage apparatus increases rapidly, and the operation rate of internal components such as the storage area and CPU inside the storage apparatus increases, resulting in a problem that the I / O performance from the host computer deteriorates. It can happen.

In Patent Document 1, data with a large amount of I / O from the host is automatically relocated to a high-performance medium. However, if I / O to the storage device has not occurred for a certain period, appropriate resource allocation is performed. Cannot be realized.

In order to solve the above problems, a management computer according to an aspect of the present invention includes a memory and a processor connected to the memory, the host computer, and the storage device. The processor associates a computer resource in the storage apparatus, a logical storage area provided by the storage apparatus using the computer resource, and an object stored in the logical storage area and executed by the host computer. Storing configuration information in the memory, storing policy information in the memory indicating the association between the logical storage area performance information indicating the performance of the logical storage area and the computer resource, and managing the target object A logical storage area that is a management operation of a target logical storage area that acquires object management operation information indicating a certain object management operation and stores the target object based on the configuration information, the policy information, and the object management operation information Generate logical storage management operation information that indicates management operations. To have.

It is possible to prevent the I / O performance of the storage apparatus from deteriorating due to the logical storage area management operation accompanying the object management operation.

1 shows an overall configuration of a computer system according to a first embodiment. The logical configuration of the host computer 5 is shown. A host management computer 35 is shown. The structure of the storage apparatus 4 is shown. The configuration of the storage management computer 1 is shown. A storage volume table 32 is shown. A storage pool table 33 is shown. A storage cache table 34 is shown. A storage CPU table 300 is shown. The storage CPU utilization rate table 2 is shown. The Tiering Policy table 36 is shown. The VM Policy table 31 is shown. The VM management task table 37 is shown. The storage configuration optimization process of Example 1 is shown. The storage configuration optimization process of Example 2 is shown.

Examples will be described with reference to the drawings. It should be noted that the embodiments described below do not limit the invention according to the scope of claims, and that all the elements and combinations thereof described in the embodiments are essential for the solution of the invention. Not exclusively.

In the following description, various types of information may be described using the expression “aaa table”, but the various types of information may be expressed using a data structure other than a table. In order to show that it does not depend on the data structure, the “aaa table” can be called “aaa information”. Further, in the following description, there are cases where the processing is simply described with the storage management computer, the host computer, and the host management computer as the subject, but these processings are performed by a processor (for example, CPU (Central Processing) included in the control device included in the computer. Unit)) indicates that it is being executed. Similarly, when the processing is simply described with the storage device as the subject, it indicates that the controller included in the storage device is executing. Further, at least one of the control device and the controller may be the processor itself, or may include a hardware circuit that performs a part or all of the processing performed by the control device or the controller. The program may be installed in each computer or storage device from a program source. The program source may be, for example, a program distribution server or a storage medium.

In the following description, an ID is used as element identification information, but other types of identification information may be used instead of or in addition thereto.

FIG. 1 shows the overall configuration of the computer system of the first embodiment.

The computer system of this embodiment has a host computer 5 and a storage device 4. The host computer 5 and the storage device 4 are connected via a data network 15. The computer system further includes a host management computer 35 and a storage management computer 1. The host computer 5, the host management computer 35, the storage management computer 1, and the storage device 4 are connected via the management network 16.

The data network 15 is, for example, a SAN (Storage Area Network), but may be an IP (Internet Protocol) network or a data communication network other than these. The management network 16 is, for example, an IP network, but may be a SAN or a data communication network other than these. Further, the data network 15 and the management network 16 may be the same network. Further, the host computer 5, the host management computer 35, and the storage management computer 1 may be the same computer.

The host computer 5 includes a storage medium group 7, a control device such as a CPU (Central Processing Unit) 8, a storage device such as a Memory 9, a M-I / F (Management-Interface) 10, and a CI / F (Communication-Interface). 11 and I / O-I / F (I / O-Interface) 12. The host computer 5 may have an input / output device (keyboard, display device, etc.). The CPU 8 executes a program stored in the Memory 9. Hereinafter, what is called a CPU executes a program stored in a Memory connected to the CPU. The MI / F 10 is an interface with the management network 16 and transmits / receives data and control commands to / from the storage apparatus 4, the host management computer 35, and the storage management computer 1. The C-I / F 11 is an interface with the data network 15 and transmits / receives data and control commands to / from the storage apparatus 4. The I / O-I / F 12 is an interface with the storage medium group 7 and transmits / receives data and control commands to / from the storage medium group 7.

The computer system of the example in this figure includes two host computers 5, two storage devices 4, one host management computer 35, and one storage management computer 1, but is not limited to this configuration. . The processing of the host management computer 35 may be executed by the host computer 5, and the host management computer 35 may be read as the host computer 5 in the following description. Further, the data network 15 may include a switch such as an FC (Fibre Channel) switch, and the switch is connected to each of the CI / F 11 of the host computer 5 and the DI / F 51 of the storage apparatus 4. Data and control commands may be sent and received between them.

FIG. 2 shows the logical configuration of the host computer 5.

The host computer 5 includes a hypervisor (hereinafter referred to as “HV”) 13 that can logically generate the VM 6 and execute the VM 6. The HV 13 can control a plurality of VMs 6 at a time. Each of the plurality of VMs 6 can execute an application as if it were a stand-alone physical computer.

FIG. 3 shows the host management computer 35.

The host management computer 35 includes a CPU 37, a display device 38 (display unit) such as an LCD (Liquid Crystal Display), a Memory 39, and an I / F 36. The host management computer 35 may have an input device (such as a keyboard). The I / F 36 is an interface with the management network 16 and transmits / receives data and control commands to / from the storage apparatus 4, the host computer 5, and the storage management computer 1. In the memory 39, a host configuration management program 40 and a setting program 41 are stored. The host configuration management program 40 is a program for managing configuration information of each host computer 5 such as the VM 6. The host configuration management program 40 communicates with a host information acquisition program 25 of the storage management computer 1 described later, and transmits / receives various types of information. The setting program 41 is a program for requesting various settings of the VM 6 of the host computer 5 and various settings of the storage. The various settings of the VM 6 are, for example, an operation for creating a VM (Create VM) and an operation for starting a VM (PowerOnVM). The various storage settings are, for example, creation and deletion of a volume for storing VM 6 data. The setting program 41 communicates with a storage configuration optimization program 30 of the storage management computer 1 to be described later, transmits / receives various information, and performs various settings.

FIG. 4 shows the configuration of the storage apparatus 4.

The storage apparatus 4 has a disk controller 55 and a physical device 52. The disk controller 55 includes a CPU 47, a Memory 48, an MI / F 45, a storage port HI / F (Host-Interface) 46, and a DI / F (Disk-Interface) 51. The MI / F 45 is an interface with the management network 16 and transmits / receives data and control commands to / from the host computer 5, the host management computer 35, and the storage management computer 1. The HI / F 46 is an interface with the data network 15 and transmits / receives data and control commands to / from the host computer 5. The DI / F 51 transmits / receives data and control commands to / from the physical device 52.

The physical device 52 includes a plurality of types of physical storage media 53 such as an HDD (Hard Disk Drive) and a Flash Memory Drive. The storage apparatus 4 creates a logical volume 54 from the storage areas of a plurality of physical storage media 53 in the physical device 52. A plurality of logical volumes 54 may be created in the storage device 4.

Further, the logical volume 54 may be a Thin Provisioning volume which is a format capable of dynamically expanding the capacity. Thin provisioning is a technology that allocates a partial area of a physical storage area (hereinafter referred to as a page) to a logical volume, dynamically expands the storage area, and effectively uses the storage area. The logical volume 54 provides a page for assignment to the Thin Provisioning volume. When the I / O from the host computer 5 is received, the storage device 4 dynamically expands the capacity of the Thin Provisioning volume by allocating pages from the logical volume 54 included in the pool to the Thin Provisioning volume. Can do. In addition, regarding a Thin Provisioning volume, after a certain page is allocated to the Thin Provisioning volume, it is dynamically changed to another page with different response performance and reliability according to the access status to the Thin Provisioning volume. Is possible. Such a Thin Provisioning volume is hereinafter referred to as a Dynamic Thin Provisioning volume. The logical volume 54 may be a dynamic thin provisioning volume.

In the Memory 48, a storage configuration management program 49 and a storage setting program 50 are stored. The storage configuration management program 49 is a program that manages the configuration information of the storage apparatus 4. In addition, the storage configuration management program 49 communicates with a storage information acquisition program 27 of the storage management computer 1 described later, and transmits / receives various information. The storage setting program 50 is a program for executing various settings of the storage device 4. The various settings include a setting for securing a cache area for temporarily storing data to be read / written for the logical volume 54, a setting for securing a processor for performing read / write processing for the logical volume 54, etc. Settings for improving the access performance to the computer 5 are included. The storage setting program 50 performs various settings by communicating with a storage configuration optimization program 30 of the storage management computer 1 described later.
The Memory 48 may also include a cache area for temporarily storing logical volume data in order to improve the I / O response speed. At this time, the cache area that can be used for each arbitrary logical volume and pool may be limited.

FIG. 5 shows the configuration of the storage management computer 1.

The storage management computer 1 includes a CPU 21, a display device 22 (display unit) such as an LCD, a Memory 24, and an I / F 23. Further, an input device (keyboard or the like) may be included. The I / F 23 is an interface with the management network 16 and transmits / receives data and control commands to / from the storage apparatus 4, the host computer 5, and the host management computer 35. In the Memory 24, programs including a host information acquisition program 25, a storage information acquisition program 27, and a storage configuration optimization program 30 are stored. Further, the Memory 24 includes a storage volume table 32, a storage pool table 33, a storage cache table 34, a storage CPU table 300, a storage CPU utilization rate table 2, a Tiering Policy table 36, a VM Policy table 31, and a VM. An information table including a management task table 37 is stored.

For convenience, this computer is referred to as “storage management computer 1”, but the management target of this computer is not limited to the storage device 4 as described in the present embodiment.

FIG. 6 shows the storage volume table 32.

This table is a table for managing information on the logical volume 54 of the storage apparatus 4. The storage volume table 32 includes a VOL ID 70 indicating a logical volume, a Storage ID 71 indicating a storage device that provides the logical volume, a VOL type 72 indicating the type of the logical volume, and a Capacity 73 indicating the total capacity of the logical volume. And a Used Capacity 74 indicating the used capacity of the logical volume, a Target Port ID 75 indicating the target port of the storage device to which the logical volume is allocated, and a device having an initiator port to which the logical volume is allocated. Initiator ID 76, VM ID 77 indicating the VM to which the logical volume is assigned, and the event to which the logical volume is assigned. Initiator Port ID 78 indicating the creator port, Source Resource ID 79 indicating the resource from which the logical volume is configured, Drive Type 80 indicating the type of the physical storage medium of the logical volume, and Read I / I to the logical volume Read rate 81 indicating the O ratio, Random rate 82 indicating the Random I / O ratio to the logical volume, VM Policy ID 83 indicating the VM Policy of the VM in which data is stored in the logical volume, and data of the logical volume Cache ID84 representing Memory area that can be stored as cache area, and Read Hi representing Read cache hit rate of the logical volume Has a Rate85, the Write Hit Rate86 representing the Write cache hit rate of the logical volume, the.

In this embodiment, when the VOL type 72 indicates “Multi Tier”, this indicates that the volume is a Thin Provisioning volume and is arranged in a plurality of Tiers. When the VOL type 72 indicates “Thin”, this indicates that the volume is a Thin Provisioning volume and is placed in one Tier. When the VOL type 72 indicates “Thick”, this indicates that the volume is a normal logical volume. When the VOL type 72 is “Multi Tier” or “Thin”, the Source Resource ID 79 is an ID of a pool managed by the storage pool table 33 described later. On the other hand, when the VOL type 72 is “Thick”, the ID of the physical storage medium of the storage device managed separately is stored in the Source Resource ID 79.

The logical volume in which the information of the physical storage medium is stored in the Source Resource ID 79 may be a volume directly provided to the host computer as a storage area, or is managed as a pool volume in the storage pool table 33 described later, and Thin It may be provided in the Provisioning volume.

When the VOL type 72 is “Multi Tier”, the Drive Type 80 can store “mix” or the like as information indicating that the types of physical storage media constituting the logical volume are mixed, for example. When the VOL type 72 is “Thin” or “Thick”, for example, “SSD” or “SATA” can be stored in the Drive Type 80 as information indicating the type of the physical storage medium constituting the logical volume.

The VM Policy ID 83 indicates a relationship with the VM Policy ID 140 of the VM Policy table 31 described later. Cache ID 84 indicates a relationship with Cache ID 100 of the storage cache table 34 to be described later.

Also, if the logical volume is not assigned to Target Port or Initiator Port, Target Port ID 75, Initiator ID 76, VM ID 77, Initiator Port ID 78, Read rate 81, Random rate 82, CPU ID 84, CPU ID 83, In addition, “null” is stored in Read Hit Rate 85 and Write Hit Rate 86. In the first embodiment, the Read rate 81 and the Random rate 82 are not necessarily stored.

FIG. 7 shows the storage pool table 33.

This table is a table for managing the information of the pools that make up the Thin Provisioning volume, which is a form of the logical volume of the storage apparatus 4. From the storage pool table 33, the correspondence relationship between the logical volume that is the Thin Provisioning volume and the logical volume that constitutes the pool can be known.

Specifically, the storage pool table 33 has the following information. The storage ID 90 is information for identifying the storage apparatus 4 having a pool. The pool ID 91 is information for identifying the pool. The Drive Type 96 is information indicating the type of physical storage medium constituting the pool. The Page ID 92 is information for identifying pages belonging to the pool. The VOL ID 93 is information for identifying the VOL having the page. The VOL LBA 94 is information indicating the position of the page in the VOL (for example, the top LBA of the page and the LBA at the end of the page). The Pool VOL ID 95 is information for identifying the logical volume 54 having the storage area to which the page is assigned. “N / A (Not / Assigned)” indicates that the page is not allocated to any storage area.

FIG. 8 shows the storage cache table 34.

This table is a table showing the cache area in the Memory 48 of the storage apparatus 4 and its utilization rate.

Specifically, the storage cache table 34 has the following information. Cache ID 100 is information for identifying a cache area. Total Capacity 101 is information indicating the capacity of the cache area. The Write Pending Rate 102 indicates the ratio of data that has not been written to the logical volume in the area where the Write data in the cache area can be stored. The larger the value of Write Pending Rate 102, the more likely that the data write processing to the logical volume and the I / O processing of the host computer will compete and the I / O performance of the host computer will deteriorate. The utility 103 is information indicating the usage rate of the cache area. The larger the value of the utility 103, the higher the possibility that a cache miss will occur, and the higher the possibility that the I / O performance of the host computer will deteriorate. For example, if the value of the initialization 103 of the cache area is large and a cache miss occurs in which the storage apparatus 4 cannot secure a cache at the time of a write request, the storage apparatus 4 writes the write data to the disk and the response time becomes long. Further, for example, when the cache area is shared by a plurality of volumes, the value of the initialization 103 is large, and a lot of data of a specific volume is cached, there is a possibility that a cache miss of data of another volume will occur. Get higher.

FIG. 9 shows the storage CPU table 300.

This table is a table representing the usage rate of the CPU 47 of the storage apparatus 4.

Specifically, the storage CPU table 300 has the following information. The CPU ID 120 is information for identifying the CPU. The utility 121 is information indicating the usage rate of the CPU. The larger the value of the utility 121, the higher the possibility that the processing performance of the CPU will decrease and the possibility that the I / O performance of the host computer will deteriorate.

FIG. 10 shows the storage CPU utilization rate table 2.

This table is a table showing the utilization rate for each logical volume of the CPU 47 of the storage apparatus 4.

Specifically, the storage CPU utilization rate table 2 has the following information. The CPU ID 130 is information for identifying the CPU. The Vol ID 131 is information for identifying a logical volume controlled by the CPU. The utility 132 is information indicating a ratio occupied by the logical volume in the usage rate of the CPU.

FIG. 11 shows the Tiering Policy table 36.

This table is a table representing the Tiering Policy set in the Dynamic Thin Provisioning volume of the storage apparatus 4. “Tiering Policy” is information that defines rules for the dynamic thin provisioning volume to assign and change pages.

Specifically, the Tiering Policy table 36 has the following information. The Tiering Policy ID 110 is information for identifying the Tiering Policy. Tier 1 rate 111 indicates the ratio of the capacity of the storage area of Tier 1 that can be used by the volume among the capacity of the dynamic thin provisioning volume in which the Tiering Policy is set. For example, if the total capacity of the volume indicated in capacity 73 of the storage volume table 32 is 100 GB and Tier 1 rate 111 is 30%, the capacity of the storage area of Tier 1 that can use the Dynamic Thin Provisioning volume is 30 GB. .

“Tier1” is information for specifying the type (performance range) of the physical storage medium that constitutes the volume 54, and it may be defined that the Drive Type 80 of the storage volume table 32 is SSD, or other types. It may be associated with the physical storage medium. Similarly, the Tier 2 rate 112 and the Tier 3 rate 113 indicate the ratio of the storage areas of the Tier 2 and Tier 3 that can be used by the Dynamic Thin Provisioning volume in which the Tiering Policy is set with respect to the total capacity of the volume. The physical storage medium may be defined as Tier1, Tier2, and Tier3 in descending order of I / O performance.

As described above, by limiting the proportion of the volume 54 that can be used for each of a plurality of types of physical storage media, it is possible to preferentially allocate a physical storage medium suitable for the use of the volume 54.

Further, each Tier number may indicate the use priority. When data is rearranged, Tiers with the highest usage priority (smaller Tier number) are allocated in order from the data with the largest number of I / Os. Tier pages with higher degrees are assigned. In addition to the example of this figure, the Tiering Policy may be a format that specifically defines a Tier assigned to the I / O number and the range of the I / O number, or any other format. Thereby, the capacity of each Tier can be distributed to a plurality of volumes that use the same pool.

FIG. 12 shows the VM Policy table 31.

This table is a table showing the resource allocation definition related to the volume 54 allocated to the VM 6 of the host computer 5.

Specifically, the VM Policy table 31 has the following information. The VM Policy ID 140 is information for identifying the VM Policy. VM Policy may represent a VM priority that is a priority of the performance of the VM. For example, VM Policy is represented by Gold, Silver, Bronze, etc. in descending order of VM priority. The VolType 141 is information indicating the type of the volume 54 and is the same as the VOL type 72 in the storage volume table 32. The Tiering Policy ID 142 is information for identifying the Tiering Policy of the volume 54. The Cache ID 143 is information for identifying a cache area to be allocated to the volume 54. The CPU ID 144 is information for identifying the CPU in charge of processing the volume 54. The VM ID 145 is information for identifying a VM to which the VM Policy is applied.

When Cache ID 143 is 000, it indicates that the cache area is a shared cache area shared by a plurality of VMs. For example, the shared cache area is shared by VMs with low VM priority. When the Cache ID 143 indicates a specific cache area that is not a shared cache area, for example, a VM with a high VM priority can occupy the cache area. When the CPU ID 144 is null, it indicates that the CPU to be used is not specified. In this case, when a volume is actually created, the CPU usage rate at that time, the number of volumes mapped to the CPU, etc. The CPU automatically determined based on this is used. When the CPU ID 144 indicates a specific CPU, for example, a VM with a high VM priority can occupy the CPU. The VM ID 145 stores one or more VM IDs, which are stored in a comma-separated format, for example.

For example, the VM administrator inputs information of the VM Policy table 31 to the host management computer 35 in advance. The host management computer 35 transmits the input information to the storage management computer 1, and the storage management computer 1 creates the VM Policy table 31.

FIG. 13 shows the VM management task table 37.

This table is a table showing a VM management operation issued by the setting program 41 of the host management computer 35.

Specifically, the VM management task table 37 has the following information. The Task ID 110 is information for identifying the VM management task. OperationType 111 is information indicating the type of VM management operation corresponding to the VM management task. The OperationType 111 is represented by, for example, VM creation (Create VM), VM power ON (PowerOn VM), VM power OFF (PowerOff VM), and the like. The information of OperationType 111 is referred to by a storage configuration optimization program described later. The VM Policy ID 112 is information for identifying the VM Policy. The VM Policy ID 112 is associated with the value of the VM Policy ID 140 in the VM Policy table 31, and indicates the VM Policy applied to the volume 54 created by the VM management operation. The VM ID 113 is information for identifying the VM 6. The VM ID 113 indicates information for identifying a VM that is a target of the VM management operation. The VM ID 113 stores information of one or more VMs, and is stored in a comma-separated format, for example.

The setting program 41 of the host management computer 35 displays the GUI for the VM management operation on the display device 38 or another computer. The host management computer 35 receives a VM management operation including a VM management operation type (OperationType 111), a VM Policy identifier (VM Policy ID 112), and a VM identifier (VM ID 113) from a user such as a host administrator. The setting program 41 creates a storage configuration change API request corresponding to the accepted VM management operation, and transmits the VM management operation and the storage configuration change API request to the storage management computer 1. Further, the host management computer 35 may store the same information as the VM Policy table 31, or may accept an API parameter of a storage configuration change API request input by the user. In this case, the setting program 41 creates a storage configuration change API request that includes VM Policy information corresponding to the VM management operation as an API parameter. The storage configuration change API request may include the same information as one entry in the VM Policy table 31, or may include information on a part of the entry such as the VM Policy ID. The setting program 41 may omit the API parameter of the storage configuration change API request. The setting program 41 may send only the VM management operation to the storage management computer 1 without creating the storage configuration change API request.

Next, details of the operation of the storage management computer 1 will be described.

First, the processing of the host information acquisition program 25 will be described. The host information acquisition program 25 acquires the configuration / performance information of the host computer 5 from the host configuration management program 40 of the host management computer 35 and stores it in the storage volume table 32 or the like. Specifically, the host information acquisition program 25 stores the configuration / performance information of each host computer 5 and the configuration information of the VM 6 in the storage volume table 32. The host information acquisition program 25 may execute these processes periodically or in response to a user operation.

The storage information acquisition program 27 acquires the configuration information of the storage apparatus 4 from the storage configuration management program 49 of the storage apparatus 4, and based on the acquired configuration information, the storage volume table 32, the storage pool table 33, and the storage cache table 34. The storage CPU table 300, the storage CPU utilization rate table 2, and the Tiering Policy table 36 are updated. With respect to the storage volume table 32, the storage information acquisition program 27 stores the acquired configuration information of the storage apparatus 4 in association with the volume identifier (VOL ID 70) stored in the storage volume table 32. Note that the storage information acquisition program 27 may execute these processes periodically or in response to a user operation.

Next, storage configuration optimization processing by the storage configuration optimization program 30 will be described. The storage configuration optimization program 30 receives the VM management operation and the storage configuration change API request, receives the received information, the storage volume table 32, the storage pool table 33, the storage cache table 34, the storage CPU table 300, and the storage CPU utilization rate. Based on the information stored in the table 2 and the tiering policy table 36, a configuration plan of the internal component considering the I / O characteristics of the host computer is generated.

FIG. 14 shows the storage configuration optimization process of the first embodiment.

First, the storage configuration optimization program 30 receives a VM management operation and a storage configuration change API request from the setting program 41 of the host management computer 35 (S11). The VM management operation indicates information in the VM management task table 37 described above. The VM management operation received in S11 is stored in the VM management task table 37 as a VM management task. Thereafter, the storage configuration optimization program 30 performs processing while referring to the VM management task table 37. The storage configuration change API request is a request for a storage configuration change API associated with a VM management operation, such as creation or deletion of a volume, and includes at least a part of information in the VM Policy table 31 as an API parameter. These pieces of information are stored in the VM Policy table 31, and thereafter, the storage configuration optimization program 30 performs processing while referring to the VM Policy table 31. Further, the VM Policy table 31 may be set in advance, and the storage configuration change API request may include a VM Policy ID. Also, a plurality of VM management operations and storage configuration change API requests can be executed simultaneously. The processing at that time will be described in the second embodiment.

Next, the storage configuration optimization program 30 determines whether or not the received VM management operation is that the target VM is created and that the target VM is in the PowerOff state (S12). For example, when the VM management operation includes VM creation and does not include VM activation, the storage configuration optimization program 30 determines Yes in S12. The VM management operations for creating a VM include Create VM, Clone VM, and the like.

If the determination result is YES in S12, the storage configuration optimization program 30 changes the API parameter of the received storage configuration change API request (S13). Specifically, the storage configuration optimization program 30 specifies the VM Policy information indicated in the VM Policy table 31 as an API parameter when the VM is created, because the VM is still in the PowerOff state. By not changing the specified VM Policy information as an API parameter as it is, but temporarily changing it to an API parameter of the same level as the VM Policy with the lowest VM priority, other VMs by securing that VM's resources Minimize adverse effects on

As a specific example, the storage configuration optimization program 30 may specify the Tier with the largest free capacity based on the storage volume table 32, the storage pool table 33, and the Tiering Policy table 36 for the Tiering Policy ID 142, or The API parameter is changed so that the Tier having the lowest utilization rate is designated. Also, the storage configuration optimization program 30 allocates a shared cache area shared by VMs with low VM priority to the target VM based on the VM Policy table 31 for the cache area, or stores the cache area in the storage cache table 34. Based on this, the API parameter is changed so that the cache area to be allocated to the target VM is changed to a write pending rate or a cache area with a low usage rate or usage amount. Further, the storage configuration optimization program 30 selects a CPU with the highest utilization rate based on the storage CPU table 300 for the CPU or a VM volume with a high VM priority based on the VM Policy table 31. The API parameter is changed so as to select a CPU having a large number of persons in charge. In addition, the case where Tiering Policy, a cache area, and a CPU are designated as API parameters has been illustrated so far, but even if the API parameter or the storage configuration change API request is omitted, the storage configuration optimization program 30 is Similarly, API parameters after changing the resource allocation configuration for minimizing the adverse effect on other VMs caused by the PowerOff VM securing resources may be created. In addition, the storage configuration optimization program 30 may lower the VM priority specified by the VM management operation or the storage configuration change API request.

Thereafter, the storage configuration optimization program 30 proceeds to S14.

When the result of the determination is NO in S12, the storage configuration optimization program 30 determines whether or not the received VM management operation is a VM activation (PowerOn VM) (S14).

If the determination result in S14 is YES, the storage configuration optimization program 30 allocates resources for maximizing the performance of the PowerOn VM target VM for API parameters not specified in S11, contrary to S13. (S15). Specific examples of the parameters not specified in S11 include a case where the Cache ID 143 and CPU ID 144 of the VM Policy table 31 are null. As a specific example of resource allocation for maximizing the performance of the VM, the storage configuration optimization program 30 changes the API parameter so that the highest performance Tier is specified for the Tiering Policy ID 142. In addition, the storage configuration optimization program 30 predefines a cache area to be allocated to a VM with a high VM priority for the cache area, and allocates the cache area, or the Write Pending Rate, the usage rate, and the usage amount. Change the API parameter so that the cache area changes to a lower cache area. In addition, the storage configuration optimization program 30 selects a CPU with the lowest utilization rate, or selects a CPU with a small number in charge of a VM volume with a high VM priority, or an assigned CPU. Is changed in advance, and the API parameter is changed so that the CPU is selected. Further, the storage configuration optimization program 30 may increase the VM priority set for the target VM. As a result, the storage management computer 1 can improve the performance of the VM whose performance has been reduced in the PowerOff state.

Thereafter, the storage configuration optimization program 30 proceeds to S18.

If the determination result in S14 is NO, the storage configuration optimization program 30 determines whether or not the received VM management operation is a VM stop (PowerOff VM) (S16).

If the determination result in S16 is YES, the storage configuration optimization program 30 reduces the resources allocated to the target VM (S17). Specifically, similarly to S13, the storage configuration optimization program 30 performs setting similar to that of the VM Policy having the lowest VM priority, so that the VM in the PowerOff state secures resources to another VM. Minimize the negative effects of. Further, the storage configuration optimization program 30 may lower the VM priority set for the target VM.

Thereafter, the storage configuration optimization program 30 proceeds to S18.

When the determination result in S16 is NO, the storage configuration optimization program 30 reflects the storage configuration change API request accepted in S11 or the storage configuration change API reflecting the API parameter changes created in S13, S15, and S17. By transmitting the request to the storage apparatus 4, a storage configuration change process is executed (S18). Then, the storage configuration optimization program 30 ends this flow. In S18, the storage configuration optimization program 30 displays the parameters of the storage configuration change API request on the display device 22 or other computers, and allows the user to select whether or not to execute the storage configuration change processing. If it is selected, S18 may be executed.

Thereafter, the storage configuration optimization program 30 may cause the host computer 5 to execute the VM management operation by transmitting the VM management operation to the host computer 5.

Note that the order of S12, S14, and S16 may be exchanged.

The storage management computer 1 can prevent the I / O performance deterioration of the storage apparatus 4 due to the VM management operation by determining the API parameter of the storage configuration change API request based on the VM management operation. Further, the storage management computer 1 determines the configuration of the internal component allocated to the volume storing the target VM in the storage apparatus 4 based on the influence of the VM management operation on the I / O processing performance of the storage apparatus 4. As a result, it is possible to prevent I / O performance deterioration due to a high load of internal components. In addition, the storage management computer 1 creates a storage configuration change API request and transmits the storage configuration change API request to the storage device 4, so that the storage device 4 can allocate an appropriate resource to the volume storing the target VM. it can. Further, the storage management computer 1 receives the VM management operation and the storage configuration change API request, and changes the storage configuration change API request based on the VM management operation, so that the I / O performance deterioration of the storage apparatus 4 due to the VM management operation Can be prevented. Further, the storage management computer 1 can determine the CPU and cache area to be allocated to the target VM according to the VM management operation, thereby preventing the I / O performance deterioration of the storage apparatus 4. Further, the storage management computer 1 can determine the Tier assigned to the target VM in accordance with the VM management operation, thereby preventing the I / O performance deterioration of the storage apparatus 4.

In addition, the host management computer 35 transmits the VM Policy designated by the VM administrator to the storage management computer 1, so that the storage management computer 1 is appropriate even if the VM administrator does not know the detailed parameters of the storage. A storage configuration change API request can be issued. In addition, since the storage management computer 1 manages the storage apparatus 4 in accordance with the management of the VM, the work of managing the storage apparatus 4 can be reduced.

The host computer 5 may execute objects such as containers and application programs instead of VMs. In this case, the object is stored in the storage apparatus 4, and the host management computer 35 performs an object management operation instead of the VM management operation.

In the following description, the configuration different from the first embodiment will be described in detail, and the detailed description of the same configuration as the first embodiment will be omitted. In the first embodiment, the storage configuration optimization program 30 of the storage management computer 1 generates the storage configuration change plan based on the type of the VM management operation, and executes the storage configuration change. The storage management computer 1 according to the second embodiment stores a storage configuration optimization program 30B instead of the storage configuration optimization program 30. In the second embodiment, when a plurality of VM management operations are requested at the same time, the storage configuration optimization program 30B generates a storage configuration change plan based on the number of simultaneous execution operations, and executes the storage configuration change. In the second embodiment, since only the behavior of the storage configuration optimization program 30B changes, this difference will be described below.

FIG. 15 shows the storage configuration optimization process of the second embodiment.

The storage configuration optimization program 30B executes the same processing as the storage configuration optimization processing of the first embodiment up to S14, and when the determination in S14 is Yes, executes processing different from that of the first embodiment. If the determination in S14 is Yes, the storage configuration optimization program 30B determines whether the number of identical VM management operations is n or more (S19). Here, the storage configuration optimization program 30B may regard the number of VMs stored in the VM ID 113 of the VM management task table 37 as the number of VM management operations, or execute simultaneously from the setting program 41 of the host management computer 35. The number to be used may be explicitly acquired. A condition that a plurality of VM management operations are regarded as the same VM management operation is that the VM Policies of the plurality of VM management operations are the same. All the resources such as the CPU and cache area of the storage apparatus 4 allocated to these VMs are selected by the same VM Policy. The threshold value n for the number of VM management operations may be defined in advance based on the performance of the storage apparatus 4, or may be designated by the user. Further, the setting program 41 of the host management computer 35 may periodically collect the received VM management operations and send them to the storage management computer 1. Further, the storage configuration optimization program 30B may periodically process the received VM management operations. The present invention is not limited to the method for determining n. The storage configuration optimization program 30B performs the processes of S19 and S20 for each VM Policy.

If the determination result in S19 is NO, the storage configuration optimization program 30B proceeds to S13 of the storage configuration optimization process of the first embodiment, and thereafter executes the same processing as the storage configuration optimization processing of the first embodiment. .

If the result of the determination in S19 is YES, the storage configuration optimization program 30B performs a configuration change that strengthens the resources allocated to the VM rather than S15 of the storage configuration optimization process of the first embodiment (S20). That is, the storage configuration optimizing program 30B changes the API parameter so as to temporarily allocate a resource having higher performance than the resource specified in the VM Policy table 31 or a resource having an empty utilization rate. As a result, the storage configuration optimization program 30B can prevent a large number of VMs from being activated at one time, and I / O from those VMs to temporarily increase rapidly, resulting in a performance failure. Note that the storage configuration optimization program 30B can use the resources of the storage apparatus more efficiently by releasing this configuration change after a certain time when the I / O load by the VM is settled. The release timing may be defined in advance by the storage configuration optimization program 30B or may be in a format that can be specified by the user. The present invention is not limited to the method for determining the release timing. Further, the storage configuration optimization program 30 may increase the VM priority of the VM management operation. Further, the storage configuration optimization program 30 may increase the VM priority of the VM management operation more than the increase step in S15.

Thereafter, the storage configuration optimization program 30 proceeds to S18, and thereafter executes the same processing as the storage configuration optimization processing of the first embodiment.

In S19, the storage configuration optimizing program 30B may calculate the impact level of the target VM management operation and may determine Yes when the impact level exceeds a preset impact level threshold. For example, the storage configuration optimization program 30B calculates the degree of influence by adding a value obtained by multiplying the number of each VM management operation by a weight according to the VM Policy. The degree of influence indicates the degree of influence that a plurality of VM management operations have on the I / O processing performance of the storage apparatus 4. As a result, the storage configuration optimization program 30B performs the process of S20 even if the number of VM management operations is small if the VM priority of each VM management operation is high. The influence threshold and the weight of each VM Policy may be defined in advance based on the performance of the storage apparatus 4 or may be designated by the user.

Furthermore, the storage configuration optimization program 30B executes S21 in any of S11-S12, S12-S14, S14-S16, and S16-S18 in the storage configuration optimization processing of the first embodiment. Perform the process.

In S21, the storage configuration optimization program 30B determines whether or not the specified VM management operation is a Clone VM. If the determination result is NO, the process proceeds to the next process of S21 in the storage configuration optimization process of the first embodiment, and the same process as the storage configuration optimization process of the first embodiment is performed thereafter.

If the determination result in S21 is YES, the storage configuration optimization program 30B determines whether the number of identical VM management operations is n or more (S22). The storage configuration optimization program 30B may regard the number of Clone VM operation tasks that have the same VM in the VM management task table 37 as the copy source as the number of VM management operations, or the setting program 41 of the host management computer 35. It is also possible to explicitly acquire the number executed simultaneously. The condition that a plurality of VM management operations are regarded as the same VM management operation is that the copy source VMs are the same. Note that the number n may be defined in advance by the storage configuration optimization program 30B based on the performance of the storage, or may be in a format that can be designated by the user, and the present invention is not limited to the n determination method. The storage configuration optimization program 30B performs the processing of S22 and S23 for each copy source VM.

If the determination result in S22 is NO, the storage configuration optimization program 30B proceeds to S18 of the storage configuration optimization process of the first embodiment, and thereafter executes the same processing as the storage configuration optimization processing of the first embodiment. .

If the determination result in S22 is YES, the storage configuration optimization program 30B changes the processing corresponding to the CloneVM to another storage configuration change API request such as Snapshot in order to respond quickly to the CloneVM (S23). In other words, the storage configuration optimization program 30B implements the Storage API that generates a volume that can access the same data as the original VM data in a short time, instead of the VM data copy that is normally performed in response to the Clone VM. As an example of the Storage API that generates a volume that can access the same data as the original VM data in a short time, there is a snapshot technology of a storage device. The storage configuration optimizing program 30B performs this substitution process, and further asynchronously performs the same data copy as in the normal Clone VM, and cancels the substitution process when the copy is completed. As a result, it is possible to prevent the host I / O performance from being lowered and the VM management operation from taking a long time due to the concentration of data copy processing load.

For example, the storage configuration optimization program 30B creates a first clone destination volume that is a virtual volume for Snapshot, creates a Snapshot of the clone source volume in the first clone destination volume, and provides it to the host computer 5. As a result, the storage apparatus 4 writes the difference between the clone source volume and the snapshot into the first clone destination volume, and writes the update to the first clone destination volume into the first clone destination volume. Refer to data not in the clone destination volume. As a result, the performance of access to the clone source volume and the first clone destination volume is temporarily lowered.

Furthermore, the storage configuration optimization program 30B creates a second clone destination volume for synchronous copy of the clone source volume, and starts synchronous copy from the clone source volume to the second clone destination volume. When the second clone destination volume matches the clone source volume by the synchronous copy, the storage configuration optimization program 30B stops the synchronous copy, reflects the snapshot volume to the second clone destination volume, and is given to the first clone destination volume. By assigning the volume ID to the second clone destination volume, the clone destination volume is switched. Thereby, the storage configuration optimization program 30B provides the host computer 5 with the second clone destination volume of the volume ID. Since the second clone destination volume is a volume having all data entities, the access performance is improved. In this way, the storage configuration optimization program 30B creates a virtual copy of the clone source volume, thereby shortening the response time of the VM management operation of the clone and creating a physical copy of the clone source volume. As a result, the subsequent access performance can be improved.

After this processing, the storage configuration optimization program 30B proceeds to S18, and thereafter executes the same processing as the storage configuration optimization processing of the first embodiment.

In S22, the storage configuration optimization program 30B may calculate the influence level of the target VM management operation, and may determine Yes when the influence degree exceeds a preset influence degree threshold, as in S19. .

If the storage management computer 1 copies VM data to a large number of Clone VMs, it takes a long time to respond to the host management computer 35, and the user of the host management computer 35 can perform the next VM management operation. Can not. By creating a replica that can be accessed in a short time by the storage management computer 1 as in S23, the user of the host management computer 35 performs the next VM management operation in a short time even when a large number of Clone VMs are issued. be able to.

Explain terms. The management computer corresponds to the storage management computer 1, the host management computer 35, and the like. The memory corresponds to the

memories

24 and 39 and the like. The processor corresponds to the

CPUs

21 and 37 and the like. The computer resources correspond to the CPU 47, the memory 48, the HI / F 46, the DI / F 51, the physical device 52, and the like. The storage device corresponds to the memory 48, the physical device 52, and the like. The logical storage area corresponds to a volume or the like. The configuration information corresponds to the storage volume table 32, the storage pool table 33, the storage cache table 34, the storage CPU table 300, the storage CPU utilization rate table 2, and the like. The policy information corresponds to the VM Policy table 31 and the like. The usage priority information corresponds to the Tiering Policy table 36 or the like. The priority corresponds to VM Policy, VM priority, and the like. The object management operation information corresponds to a VM management operation, an entry in the VM management task table 37, and the like. The logical storage area management operation information corresponds to a storage configuration change API request issued by the storage computer. The logical storage area management operation request information corresponds to a storage configuration change API request issued by the host management computer 35. The logical storage area performance information corresponds to VM Policy ID, VM Policy, VM priority, and the like. The storage device restriction information corresponds to Tiering Policy ID, Tiering Policy, and the like. Creating a virtual copy corresponds to a snapshot or the like.

As mentioned above, although embodiment of this invention was described, this is an illustration for description of this invention, Comprising: It is not the meaning which limits the scope of the present invention to the said structure. The present invention can be implemented in various other forms.

1 ... Storage management computer, 4 ... Storage device, 5 ... Host computer, 15 ... Data network, 16 ... Management network, 21 ... CPU, 22 ... Display device, 23 ... I / F, 24 ... Memory, 35 ... Host management computer

Claims

Memory,
A processor connected to the memory, host computer, and storage device;
With
The processor is
Configuration information indicating an association between a computer resource in the storage device, a logical storage area provided by the storage device using the computer resource, and an object stored in the logical storage area and executed by the host computer In the memory,
Storing policy information indicating the association between the logical storage area performance information indicating the performance of the logical storage area and the computer resource in the memory;
Obtain object management operation information indicating the object management operation that is the target object management operation
Based on the configuration information, the policy information, and the object management operation information, logical storage area management operation information indicating a logical storage area management operation that is a management operation of a target logical storage area that stores the target object is generated. ,
Configured as
Management computer.
The processor is configured to cause the storage apparatus to execute the logical storage area management operation by transmitting the logical storage area management operation information to the storage apparatus.
The management computer according to claim 1.
The processor determines a target computer resource for the target logical storage area based on the configuration information, the policy information, and the object management operation information, and allocates the target computer resource to the target logical storage area. Is determined as the logical storage area management operation,
The management computer according to claim 2.
When the object management operation is activation of the target object, the I / O processing performance of the target computer resource is higher than the I / O processing performance of the computer resource allocated to the target logical storage area.
The management computer according to claim 3.
When the object management operation is to stop the target object, the I / O processing performance of the target computer resource is lower than the I / O processing performance of the computer resource allocated to the target logical storage area.
The management computer according to claim 4.
When the object management operation is a new creation of the target object and the object management operation information indicates logical storage area performance information for the target logical storage area, the I / O processing performance of the target computer resource is Lower than the I / O processing performance of the computer resource corresponding to the logical storage area performance information for the target logical storage area;
The management computer according to claim 5.
The processor acquires logical storage area management operation request information indicating a management operation of the target logical storage area;
The processor generates the logical storage area management operation information by changing the logical storage area management operation request information based on the configuration information, the policy information, and the object management operation information.
The management computer according to claim 6.
The object is a virtual machine;
The management computer according to claim 7.
The storage apparatus includes a storage device and a storage processor,
The computer resource is at least one of the storage device and the storage processor.
The management computer according to claim 8.
The policy information includes storage device restriction information indicating a restriction on allocation of a plurality of storage devices in the storage device,
The processor is
Based on the configuration information, the policy information, and the object management operation information, determine target storage device restriction information for the target logical storage area;
Allocating a storage device to the target logical storage area according to the target storage device restriction information is determined as the logical storage area management operation;
Configured as
The management computer according to claim 9.
The object management operation information indicates a plurality of object management operations respectively corresponding to a plurality of target objects,
The processor is
Based on the policy information, the degree of influence indicating the magnitude of the influence of the plurality of object management operations on the I / O of the storage apparatus is calculated,
Determining the logical storage management operation based on the degree of influence;
Configured as
The management computer according to claim 2.
When the plurality of object management operations are to start the plurality of target objects, respectively, and the influence degree is larger than a preset influence degree threshold, the processor stores a plurality of the plurality of target objects, respectively. Determining a plurality of target computer resources for the target logical storage area, and assigning each of the plurality of target computer resources to the plurality of target logical storage areas as the logical storage area management operation,
The I / O processing performance of the plurality of target computer resources is higher than the I / O processing performance of the plurality of computer resources respectively allocated to the plurality of target logical storage areas.
The management computer according to claim 11.
When each of the plurality of object management operations is object duplication and the influence degree is larger than a preset influence degree threshold, the processor stores data in a logical storage area storing the original object of the duplication To create a virtual replica of the logical storage area management operation,
The management computer according to claim 11.
A computer system management method including a host computer and a storage device,
Configuration information indicating an association between a computer resource in the storage device, a logical storage area provided by the storage device using the computer resource, and an object stored in the logical storage area and executed by the host computer In memory,
Storing policy information indicating the association between the logical storage area performance information indicating the performance of the logical storage area and the computer resource in the memory;
Obtain object management operation information indicating the object management operation that is the target object management operation
Based on the configuration information, the policy information, and the object management operation information, logical storage area management operation information indicating a logical storage area management operation that is a management operation of a target logical storage area that stores the target object is generated. ,
To be prepared,
Management method.