JP5597266B2 - Storage system - Google Patents

Description

The present invention relates to a control method and a management method of a storage system accessed by a computer in a computer system, and in particular, a method for allocating a storage area of a nonvolatile memory mounted in a storage system, and a storage area management method About.

With the deregulation of computerization of documents, the digitization of procedures, and the expansion of the Internet business, information handled by corporations has increased explosively. In addition to such information increase, data backup from a disk device to another disk device (Disk-to-Disk Ba)
The amount of data stored in storage devices has increased dramatically due to increasing customer demands such as long-term storage of corporate business activity records (transaction information, e-mails, etc.) due to response to audits and audits. Is invited.

Along with this, in the enterprise information system, while increasing the storage of each department / system, it is required to simplify and improve the management of the complicated IT infrastructure. In particular, there is an increasing expectation for simplifying management of storage devices and optimizing total cost by utilizing optimal storage according to the value of data.

One method for reducing the management cost of a large-scale storage system is a storage virtualization technique as disclosed in Patent Document 1. In Patent Document 1, a first storage is connected to one or more second storages, and a device (hereinafter referred to as a logical device) provided by the second storage to a host or the like via the first storage is passed through the first storage. A storage virtualization technique (hereinafter referred to as an external storage connection method) provided to a host as a logical device of the first storage is disclosed. When the first storage receives an I / O request to the logical device from the host, the access target device corresponds to either the logical device of the second storage or a physical device such as a disk device in the first storage. I / O requests are distributed to the appropriate access destination according to the determination result.

Further, as one method for simplifying the storage system design, there is a volume management technique (hereinafter referred to as an expansion volume) that can dynamically expand the capacity as disclosed in Patent Document 2. In Patent Document 2, when a logical device is defined, a physical storage area (HDD or the like) corresponding to the device capacity requested by the user is not fixedly allocated, but a pool area is provided in the storage, and update access to the logical device is performed. Is performed, a specified amount of physical storage area is dynamically added from the pool area only to the updated portion. As a result, a large-capacity logical device can be introduced with a small physical storage area, and the storage capacity design can be simplified.

On the other hand, as one of the methods for reducing the introduction / operation cost of a large-scale storage system that stores a large amount of data as described above, there is a storage packaging technology that employs a nonvolatile memory such as a flash memory disclosed in Patent Document 3. Conceivable. Patent Document 3 discloses a countermeasure against a write failure triggered by an increase in the number of rewrites, which is a drawback of a flash memory. By using the technique of Patent Document 3, a nonvolatile memory such as a flash memory is mounted instead of or in addition to the HDD, and a storage that stores data written to or read from the host by the nonvolatile memory is conceivable. By storing the host data in the non-volatile memory, it is possible to provide a storage apparatus that can achieve both high-speed I / O processing and low power consumption as compared with a storage apparatus equipped with a disk device. This makes it possible to reduce the size of the introduction system and reduce the system running cost (power consumption). Note that flash memory, which is currently the mainstream nonvolatile memory, is DRAM or S-RAM used as the main memory of computers.
Compared to the above, it has characteristics that affect usability, such as a low speed and a limited number of times of writing. Therefore, it is used as a secondary storage device such as disk device replacement.

JP 2005-11277 A JP 2003-15915 A Patent 3507132

In the first storage shown in Patent Document 1, it is necessary to mount a cache memory for input / output processing of the second storage, which is an external storage. The cache memory temporarily stores data to be read or written during input / output processing to the disk device installed in the storage, and realizes high-speed read and write processing due to a cache hit. Most storage systems It is mounted on. In this specification, a storage area that temporarily stores data, such as a cache memory, is referred to as a “temporary area”, and a storage area that is the final storage destination of data, such as a disk device, is referred to as a “permanent area”. To do.

In order to execute I / O to external storage at high speed, it is necessary to allocate cache memory (temporary area) with a capacity necessary for I / O processing according to the capacity and access characteristics (locality, etc.) of external storage. .

Conventionally, memory devices such as DRAM and SRAM have been used for this cache memory (temporary area), and as a storage medium with completely different characteristics from disk devices (permanent area), such as high cost, high speed, and small capacity, Contributed to speeding up. On the other hand, in order to meet the characteristics required for the temporary area, in the first storage described above, a memory is provided as a temporary area for external storage (and the disk apparatus mounted inside) separately from the mounted disk apparatus (permanent area). I had to install the device.

However, when the first storage having the external storage connection means described above is a storage that uses a nonvolatile memory as a permanent area, it is possible to use the nonvolatile memory as a temporary area of the external storage as in the permanent area. By using nonvolatile memory as both the temporary area and the permanent area, it is not necessary to mount both the disk device and the memory, the hardware is simplified, and the hardware cost can be reduced. However, for this realization, a management technique capable of determining and managing the capacity / location of the temporary area and the permanent area allocated on the nonvolatile memory is required. Further, according to the increase / decrease of the external storage and the change of access characteristics, it is necessary to change to temporary area allocation that satisfies the input / output processing requirements of the external storage.

For next-generation non-volatile memory such as MRAM, which is being studied as a future technology, DRAM
It is planned to achieve high speed comparable to the above and to eliminate write restrictions, and new nonvolatile memory will require a new architecture.

An object of the present invention is to determine the temporary area capacity according to the input / output processing requirements of the external storage and allocate the area in a storage system having external storage connection means and having a nonvolatile memory as a permanent area and an external storage. It is possible to select and manage.

The management server and the storage system respond to an instruction from the user or application program or a configuration change such as storage device increase / decrease. The capacity of the storage area for storing data is determined, and the temporary area is allocated to a specific area of the nonvolatile memory.

Further, the capacity of the temporary area allocated to the nonvolatile memory is increased / decreased in accordance with a change in capacity of a data permanent area such as a disk device or a nonvolatile memory, that is, a storage area for storing host data or a change in access characteristics.

According to the storage system of the present invention, it is possible to determine the temporary area capacity according to the configuration of the virtual storage system, the capacity of the virtualized device and the access characteristics, and to use the nonvolatile memory separately for the temporary area and the permanent area. Become.

It is a figure which shows an example of the hardware constitutions of a computer system. It is a figure which shows an example of the software configuration of a storage. It is a figure which shows an example of logical device management information. It is a figure which shows an example of LU path management information. It is a figure which shows an example of physical device management information. It is a figure which shows an example of external device management information. It is a figure which shows an example of extended volume management information. It is a figure which shows an example of pool volume management information. FIG. 10 is a diagram illustrating an example of a processing flow of a temporary area definition instruction process 243 executed by the management server 110 in the first embodiment. FIG. 10 is a diagram illustrating an example of a processing flow of a temporary area definition process 224 executed in the storage 130 in the first embodiment. 6 is a diagram illustrating an example of a processing flow of physical device definition processing 226 executed in the storage 130 in Embodiment 1. FIG. FIG. 10 is a diagram illustrating an example of a process flow of a device definition instruction process 241 executed by the management server 110 in the first embodiment. FIG. 10 is a diagram illustrating an example of a process flow of a device definition process 222 executed in the storage 130 in the first embodiment. FIG. 10 is a diagram illustrating an example of a processing flow of a temporary area change instruction process 244 executed by the management server 110 in the second embodiment. FIG. 10 is a diagram illustrating an example of a processing flow of a temporary area deallocation process 225 executed in the storage 130 in the second embodiment. FIG. 10 is a diagram illustrating an example of a processing flow of device migration processing 223 executed in the storage 130 in the second embodiment. FIG. 10 is a diagram illustrating an example of a processing flow of physical device definition cancellation processing 227 executed in the storage 130 in the second embodiment. FIG. 10 is a diagram illustrating an example of a process flow of a device migration instruction process 242 executed by the management server 110 in the second embodiment. FIG. 11 is a diagram illustrating an example of a processing flow of an I / O processing process 221 executed in the storage 130 in the first and second embodiments. 3 is a diagram illustrating an example of a configuration of a nonvolatile memory 134 mounted in a storage 130 in Embodiments 1 and 2. FIG. In Example 2, it is a figure which shows an example of the address conversion table which a management server hold | maintains. In Example 2, it is a figure which shows an example of the file collection information which the host computer 100 collects. FIG. 6 is a diagram illustrating an example of a hardware configuration of a computer system according to a modification common to the first and second embodiments.

  The best mode for carrying out the present invention will be described.

As embodiments of the computer system, storage and device control method of the present invention, first to second embodiments will be described.

The first embodiment is a computer system in which a host computer, a first storage, a second storage connected between the host computer and the first storage, and a management server are connected via a network. Storage is equipped with a non-volatile memory as a permanent area, and a storage virtualization function that can virtually integrate other storage devices as its own device,
The management server and the second storage determine the capacity of the temporary area in accordance with a user or application instruction, and secure and allocate the temporary area of the capacity on the nonvolatile memory. In addition, device definition processing and I / O processing according to the present invention will be described.

The second embodiment has the same configuration as the first embodiment, and is allocated according to the capacity and access characteristics of the permanent area on the disk device and the nonvolatile memory mounted in the first storage device or the second storage. Calculate the required temporary space capacity and according to the comparison result with the allocated capacity
Control to increase or decrease the temporary area allocation requirement. A device migration instruction process according to the present invention will also be described.

First, the first embodiment will be described with reference to FIGS. 1 to 13, 19 and 20. FIG. FIG. 1 is a diagram showing an example of a hardware configuration of a computer system to which the first embodiment of the present invention is applied.

The computer system includes one or more host computers 100, a management server 110, a fiber channel switch 120, a storage 130, a management terminal 140, an external storage 1
50a and 150b (collectively referred to as external storage 150). Host 100
The storage 130 and the external storage 150 are connected to the port 121 of the fiber channel switch 120 via ports 107, 131, and 151, respectively. Host 10
0, the storage 130, the external storage 150, and the fiber channel switch 120 are connected to the management server 110 from the interface control units (I / F) 106, 138, 157, and 123 via the IP network 175, and operate on the management server 110. Integrated management by storage management software (not shown). In this embodiment, the storage 130 is connected to the management server 110 via the management terminal 140.
The storage 130 may be directly connected to the IP network.

The host computer 100 is a computer having a CPU 101, a memory 102, and the like. Software such as an operating system and application programs stored in a storage device 103 such as a disk device or a magneto-optical disk device is read into the memory 102, and the CP
The U101 reads out from the memory 102 and executes it to achieve a predetermined function. An input device 104 such as a keyboard and a mouse and an output device such as a display 105 are provided. The input / output device 104 receives input from a host administrator or the like, and the output device outputs information instructed by the CPU 101. In addition to the port 107, the host 100 has an IP network 17
And one or more interface control units 106 for connection to the network.

The management server 110 is also a computer having a CPU 111 and a memory 112. The storage management software stored in the storage device 113 such as a disk device or a magneto-optical disk device is read into the memory 112, and the CPU 111 reads and executes it. Thus, predetermined functions such as operation and maintenance management of the entire computer system are achieved. When the storage management software is executed by the CPU 111, the management server 110 executes the interface control unit 11.
6 through the IP network 175, configuration information, resource utilization, performance monitoring information, failure logs, etc. are collected from each device in the computer system. Then, the management server 110 outputs the collected information to an output device such as the display 115 and presents it to the storage administrator. Further, the management server receives an instruction from the storage administrator via the input device 114 such as a keyboard and a mouse, and transmits the received operation / maintenance instruction to each device via the interface control unit 116.

The storage 130 includes one or more ports 131, one or more control processors 132, one or more memories 133 each connected to the control processor 132, one or more non-volatile memories 134, and one or more The control memory 135, one or more ports 136, one or more disk devices 137 connected to the ports 136, and the interface control unit 138 are connected by an internal connection network.

The control processor 132 identifies an access target device for the input / output request received from the port 131, and the disk device 137 or external storage 150 to which the device corresponds.
Process I / O requests to internal devices. At that time, the control processor 132 specifies an access target device from the port ID and LUN (Logical Unit Number) included in the received input / output request. In this embodiment, the port 131 is a SCSI.
A port corresponding to a fiber channel interface using (Small Computer System Interface) as an upper protocol is assumed.
May be a port corresponding to another network interface for storage connection, such as an IP network interface (for example, iSCSI).

The storage 130 of this embodiment has the following device hierarchy. First, a storage device array having a plurality of disk devices 137 or a plurality of permanent areas of the nonvolatile memory 134 is configured, and this storage device array is managed by the control processor 132 as one physical device. Further, in the control processor 132 of the present invention, a logical device is directly assigned to a physical device mounted in the storage 130 (that is, the control processor 132 associates the physical device with the logical device). The logical device is managed in the storage 130, and the number is managed independently for each storage 130. The logical device has each port 13
1 is associated with the LUN assigned to 1, and is provided to the host computer 100 as a device of the storage 130. That is, the host computer recognizes the logical device of the storage 130, and the host computer 100 accesses the data stored in the storage 130 using the LUN of the port 131 corresponding to the logical device. In this example,
The control processor 132 has an external storage connection method for managing a logical device of the external storage 150 as an external device and virtualizing it as a device of the storage 130. One or more of the external devices captured using this external storage connection are directly associated with the logical devices of the same storage 130 in the same way as physical devices. External devices are also managed independently in each storage 130. Further, in this embodiment, it is also possible to apply an extended volume technique for a logical device, that is, a virtual volume management technique capable of dynamically allocating a physical storage area to an update access part. In that case, a pool volume is allocated to a physical device or an external device, an expansion volume is defined corresponding to the pool volume, and a logical device is allocated to the expansion volume. Here, the expansion volume is a virtual device using the expansion volume technology, and corresponds to the logical device on a one-to-one basis. A pool volume is a virtual device that manages a physical storage area that is dynamically allocated to an expansion volume. In this embodiment, for simplicity, the pool volume corresponds to one physical device or an external device, but one pool volume may be composed of a plurality of physical devices or external devices. In order to realize the device hierarchy as described above, the control processor 132 includes a logical device, an expansion volume, a pool volume, a physical device, a disk device 137, a permanent area of the nonvolatile memory 134, an external device, and a logical device of the external storage 150, respectively. Of managing the correspondence relationship between the devices, converting the access request for the logical device into the access request to the disk device 137, the permanent area of the nonvolatile memory 134, and the logical device of the external storage device 150, and transmitting to the appropriate device I do. As described above, the storage 130 in this embodiment defines one or a plurality of physical devices by combining a plurality of disk devices 137 or nonvolatile memory 136 permanent areas (that is, a plurality of disk devices 137 or a nonvolatile memory). Memory 1
A plurality of 36 permanent areas are collectively associated with one or more physical devices) and one logical device or one pool volume is allocated to one physical device and provided to the host computer 100. However, of course, the individual disk device 137 or the non-volatile memory 1
One area of the 36 permanent areas may be one physical device and one logical device or pool volume.

In addition to the input / output processing for the device, the control processor 132 executes various processes for realizing data linkage between devices, such as data replication and data relocation.

In addition, the control processor 132 transmits configuration information to be presented to the storage administrator to the management terminal 140 connected via the interface control unit 138, and the management terminal 140 sends the configuration information to the management terminal 140.
The maintenance / operation instruction input to is received from the management terminal 140, and the configuration of the storage 130 is changed.

The nonvolatile memory 134 is used by being divided into a temporary area and a permanent area. The temporary area increases the processing speed in response to an access request from the host computer 100.
Data that is frequently read from the host computer is stored in advance, or write data received from the host computer 100 is temporarily stored. When performing write after using the temporary area of the non-volatile memory 134, that is, after the write data received from the host computer 100 is stored in the temporary area of the non-volatile memory 134, the host computer 100 is not actually written to the disk device 137. When the response to the write request is returned, the write data stored in the temporary area of the non-volatile memory 134 is prevented from being lost before being written to the disk device 137. It is desirable to improve availability, such as by duplicating to improve tolerance. On the other hand, the permanent area is the disk device 137.
Like the above, it is used for finally storing the host computer 100 data. Therefore,
In order to increase the bit cost and I / O processing performance, the temporary area is simply duplicated to achieve a small capacity but high speed access, and the permanent area adopts a redundant array configuration such as RAID 5 like the disk device 137. Therefore, it is desirable to provide a storage area with relatively low speed access but large capacity and low cost. In the present invention, for the sake of simplicity, the permanent area of the nonvolatile memory 134 is managed by being assigned an identification number for each small area of a prescribed size, and physical devices are defined in all areas of the permanent area.

FIG. 20 shows an implementation example of the nonvolatile memory 134. Each of the nonvolatile memories 134 includes a plurality of switches 2001 and a plurality of memory packages 2002. The memory package 2002 includes a plurality of nonvolatile memory modules 2004 and the memory modules 2004.
A memory controller 2003 for controlling access to the memory is installed. The memory controller 2003 is coupled to the switch 2001 and other components in the storage 130 via an internal coupling network, and performs memory access and data transfer.

The control memory 135 stores control information for managing the attribute of each device for realizing the above-described device hierarchy and the correspondence relationship between devices, and disk reflected or unreflected data held in the temporary area of the nonvolatile memory 134. Stores control information for management. If the control information stored in the control memory 135 is lost, the data stored in the permanent area of the disk device 137 or the nonvolatile memory 134 cannot be accessed.
It is desirable that 5 has a configuration for high availability, such as non-volatility by battery backup or the like, or duplexing to improve resistance to medium failure.

As shown in FIG. 1, each part in the storage 130 is connected by an internal connection network, and exchanges data, control information, and configuration information between the parts. With this internal connection network, the control processors 132 can share and manage the configuration information of the storage 130.
It is desirable that the internal connection network is also multiplexed from the viewpoint of improving availability.

The management terminal 140 includes a CPU 142, a memory 143, a storage device 144, and a storage 130.
Interface controller 141 connected to the IP network 175, interface controller 147 connected to the IP network 175, input device 145 that receives input from the storage administrator, and display 146 that outputs storage 130 configuration information and management information to the storage administrator.
Output device. The CPU 142 reads the storage management program stored in the storage device 144 into the memory 143 and executes it, thereby referring to the configuration information.
A configuration change instruction, a specific function operation instruction, and the like are performed, and the storage manager or an interface between the management server 110 and the storage 130 is used for the maintenance operation of the storage 130. The management terminal 140 may be omitted, the storage 130 may be directly connected to the management server 110, and the storage 130 may be managed using management software that operates on the management server 110.

The external storage 150 is connected to the storage 130 via the fiber channel switch 120.
One or a plurality of ports 151 connected to the port 131, a control processor 152, a memory 153, a disk cache 154, one or a plurality of disk devices 156, and one or a plurality of ports 155 connected to the respective disk devices 156. The control processor 152 executes the program stored in the memory 153 to process the input / output request to the disk device 156 received from the port 151. In this embodiment, the external storage 150
Is a storage having a configuration smaller than that of the storage 130 without the control memory. However, a storage having the same configuration as the storage 130 may be used.

In this embodiment, since the port 131 of the storage device 130 and the port 151 of the external storage 150 are connected via the fiber channel switch 120 as shown in FIG. 1, direct access from the host computer 100 to the external storage 150 is suppressed. Therefore, it is desirable to set the zoning of the fiber channel switch 120. Further, the port 131 and the port 151 may be directly connected without using the fiber channel switch 120.

Next, the software configuration of the storage 130 in the embodiment of the present invention will be described. FIG. 2 is a software configuration diagram showing an example of control information and a program for storage control processing stored in the control memory and the memory of the storage 130 and the management server 110.

The configuration management information of the storage 130 includes logical device management information 201, LU path management information 202, physical device management information 203, external device management information 204, cache management information 205, nonvolatile memory management information 206, and extended volume in the control memory 135. There are management information 207, pool volume management information 208, and device operation information 209. These pieces of information are stored in the control memory 135 in this embodiment in order to prevent information loss. These pieces of control information can be referred to and updated from the control processor 132, and access via the interconnection network is required at that time. Therefore, a copy of control information necessary for processing executed by each control processor 132 may be held in the memory 133 in order to improve processing performance. Further, the configuration information of the storage 130 is also transmitted and held in the control terminal 140 and the management server 110.
When the management server 110 or the management terminal 140 changes the configuration of the storage 130 in response to an instruction from the storage management software or the storage administrator,
When the configuration of each part in 30 changes due to failure or automatic replacement, the control processor 13
One of the two updates the corresponding configuration information in the control memory 135. Then, the control processor 132 notifies the other control processor 132, the management terminal 140, and the management server 110 via the interconnection network that the control information has been updated due to the configuration change, and the latest information is sent to the control memory 135.
To be taken into the memory of other parts.

The external storage 150, like the storage 130, uses logical device management information 251 and L for device management and data management of logical devices and physical devices inside the own device.
U path management information 252, physical device management information 253, and cache management information 254 are held. These contents are used for the same purpose as the same name information in the storage 130.

Further, the management server 110 holds, in the memory 112, storage management information 232 indicating a copy of device management information collected from the storage 130 or the external storage 150 and an attribute of each storage. Such information may be held in the storage device 113 mounted on the management server 110 in order to avoid data loss.

Next, each management information will be described. First, the logical device management information 201 will be described. FIG. 3 shows an example of the logical device management information 201. The logical device management information 201 holds an information set from the logical device number 301 to the data migration flag 310 for each logical device. The logical device number 301 stores a number assigned by the control processor 132 to the logical device in order to identify the logical device. For size 302,
The capacity of the logical device specified by the logical device number 301 is stored. In the corresponding lower device number 303, the number of the physical device, external device, or expansion volume associated with the logical device, that is, the physical device management information 203, the external device management information 204, or the expansion information, which are the respective management information The entry number of the volume management information 207 is stored. In this embodiment, logical devices and physical devices / external devices / expansion volumes have a one-to-one correspondence. However, when a plurality of physical devices or external devices are combined to form one logical device, logical device management information is used. An entry for storing a physical / external device number list corresponding to each logical device and the number thereof is required in 201. When a logical device is undefined, an invalid value is set in the corresponding lower device number 303. The type 304 stores device type identification information of the logical device. The storage 130 can define logical devices of a plurality of device types, such as a data management unit on the cache and a storage form of device management information (whether storage information is stored in the disk space, storage form, etc.) The type 304 holds which device type the logical device is.

In the device status 305, information indicating the status of the logical device is set. The states include “attached”, “detached”, “unmounted”, and “blocked”. “Attached” indicates that the logical device is operating normally, LU paths are defined in one or more ports 131, and can be accessed from the host computer 100. "Detached"
Indicates that the logical device is defined and operating normally, but cannot be accessed from the host computer 100 because the LU path is undefined. “Unmounted” indicates that the logical device is not defined for a physical device, an external device, or an expansion volume, and is not accessible from the host computer 100. "
“Blocked” indicates that the logical device has failed and cannot be accessed from the host computer 100. The initial value of the device state 305 is “not mounted”, and is changed to “detached” by the logical device definition process, and further changed to “attached” by the LU path definition process.

In the port number of the entry 307, information indicating to which port the logical device is LUN-defined among the plurality of ports 131, that is, identification information of the port 131 used for accessing the logical device is set. . Here, the identification information of the port 131 is a number unique within the storage 130 allocated to each port 131. The target ID and LUN stored in the entry 307 are identifiers for identifying the logical device. In this embodiment, as an identifier for identifying a logical device,
SCSI used when accessing device from host computer 100 on SCSI
-ID and LUN are used. Information is set in the entry 307 when the LU path definition for the logical device is executed. The connected host name 308 is a host name that identifies the host computer 100 that is permitted to access the logical device. As the host name, the WWN (World Wide N) assigned to the port 107 of the host computer 100 is used.
any value that can uniquely identify the host computer 100 or the port 107, such as “Ame)”. In addition, the storage 130 holds management information related to port attributes such as WWN of each port 131. The entry 308 is set by the storage administrator when defining the logical device. In the lower device number 308 being migrated, if the physical device / external device / expansion volume currently supported by the logical device is being migrated to another physical device / external device / expansion volume, the device number of the migration destination Is stored. If the logical device is in the process of data migration, the data migration flag 310 will be
ON "is stored, and address information indicating the last area where the data migration is completed is stored in the data migration progress pointer 309. The entries 308 and 309 are the data migration flag 3
Effective only when 10 is “ON”.

Secondly, the LU path management information 202 will be described. FIG. 4 shows LU path management information 202.
It is an example. The LU path management information 202 holds information for valid LUNs defined for each port for each port 131 in the storage 130. Target ID / LU
The LUN defined in (assigned to) the port 131 is stored in the N402 entry. The corresponding logical device number 403 stores the number of the logical device to which the LUN is assigned. The connection host name 404 is the LU defined for the port 131.
Information indicating the host computer 100 permitted to access N is stored. As the information indicating the host computer 100, for example, the WWN assigned to the port 107 of the host computer 100 described above is used. In some cases, LUNs of a plurality of ports 131 are defined (assigned) to one logical device, and the logical devices can be accessed from the plurality of ports 131. In this case, the L for each LUN of the plurality of ports 131
A union of the connection host names 404 of the U path management information 202 is held in the connection host name 308 of the logical device management information 201 related to the logical device.

Third, the physical device management information 203 will be described. Physical device management information 203
Is used for management of a physical device composed of one or more disk devices 137 in the storage 130 or one or more areas of the nonvolatile memory 134 permanent area. FIG. 5 is an example of the physical device management information 203. Each storage 130 has a disk / memory size and an offset 507 from the physical device number 501 for each physical device existing in the own device.
Holds the information set. In the physical device number 501, an identification number for identifying the physical device is registered. The size 502 stores the capacity of the physical device specified by the physical device number 501. The corresponding upper device number 503 stores the logical device or pool volume number associated with the physical device when the logical device or pool volume is defined. If the physical device is not assigned to a logical device or pool volume, an invalid value is set in the entry 503. In the device status 504, information indicating the status of the physical device is set. The states include “attached”, “detached”, “unmounted”, and “blocked”. “Attached” indicates a state in which the physical device is operating normally and assigned to a logical device or a pool volume. “Detached” indicates that the physical device is defined and operating normally, but is not allocated to a logical device or a pool volume. “Not mounted” indicates that a physical device corresponding to the physical device number is not defined on the disk device 137. “Blocked” indicates that the physical device has failed and cannot be accessed. The initial value of the device state 504 is “unmounted”, and is changed to “detached” by the physical device definition process, and further changed to “attached” when the logical device is defined. The RAID configuration 505 includes the RAID level of the disk device 137 or nonvolatile memory 134 permanent area to which the physical device is assigned, the number of data disk devices / memory areas and the number of parity disk devices / memory areas, the size of stripes as data division units, and the like. Information related to the RAID configuration is retained. The disk / memory number list 506 holds the identification numbers of a plurality of disk devices 137 or a plurality of permanent areas of the nonvolatile memory 134 constituting the RAID to which the physical device is assigned. The identification numbers of the plurality of areas in the permanent area of the disk device 137 and the nonvolatile memory 134 are unique values assigned to identify the plurality of areas in the permanent area of the disk device 137 and the nonvolatile memory 134 in the storage 130. The in-disk / memory size / offset 507 is information indicating which area in the specific area of each disk device 137 or the nonvolatile memory 134 permanent area the physical device is assigned to. In this embodiment, for the sake of simplicity, it is assumed that the offset and size in a specific area of each disk device 137 or nonvolatile memory 134 constituting the RAID are the same for all physical devices.

Fourth, the external device management information 204 will be described. External device management information 204
Is used to manage a logical device of the external storage 150 connected to the storage 130 as an external device. FIG. 6 is an example of the external device management information 204.
The storage 130 holds an information set from the external device number 601 to the target port ID / target ID / LUN list 608 for each external device managed by the storage 130. The external device number 601 stores a unique value in the storage 130 that is assigned to the external device by the control processor 132 of the storage 130. The size 602 stores the capacity of the external device specified by the external device number 601. In the corresponding upper device number 603, the number of the logical device or pool volume in the storage 130 associated with the external device is registered. Device state 60
4, the status of the external device is set in the same manner as the device status 504 of the physical device management information 203. Since the storage 130 is not connected to the external storage 150 in the initial state, the initial value of the device state 604 is “unmounted”. The storage identification information 605 holds identification information of the external storage 150 in which the external device is mounted. As the identification information, a combination of the vendor identification information of the storage and a manufacturing serial number uniquely assigned by each vendor can be considered. In the external storage device number 606, an identification number assigned in the external storage 150, that is, a logical device number is stored for the logical device of the external storage 150 corresponding to the external device. The initiator port number list 607 includes a storage 1 that can access the external device.
30 port 131 identification numbers are registered. When the external device can be accessed from a plurality of ports 131, a plurality of port identification numbers are registered. Target port ID
/ Target ID / LUN list 608 indicates that the external device is external storage 150
When one or more ports 151 are defined as LUNs, one or a plurality of port IDs of the ports 151 and target IDs / LUNs to which the external devices are assigned are held. When the control processor 132 of the storage apparatus 130 accesses an external device (when the control processor transmits an input / output request to the external device from the port 131), the external storage 150 to which the external device belongs to the external device. The assigned target ID and LUN are used as information for identifying the external device.

Fifth, the extended volume management information 207 will be described. In this embodiment, a physical storage area is not allocated at the time of definition of an expansion volume, and a pool volume storage area is dynamically added as a physical storage area corresponding to the access part in response to an update access from a host computer or the like. A virtual volume to be allocated. This information is used to manage the correspondence between the expansion volume, logical device, and pool volume.

FIG. 7 is an example of the extended volume management information 207. The storage 130 holds an information set from the expansion volume number 701 to the real area allocation total amount 707 for every expansion volume. In the expansion volume number 701, a unique value in the storage 130 assigned to the expansion volume by the control processor 132 of the storage 130 is stored.
The size 702 stores the apparent capacity from the host computer of the expansion volume specified by the expansion volume number 701. In the corresponding logical device number 703, the number of the logical device in the storage 130 with which the expansion volume is associated is registered.
In the volume status 704, the status of the extended volume is set as in the information with the same name in the physical device management information 203. In the present embodiment, the initial value of the volume state 704 is “unmounted”. The corresponding pool number list 705 holds the numbers of all pool volumes that store the physical storage area of the expansion volume. The pool entry correspondence management information 706 stores information indicating the correspondence between the storage space provided by the expansion volume to the host computer and the physical storage area in the corresponding pool volume. In the initial state where the extended volume has just been defined, no physical storage area is allocated to the volume.
All storage areas of the volume are in an unallocated state. When there is an update access to the logical device corresponding to the expansion volume from a host computer or the like, the access target position in the expansion volume is calculated, and whether or not the pool volume storage area is assigned to the corresponding storage area entry is checked. When the physical storage area for the access target storage area is not assigned, an appropriate storage area is selected from one or more pool volumes indicated by the corresponding pool number list 705 and assigned as the physical storage area of the access target area of the expansion volume. . The final real area allocation total amount 707 stores the total capacity allocated physical storage areas to the expansion volume.

Sixth, the pool volume management information 208 will be described. In the present invention, a pool volume is a virtual volume that provides and manages a storage area allocated as a physical storage area of an expansion volume. This information is used to manage the correspondence between the pool volume, the expansion volume, and the physical / external device.

FIG. 8 shows an example of the pool volume management information 208. For each pool volume, the storage 130 stores the corresponding expansion volume number list 8 from the pool volume number 801.
Holds information sets up to 08. The pool volume number 801 stores a unique value in the storage 130 that is assigned to the pool volume by the control processor 132 of the storage 130. The size 802 stores the capacity of the pool volume specified by the pool volume number 801, and the free capacity 803 stores the total capacity of the storage area that is not allocated to the expansion volume. The volume status 804 includes extended volume management information 207.
Similarly to the information with the same name, the status of the pool volume is set. The physical / external device number 805 holds the number of the physical / external device to which the pool volume is assigned. The pool entry size 806 stores information indicating the size of an allocation unit (entry) when allocating the storage area of the pool volume as a physical storage area of the expansion volume. The storage area of the expansion volume is divided into a plurality of entries according to the entry size of the corresponding pool volume, and the correspondence state between each entry of the expansion volume and each entry of the pool volume is held in the pool entry allocation state management information 807. The corresponding extended volume number list 808 holds the numbers of one or more extended volumes to which the storage area of the pool volume is assigned as a physical storage area.

Seventh, the device operation information 209 will be described. In this embodiment, the I / O frequency from the host computer 100, the data length, the locality / continuity of the access range, etc. are collected for each logical device. Based on this information, the required amount of cache memory to the physical device or external device corresponding to each logical device, that is, the temporary area of the nonvolatile memory 134 is calculated.

In this embodiment, the storage 130 manages devices / volumes using the above seven device / volume management information. As an initial state of the storage 130, when the device is shipped from the factory, physical devices are already defined in the permanent areas of the disk devices 137 and the nonvolatile memory 134, and logical devices, external devices, expansion volumes, and pool volumes are not defined. is there. Also, the user or storage administrator defines the logical device of the external storage 150 connected to the storage 130 at the time of introduction of the storage 130 as an external device, defines the logical device and the pool volume on the physical device and the external device, and In the case of a volume, an expansion volume and a logical device are further defined, and a LUN is defined for each port 131 for the logical device.

The storage 130 also stores nonvolatile memory management information 206 and cache management information 205 in the control memory 135 in order to manage the nonvolatile memory 134 and the temporary area allocated to the nonvolatile memory 134. The non-volatile memory 134 management information holds information indicating the allocation status of each area of the memory space provided by the mounted non-volatile memory 134, that is, whether the area is allocated to a temporary area or a permanent area, or is not allocated. The cache management information 205 includes area management information of the temporary area of the nonvolatile memory 134, hit miss determination information indicating which data in the address space of the physical device or the external device is stored in which temporary area, and the like. As a temporary area management method, for example, the temporary area is divided into small areas (segments) of a prescribed size, information corresponding to the physical device or external device address space is held for each segment, and information held by each segment It may be possible to hold information indicating whether the device is not reflected (dirty) or reflected (clean).

Next, returning to FIG. 2 again, the storage 130, the external storage 150, and the management server 110.
The programs stored in the memories 133, 153, and 112 will be described. In addition to the above-mentioned copy of control information, each memory has control processors 132, 152 and C in each memory.
A program operating on the PU 111 is stored.

In the present embodiment, the allocation process of the temporary area of the nonvolatile memory 134 of the storage 130, the accompanying physical device allocation process, and the device definition process will be described. In a second embodiment to be described later, the temporary area allocation change process and the device migration process associated therewith will be described. These processes are executed in cooperation with the management server 110, the storage 130, and the external storage 150 in response to a request from a user or an application.

First, in correspondence with the temporary area allocation process of the nonvolatile memory 134, the management server 110 has a temporary area definition instruction process 243 in the memory 112 and the storage 130 has a temporary area allocation process 22.
4 and physical device definition processing 226 are stored in the memory 133, respectively.

FIG. 9 is a diagram illustrating an example of a processing flow of the temporary area definition instruction processing 243 executed by the management server 110. The management server 110 calculates the capacity of the temporary area that must be allocated to the nonvolatile memory 134 of the storage 130 based on the information in the device management information copy 231 (step 901). More specifically, the physical / external device capacity is calculated from the physical device management information 203 and the external device management information 204, and the device management information 2
The access characteristics (frequency, locality, etc.) of the physical / external device corresponding to each logical device are obtained from 09, and the temporary area necessary amount based on the access characteristics is calculated. For example, a method of estimating the cache hit improvement rate for the temporary area allocation capacity from the access frequency and locality, and controlling so as to achieve the performance expected for each logical device can be considered. There are various algorithms for calculating the capacity of these temporary areas. When the required temporary area is calculated based on a specific algorithm, the storage 130 is instructed to allocate a temporary area to the nonvolatile memory 134 (step 902). When the allocation of the temporary area is completed, next, in the nonvolatile memory 134,
The amount and ratio of areas that are not allocated as temporary areas or permanent areas (hereinafter referred to as unallocated areas) are calculated. The storage 130 is instructed to define and define the physical device (step 903). As a result, the excess nonvolatile memory 134 is counted as a free device managed by the storage 130 as a physical device.

FIG. 10 is a diagram illustrating an example of a processing flow of the temporary area allocation processing 224 executed in the storage 130. The storage 130 that has received the temporary area allocation instruction from the management server 110
The control processor 132 selects the non-allocated area of the non-volatile memory 134 for the instructed capacity based on the non-volatile memory management information 206, and changes the unallocated area to a temporary area (steps 1001 and 1002). Further, the newly allocated temporary area is registered in the cache management information 205 so as to be divided into segments (step 1003). Here, the control processor 132 registers in the control memory 135. If the control processor 132 also has the cache management information 205 in the memory 133, the control processor 132 also registers in the memory 133.

FIG. 11 is a diagram illustrating an example of a processing flow of the physical device definition processing 226 executed in the storage 130. The control processor 132 of the storage 130 that has received the physical device definition instruction from the management server 110 to the unallocated area of the nonvolatile memory 134, based on the nonvolatile memory management information 206, unallocated the nonvolatile memory 134 for the instructed capacity. An area is selected and the unallocated area is changed to a permanent area (steps 1101 and 1102). Further, the control processor 132 defines the physical device in the newly allocated permanent area.
The physical device management information 203 is set. Specifically, the control processor 132 adds the size of the permanent area defining the physical device, the identification number of the permanent area, and the offset and size in each area to the physical device management information 203 corresponding to the specific physical device. Register each.

As described above, the temporary area of the nonvolatile memory 130 is defined by the cooperation of the management server 110 and the storage 130. Next, logical device definition processing and I / O processing will be described as an example of normal processing associated with storage operation management in such a system environment.

First, in response to the logical device definition process, the management server 110 has a device definition instruction process 2
41 is stored in the memory 112, and the device definition process 222 is stored in the memory 133 in the storage 130.

FIG. 12 is a diagram illustrating an example of a process flow of the device definition instruction process 241 executed by the management server 110. The management server 110 receives a logical device definition request for the storage 130 from the user or application program via the IP network 175 and the interface control unit 116 (step 1201). Information included in the request includes, for example, storage 130 identification information, logical device number, access source HBA identification information (WWN, etc.), port identification information (port ID of port 131, target ID, LU
N list), device attribute information (normal volume / expansion volume, usage, etc.)
And so on. From these pieces of information, the management server 110 first determines that the storage 130
From the management access right setting etc. set in the above, it is confirmed that the request is from a user having the management access right of the storage 130 or an application program on the host computer 100, and whether or not the request can be executed. judge. Subsequently, the management server 110 determines whether to define the logical device as a normal volume or an extended volume from the information included in the device definition instruction (step 1202). When allocating the logical device as an expansion volume, the management server 110 sets the remaining capacity of all pool volumes corresponding to the physical / external device, that is, the total capacity of the unallocated area as the physical storage area of the expansion volume. If so, an expansion volume corresponding to one of the already defined pool volumes is defined, and the storage 130 is instructed to allocate the logical device to the expansion volume (step 1203).
1204). If the capacity is not sufficient, the management server 110 is a device that is not allocated to either a logical device or a pool volume among physical / external devices (hereinafter referred to as a free device).
If there is a corresponding device, a new pool volume and expansion volume are defined, and the storage 130 is instructed to allocate the logical device (step 12).
05). If there is no free device, the management server 110 reports an error of the logical device definition process to the request source (step 1208). On the other hand, if it is determined in step 1202 that the logical device is defined as a normal volume instead of an expansion volume, a free physical / external device is searched, an allocation target device is determined, and the logical device allocation to the device is stored. 130 is instructed (step 1206). After instructing the storage 130, the management server 110 reports the completion to the request source (step 1207).
).

FIG. 13 is a diagram illustrating an example of a process flow of the device definition process 222 executed in the storage 130. First, the storage 130 receives a logical device definition request from the management server 110 (step 1301). At this time, the management server 110 uses the IP network 1
75 and the logical device definition are transmitted to the storage 130 via the management terminal 140. The control processor 132 stores the request in the control memory 135. As information included in the request, in addition to the information received by the management server 110 in the device definition instruction processing 241,
In the case of a normal volume, a logical device allocation target physical device or an external device number is included. On the other hand, in the case of an expansion volume, a newly allocated expansion volume number and expansion volume allocation target pool volume number, or a newly allocated expansion volume number / pool volume number and pool volume allocation target physical device or external device number are included. Control processor 1 when the logical device is an expansion volume
If necessary, 32 defines a pool volume for the specified physical device or external device (steps 1302, 1303, 1304), and defines an expansion volume for the pool volume (step 1305). Here, as the pool volume definition, specifically, the control processor 132 sets the physical device number or external device number to be assigned to the physical device number or external device number for the information entry corresponding to the designated pool volume number. Set to 805 and refer to the device management information for size 802
Is set, the free space 803 is set to the value of the size 802, and the volume status 804 is "detached"
In addition, the corresponding extended volume number list 808 is initialized to invalid values. The control processor 132 may use a fixed value in the storage 130 for the pool entry size 806, or may set information indicating the same size in a logical device allocation instruction from the user. Further, the control processor 132 initializes the pool entry allocation state management information 807 so that all entries are unallocated to the expansion volume. Further, as the extended volume definition, specifically, the corresponding extended volume number list for the information entry of the pool volume management information 208 of the pool volume corresponding to the information entry of the extended volume management information 207 corresponding to the designated extended volume number The expansion volume number is registered in 808, and “attach” is registered in the volume state 804. Further, the control processor 132 sets the size of the logical device corresponding to the expansion volume to the size 702, sets an invalid value to the corresponding logical device number 703, “detach” to the volume status 704,
“On” is set in the stop flag, the current time is set in the last start / stop time 706, the pool volume number is set in the corresponding pool number list 707, and 0 is set in the real area allocation total amount. The pool entry correspondence management information 708 is initialized so that all entries are in an unallocated state. Next, after defining the logical device for the expansion volume, the control processor 132 performs LUN path definition for the designated port 131 and completes the processing to the management server 1.
10 (steps 1306 and 1307). Specifically, as the logical device definition, the control processor 132 sets the corresponding logical device number 703 and “attach” to the corresponding logical device number 703 for the corresponding logical device management information 201 and the extended volume management information 207. "Is set respectively. Furthermore, the allocation instruction content for the size 302 and the type 304, the extension volume number for the corresponding lower device number 303, the “detached” device status, the invalid values in the entries 306 to 309, the “off” in the data migration flag 310 are displayed. "Is set respectively. Further, specifically, as the LUN path definition, the control processor 132 sets the corresponding logical device management information 201 and LU path management information 202.
, The port information designated by the assignment instruction and the connected host computer 100 information are set in the entries 306, 307 and 402, 404, respectively. Corresponding logical device number 4
Set the logical device number to 03. On the other hand, if the logical device is a normal volume, the logical device is defined for the designated free physical device or external device, LUN path definition is performed, and processing completion is reported (steps 1302, 1306, 1307).
). Here, specifically, the logical device management information 201 corresponding to the logical device definition
Necessary information is set in the physical device management information 202 or the external device management information 203.

Next, in response to the I / O processing from the host computer to the logical device, the storage 130
The I / O processing 221 is stored in the memory 133.

FIG. 19 is a diagram illustrating an example of a processing flow of the I / O processing 221 executed in the storage 130. The storage 130 receives the I / O request from the host computer 100 (step 1901). The control processor 132 separates the process according to the type of the I / O request (step 1902).

If the control processor 132 determines in step 1902 that the request is a read request, the control processor 132 determines whether or not the read target data is a cache hit miss based on the cache management information 205 (step 1903). In the case of a cache miss, the cache management information 205 is manipulated to allocate a free segment from the non-volatile memory temporary area (step 1904), and the disk 137 or the external storage 15 as a physical device.
The data staged from 0 is stored in the nonvolatile memory 134 (step 1905), and the data is transferred to the host computer (step 1906). In the case of a cache hit, the control processor 132 reads data from the nonvolatile memory 134.

In step 1902, the control processor 132 determines that the I / O request is a write request, and determines whether it is a cache miss (step 1908). In the case of a cache miss, the control processor 132 allocates a free segment (step 1909).
Here, the cache miss refers to the nonvolatile memory 1 for the data area in which the logical device is located.
34 is a state in which no segment is assigned. After that, the control processor 132 stores the write data from the host computer in the segment (step 1910).
Destage to a permanent area such as a physical device or external device (step 1911)
). On the other hand, if a cache hit occurs in step 1908, a permanent area exists in the nonvolatile memory 134, so data is stored at the corresponding address.

On the other hand, if it is determined in step 1902 that the control processor 132 is neither a write request nor a read request (eg, a sense command), the control processor 132 executes the requested process (step 1912).

Finally, the control processor 132 that executed the process reports the process completion to the host computer 100 and ends the process (step 1907).

Since the present embodiment is configured as described above, the following effects can be obtained. The temporary area capacity is determined according to the configuration of the virtual storage system, the capacity of the virtualized device, and the access characteristics, and the nonvolatile memory can be used separately for the temporary area and the permanent area.

In the first embodiment, the method of calculating the necessary amount of the temporary area from the capacity and access characteristics of the physical device and the external device managed by the storage 130 and allocating the necessary temporary area has been described. In the second embodiment, a method for recognizing a temporary area allocation amount by detecting an increase / decrease in external storage or a change in access characteristics to each device will be described.

The second embodiment will be described with reference to FIGS. 1 to 8 and FIGS. 14 to 18. 1st and 2nd
Since this embodiment has many common points, only the difference between them will be described.

The software configuration in the second embodiment is shown in FIG. 2 as in the first embodiment. Second
In the embodiment, first, in response to the temporary area allocation change process in the nonvolatile memory 134, the management server 110 has a temporary area change instruction process 244 in the memory 112, and the storage 130 has a temporary area deallocation process 225 and a device. Migration processing 223 and physical device definition cancellation processing 227
Are stored in the memory 133, respectively.

FIG. 14 is a diagram illustrating an example of a processing flow of the temporary area change instruction processing 244 executed by the management server 110. The CPU 111 of the management server 110 calculates the capacity of the temporary area already allocated to the nonvolatile memory 134 at a certain time based on the cache management information 205 and the nonvolatile memory management information 206 (step 1401). Furthermore, the temporary area allocation required amount at that time is calculated by the same method as in Step 1001 of the temporary area allocation processing 224 (
Step 1402). When these values are compared and the difference between the two exceeds a specified value, the CPU 11
1 controls the allocation amount of the temporary area. If the capacity of the allocated temporary area of the nonvolatile memory 134 is equal to or greater than a predetermined value with respect to the required amount (step 1403), a temporary area for the excess capacity is selected as a release candidate (step 1408). As a method for selecting a temporary area, a temporary area that includes many unused segments may be selected, or a temporary area that has few segments storing dirty data may be selected. The CPU 111 adds information for specifying the selected area and instructs the storage 130 to deallocate the specified temporary area (step 1409).
. At this time, the temporary area to be deallocated may be determined by adding only the capacity of the temporary area that needs to be deallocated without selecting the temporary area to be deallocated. When the allocation of the specific temporary area is released, the CPU 111 causes the nonvolatile memory 13 that has become an unallocated area.
The storage area 130 is instructed to set the permanent area and the physical device definition for the area 4 (step 1410). If the capacity of the allocated temporary area is smaller than the required value with respect to the required amount, the physical device defined in the nonvolatile memory 134 corresponding to the excess capacity is selected, and the logical device corresponding to the physical device is selected as the disk. Instruct the storage 130 to migrate to a physical device or an external device defined in the device 137 (step 1405)
. When the migration is completed, the CPU 111 cancels the deallocation of the physical device.
(Step 1406). When it is completed, the management server 110 stores the storage 1
30 is instructed to allocate the unallocated area to the temporary area (step 1407).
.

FIG. 15 is a diagram illustrating an example of a processing flow of the temporary area deallocation processing 225 executed in the storage 130. Sends a temporary area deallocation instruction from the management server 110 to the management terminal 1
Receive via 40. Thereafter, the control processor 132 selects a temporary area of the non-volatile memory 134 to be deallocated as necessary (steps 1501 and 1502), and the dirty data for the temporary area to be deallocated is based on the cache management information 205. Whether there is a segment storing host update data that has not been reflected in the permanent area. If there is a segment, the unreflected data (dirty data) is destaged to the corresponding physical device or external device (step 1503). ). When the destage of all the dirty data in the target temporary area is completed, the corresponding entry in the nonvolatile memory management information 206 is changed to indicate the unallocated area, and all the corresponding cache management information 205 is initialized (step 1504). ).

FIG. 16 is a diagram illustrating an example of a process flow of the device migration process 223 executed in the storage 130. The control processor 132 periodically activates this process to check whether a device migration request is registered (step 1601). As described above, the device migration request is generated when the logical device corresponding to the physical device defined in the permanent area for expanding the temporary area of the nonvolatile memory 134 is migrated to the disk device 137 or an external device which is another physical device. be registered. Further, as will be described later, the storage 130 is constructed.
It is registered when data transfer between tiers is required based on the data value in the permanent area of the non-volatile memory 134, tier storage such as the disk device 137 or external device. If the device migration request is not registered, this process is terminated without doing anything (step 160).
1). If the request is registered, the process is separated depending on whether the target logical device is an expansion volume or a normal volume (step 1602). When the logical device is a normal volume, if the request process is a migration to an empty device, the data of the migration source logical device is migrated sequentially from the start address to the migration destination physical device or external device. The device management information is updated so as to correspond from the physical device or external device that currently supports the migration source logical device to the migration destination free physical device or external device (steps 1606 and 1607). When the request processing is replacement with another logical device, the data of the migration source / destination logical device is read sequentially from the head address to another storage device such as a temporary area of the nonvolatile memory 134, and the data of the migration source logical device is read. The data of the migration destination logical device is replaced with the migration source logical device and written to the migration destination logical device. When the data replacement is completed for the entire area of the logical device, the device management information is updated so that the correspondence relationship between the migration source logical device and the migration destination logical device and the corresponding physical device or external device is switched (step 1604, 1605).

On the other hand, if the logical device is an expansion volume, if the request process is a migration process to another pool volume, the control processor 132 has assigned real areas to the corresponding expansion volume sequentially from the first entry of the logical device. The entry is searched, and only the data of the entry is added to the free entry of the migration destination pool volume. More specifically, the migration destination pool volume is added to the corresponding pool volume of the expansion volume, the data read out from the entry of the migration source pool volume is assigned and written to the free entry of the migration destination pool volume, and the entry of the migration destination volume is written. The management information is updated so as to correspond to the entry of the extension volume. This process is executed sequentially using the progress pointer of the logical device management information, and when the migration process is completed for all areas, the migration source pool volume (initially the logical device is supported) from the corresponding pool volume list of the expansion volume. Delete pool volume) (steps 1609 and 1610). When the request process is a replacement instruction with another logical device, the control processor 132 searches for an entry to which a real area (pool volume entry) is assigned in order from the first entry of the migration source / destination logical device. For the target entry, a free entry in the other pool volume is newly allocated and the data of the entry is migrated. If there is a large difference in the number of migration entries in the migration destination / source logical device during this migration process, the number of free entries in one pool volume may be exhausted. Therefore, it is necessary to control the progress so that the number of migration execution entries does not greatly deviate between the migration destination / source logical devices. When the replacement processing of all areas is completed in this way, the device management information is updated so that the correspondence between the migration destination / migration source logical device and each pool volume is switched (steps 1611 to 1616).

FIG. 17 is a diagram illustrating an example of a processing flow of the physical device definition cancellation processing 227 executed in the storage 130. The control processor 132 initializes the physical device management information 203 corresponding to the designated physical device, and cancels the correspondence between the permanent area set in the nonvolatile memory 134 and the unallocated area (step 1701). Next, the nonvolatile memory management information 206 is updated so that the permanent area of the nonvolatile memory 134 becomes an unallocated area, and the process ends (
Step 1702).

As described above, the allocation state of the temporary area of the nonvolatile memory 134 is changed by the cooperation of the management server 110 and the storage 130. Next, a logical device migration process will be described as an example of a normal process associated with storage operation management in such a system environment.

Corresponding to the logical device migration process, the management server 110 has a device migration instruction process 242.
Is stored in the memory 112, and the device migration process 223 is stored in the memory 133 in the storage 130. Among these, the device migration instruction processing 242 that has not been described so far will be described.

FIG. 18 is a diagram illustrating an example of a process flow of the device migration instruction process 242 executed by the management server 110. The management server 110 determines whether it is necessary to replace another logical device corresponding to another physical device or external device from among the logical devices (step 18).
01). Specifically, as in DLCM (Data Life Cycle Management), between the tiers of the tiered storage managed by the storage 130 based on the value of data stored in each logical device, access frequency, data retention period, etc. The data arrangement between devices is reconsidered so as to obtain a more suitable data arrangement. If there is a device that needs to be replaced as a result of determination (
In step 1802), the management server 110 instructs the storage 130 to replace the device to which information specifying the migration source and destination devices is added (step 1803). When the replacement is completed, the management server 110 stores the latest information on the device from the storage 130 in the memory. 112, and the device management information copy 231 is updated (step 1804).

In step 1801, the control processor 132 records the access frequency per device unit, for example, when data is arranged according to the access frequency in device units. The CPU 111 of the management server 110 acquires this access frequency information, and instructs data relocation according to the algorithm stored in the device migration instruction processing 242.

Next, a case where data relocation is executed in units of files instead of in units of devices will be described. FIG. 21 is a diagram showing an example of the address conversion table 2100 that the management server 110 has. The CPU 111 can create the address conversion table 210 by acquiring information from the host computer. By using the address conversion table 210, the CPU 111 can read and write data in units of files. The address conversion table is created for each operating system.

FIG. 22 is a diagram illustrating an example of information that can be collected as a result of execution by the CPU 101 of the host computer 100 according to an instruction from the management server 110. The CPU 101 of the host computer 100 collects the file name, file creation date / time, last access date / time, file size, file address, storage destination LU number, and the like for each file. CPU after collecting
101 transmits to the management server 110. A file that has not been accessed for a long time can be rearranged according to the file generation date and the last access date and time of this information. According to the configuration of the second embodiment, in addition to the effects of the first embodiment, it is possible to dynamically change the allocation of the temporary / permanent area of the nonvolatile memory in accordance with the system configuration change or the access characteristic change.

As mentioned above, although several suitable embodiment of this invention was described, these are the illustrations for description of this invention, Comprising: It is not the meaning which limits the scope of the present invention only to these embodiment. The present invention can be implemented in various other forms.

For example, the present invention can be applied to a mainframe system instead of an open system.

Further, for example, the function of the management server 110 can be incorporated into the host computer 100. In this case, the management server 110 is not necessary, and both operation and management can be performed by the host computer. In addition, when instructing the storage 130, the management server 110 can execute the instruction in-band instead of out-band via the fiber channel switch to the management server 110.

Further, for example, a NAS adapter can be built in the storage system. FIG.
These are figures which show an example of the storage system which incorporated the NAS adapter. The difference from FIG. 1 is that the NAS adapter 2312 is built in the storage system 130 and that I
The P switch 122 is added. The processor 2 of the NAS adapter 2310
311 can convert the file access from the host computer into block access by the operating system or file system stored in the memory 2312 and store the data in the nonvolatile memory 134, the disk 137, and the external storage 150.
When the NAS adapter 2310 is built in, the file collection information 2200 is collected by the processor 2311 and transmitted to the management server 110 via the management terminal 140.

DESCRIPTION OF SYMBOLS 100 ... Host computer 110 ... Management server 120 ... Fiber channel switch 130 ... Storage 140 ... Management terminal 150 ... External storage 175 ... IP network 201 ... Logical device management information 202 ... LU path management information 203 ... Physical device management information 204 ... External device Management information 205 ... Cache management information 206 ... Non-volatile memory management information 207 ... Extended volume management information 208 ... Pool volume management information 209 ... Device operation information 221 ... I / O processing 222 ... Device definition processing 223 ... Device migration processing 224 ... Temporary area Allocation process 225 ... Temporary area deallocation process 226 ... Physical device definition process 227 ... Physical device definition deallocation process 261 ... I / O process 241 ... Device definition instruction process 242 ... Device migration instruction process 243 ... Temporary area Righteousness instruction processing 244 ... temporary area change instruction processing

Claims (20)

One or more flash memory drives and one or more disk drives included in a pool for storing data sent from the computer;
A processor connected to the one or more flash memory drives and the one or more disk drives and providing a virtual volume to the computer;
Have
The processor is
Assigning a first area of the one or more flash memory drives as a cache area for temporarily storing data;
A second area of the one or more flash memory drives and a third area of the one or more disk drives, which are different from the first area, are allocated to a pool for storing data. ,
In response to a write request from the computer to the virtual volume, a storage area having a predetermined size is allocated to the virtual volume from the second area or the third area included in the pool.
A storage system characterized by that. The storage system according to claim 1, wherein
The processor is
Allocating the first area as the cache area based on a first request specifying the capacity of the first area;
The storage system according to claim 1, wherein the second area is allocated to the pool based on a second request designating a capacity of the second area. The storage system according to claim 2, wherein
The processor is
Managing the plurality of disk drives as the third area;
A storage system that stores data stored in the third area in the first area based on a frequency of access requests. The storage system according to claim 3, wherein
The storage system, wherein the processor writes data stored in the first area to the third area. The storage system according to claim 3, wherein
When the processor receives the release request for the first area, the processor stores data that has not been written to the second area or the third area among the data stored in the first area. A storage system, wherein data is written in the second area or the third area. The storage system according to claim 1, wherein
The storage system according to claim 1, wherein the data stored in the first area is duplicated. The storage system according to claim 2, wherein
The processor is
The storage system, wherein after assigning the first area as the cache area, the second area is assigned to the pool. The storage system according to claim 1,
The processor is
A storage system that manages the capacity of the pool and the capacity of a storage area that is not allocated to the virtual volume among the capacity of the pool. The storage system according to claim 1,
The processor is
A storage system that measures the access frequency from the computer to the virtual volume. The storage system according to claim 1,
The processor is
A storage system, wherein the data is moved between the second area and the third area based on the frequency of access to the data. A storage control method for a storage system including one or more flash memory drives and one or more disk drives included in a pool for storing data sent from a computer,
Assigning a first area of the one or more flash memory drives as a cache area for temporarily storing data;
A second area of the one or more flash memory drives and a third area of the one or more disk drives, which are different from the first area, are allocated to a pool for storing data. ,
In response to a write request from the computer to the virtual volume , a storage area having a predetermined size is allocated to the virtual volume from the second area or the third area included in the pool.
A storage control method. The storage control method according to claim 11, comprising:
Allocating the first area as the cache area based on a first request specifying the capacity of the first area;
A storage control method, comprising: allocating the second area to the pool based on a second request designating a capacity of the second area. A storage control method according to claim 12 , comprising:
Managing the plurality of disk drives as the third area;
A storage control method comprising: storing data stored in the third area in the first area based on a frequency of access requests. The storage control method according to claim 13, comprising:
A storage control method comprising writing data stored in the first area to the third area. The storage control method according to claim 13, comprising:
When receiving the release request for the first area, the data that has not been written to the second area or the third area among the data stored in the first area is transferred to the second area or A storage control method, wherein data is written in the third area. The storage control method according to claim 11, comprising:
A storage control method, wherein data stored in the first area is duplicated. A storage control method according to claim 12, comprising:
A storage control method comprising: allocating the second area to the pool after allocating the first area as the cache area. The storage control method according to claim 11, comprising:
A storage control method comprising managing the capacity of the pool and the capacity of a storage area not allocated to the virtual volume among the capacity of the pool. The storage control method according to claim 11, comprising:
A storage control method, comprising: measuring an access frequency from the computer to the virtual volume. The storage control method according to claim 11, comprising:
A storage control method, wherein the data is moved between the second area and the third area based on an access frequency to the data.

Images