JP5146351B2 - Storage device, data relocation method, and program - Google Patents

Description

  The present invention relates to a storage apparatus having a plurality of disk pools, and more particularly to data relocation performed in the storage apparatus.

  A storage device configured by combining a plurality of physical disks forms a logical disk, which is a virtual storage area, by combining a large number of physical blocks on the physical disks, and the host computer responds to the addresses on the logical disks. In many cases, a command for reading or writing data is issued.

  In these storage apparatuses, in recent years, more and more physical disks have been combined in response to an increase in required storage capacity. Accordingly, in order to reduce the cost of the storage device, an expensive high-speed physical disk instead of a high data access speed, and an inexpensive low-speed physical disk that is less expensive than a high-speed physical disk, A plurality of storage devices are constructed by combining them.

  At this time, there is a technique in which frequently accessed data is stored in a high-speed physical disk and data that is not so stored in a low-speed physical disk in accordance with the access frequency accessed from the host computer. Since it is said that “80% of data access occurs with respect to 20% of the physical area”, it is expected that this technology can realize cost reduction without impairing the performance of the storage device.

  However, the above-described data rearrangement according to the prior art is normally performed in units of logical disks. Therefore, especially when the size of the logical disk is large, frequent relocation of a large amount of data imposes an excessive load on the storage device and the physical disk, resulting in normal data read and write processing. Adversely affected.

  When there is a bias in data access in the logical disk, there are areas where the effect can be obtained by the rearrangement and areas where the effect cannot be obtained. If data is relocated to such a logical disk in units of logical disks, the load on the storage device and physical disk described above, and the problem of adverse effects on normal processing will be in an area where the need for relocation is low. It also occurs. In addition, there is a problem that it is extremely difficult to determine whether the necessity of rearrangement is high or low only by information in units of logical disks.

  The following technical literature is related to this. Patent Document 1 discloses an example of a method for reallocating data in units of physical blocks between physical disks corresponding to logical disks. Patent Document 2 discloses an example of a method for rearranging a logical disk to a higher-speed physical disk according to the access frequency.

  In the technique of the above-described Patent Document 2, it is determined whether or not data relocation is possible to relocate a logical disk to a higher-speed physical disk based on “access information” that records information about access frequency. Combining this with the technology for reallocation of data in physical block units of Patent Document 1 will solve the above-mentioned problems associated with data relocation in logical disk units.

JP 2003-131816 A JP 2003-271425 A

  However, in the technique of the above-mentioned Patent Document 2, since the process of rearranging frequently accessed data to the high-speed physical disk is performed separately from normal data reading and writing, particularly when a large amount of data is rearranged This greatly increases the load on the storage device and the physical disk.

  In the technique disclosed in Patent Document 2, data with high access frequency can be relocated to a high speed physical disk, but data with low access frequency cannot be relocated to a low speed physical disk. Therefore, if the data rearrangement is continued by the technique combining Patent Documents 1 and 2, data is concentrated only on the high-speed physical disk. Therefore, the meaning is to combine the high-speed physical disk and the low-speed physical disk. Itself disappears.

  Even if this is combined with the technology of reallocation of data in physical block units in Patent Document 1, the above two problems cannot be solved.

  An object of the present invention is to provide a storage device, a data relocation method, and a program that enable processing to relocate frequently accessed data to a high-speed physical disk while suppressing a load on the storage device and the physical disk. is there.

  In order to achieve the above object, a storage apparatus according to the present invention includes a first disk pool including one or more first physical disk apparatuses, and a second data access speed lower than that of the first physical disk apparatus. A second disk pool including one or more physical disk devices, a cache memory for temporarily storing data, and a physical block on the first disk pool and the second disk pool are associated with addresses on the logical disk. A control unit that reads / writes data from / to a physical block in response to a read request / write request from the host computer to the logical disk using a logical disk allocation table. Correspondingly, data is read from the first physical block and cache memo is read. If the first physical block is on the second disk pool and the data on the cache memory is written to the second physical block on the first disk pool, the first physical block is stored Data read means for releasing and updating the logical disk allocation table is included.

  In order to achieve the above object, a data relocation method according to the present invention includes a first disk pool including one or more first physical disk devices, and a data access speed lower than that of the first physical disk device. A method of rearranging data in a storage device having a second disk pool including one or more physical disk devices and a cache memory for temporarily storing data, on a logical disk from a host computer The first request for specifying the first physical block on the first disk pool or the second disk pool corresponding to the address on the logical disk by the logical disk allocation table prepared in advance is received. Physical block identification step and cache memory by reading data from the first physical block A data read step for storing data in the first disk pool, a first disk pool specifying step for specifying in which of the first disk pool and the second disk pool the first physical block exists, and a first physical block When the data is on the second disk pool, the logical disk allocation table is updated by releasing the first physical block after writing the data on the cache memory to the second physical block on the first disk pool. 1 data rearrangement step.

  In order to achieve the above object, a data relocation program according to the present invention includes a first disk pool including one or more first physical disk devices and a data access speed lower than that of the first physical disk device. A read request for an address on a logical disk from a host computer to a computer that controls a storage device having a second disk pool including one or more physical disk devices and a cache memory for temporarily storing data. And a first physical block specifying process for specifying the first physical block on the first disk pool or the second disk pool corresponding to the address on the logical disk by using a logical disk allocation table prepared in advance. Read data from the first physical block and put it in the cache memory Existing data read processing, first disk pool specifying processing for specifying in which of the first disk pool and the second disk pool the first physical block is present, and the first physical block being the second The first physical block is released and the logical disk allocation table is updated after writing the data on the cache memory to the second physical block on the first disk pool. Data relocation processing is executed.

  In the present invention, after data is stored on the cache memory, when the physical block exists on the low-speed physical disk, the data on the cache memory is recorded on the high-speed physical disk to release the area on the low-speed physical disk. Since it is configured, data rearrangement can be performed during data read / write processing. As a result, an unprecedented superior storage device, data relocation method, and program capable of relocating frequently accessed data to a high-speed physical disk while suppressing the load on the storage device and the physical disk Can be provided.

It is a block diagram which shows the structure of the storage apparatus which concerns on embodiment of this invention. FIG. 2 is a functional block diagram illustrating a configuration of functions executed by a control unit in the storage apparatus illustrated in FIG. 1. FIG. 3 is a conceptual diagram illustrating a data configuration of a logical disk allocation table illustrated in FIG. 2. It is a conceptual diagram which shows the data structure of the read frequency table shown in FIG. It is a conceptual diagram which shows the data structure of the write frequency table shown in FIG. It is a conceptual diagram which shows the data structure of the no access status recording table shown in FIG. It is a flowchart which shows operation | movement of the data reading means shown in FIG. 3 is a flowchart showing the operation of the data writing means shown in FIG. It is a flowchart which shows operation | movement of the data rearrangement means shown in FIG.

  FIG. 1 is a block diagram showing the configuration of the storage apparatus 100 according to the embodiment of the present invention. The storage apparatus 100 is connected to the host computer 10 via a network, and reads or writes data according to a request from the host computer 10.

  The storage apparatus 100 is connected to the host computer 10 to exchange data, a cache memory 140 that temporarily stores unprocessed requests and data, a high-speed disk pool 102 and a low-speed disk pool 112 described later, And a control unit 150 that controls the operation of each unit described above.

  The high speed disk pool 102 includes one or more high speed physical disks 101 having a high data access speed. The low-speed disk pool 112 includes one or more low-speed physical disks 111 whose data access speed is low. Here, “high speed” and “low speed” are merely concepts in which the data access speed is relatively compared between them.

  The control unit 150 is a microcomputer having a processor, a ROM, a RAM, and the like. The physical blocks 103a, 103b... On the high-speed disk pool 102 and the physical blocks 113a, 113b. A physical block is configured by partitioning, and a logical disk 120, which is a logical storage area, is configured from each set of physical blocks.

  FIG. 2 is a functional block diagram showing a configuration of functions executed by the control unit 150 in the storage apparatus 100 shown in FIG. These functions are preferably realized as a program executed by a computer constituting the control unit 150. The control unit 150 executes functions such as a data reading unit 201, a data writing unit 202, and a data rearrangement unit 203 for the logical disk 120. Also, each table such as a logical disk allocation table 301, a logical disk read frequency table 302, a logical disk write frequency table 303, and a no access status recording table 304, which will be described later, is managed.

  The data reading unit 201 operates when a read request is received from the host computer 10, reads data recorded from an area corresponding to the requested address for the target logical disk, and stores the data in the cache memory 140. After that, the response is returned to the host computer.

  However, if the read area is a physical block on the low-speed disk pool 112, the data reading unit 201 secures a new physical block on the high-speed disk pool 102 and transfers the data on the cache memory 140 to the high-speed disk pool. Write to the physical block, and allocate the physical block as a new storage area for the logical disk.

  The data writing unit 202 operates when a write request is received from the host computer 10, and when the area corresponding to the requested address for the target logical disk is a physical block of the low-speed disk pool 112, A new physical block on the high-speed disk pool 102 is secured, data on the cache memory 140 is written into the physical block of the high-speed disk pool, and the physical block is allocated as a new storage area for the logical disk.

  The data relocation unit 203 operates periodically by a timer or the like (or can be operated by a relocation request from the outside), selects a target logical disk, and then selects a logical disk for the logical disk. With reference to the read frequency table 302 and the logical disk write frequency table 303, the presence / absence of access is calculated, and the state of no access to the logical disk for a certain period is detected.

  When such a state is detected and a physical block of the high-speed disk pool 102 is allocated to a non-accessed area, the data relocation unit 203 renews the physical block on the low-speed disk pool 112. And the data in the area is read and written to the newly reserved physical block on the low-speed disk pool 112, and the physical block is allocated as a new storage area for the logical disk.

  FIG. 3 is a conceptual diagram showing the data structure of the logical disk allocation table 301 shown in FIG. The logical disk allocation table 301 is a table for managing the allocation status of logical disks, and includes data such as a logical disk ID 301a, an allocation address 301b, a performance category 301c, storage pool identification information 301d, and storage address information 301e.

  The logical disk ID 301a is information for identifying the logical disk 120. The allocation address 301b indicates an address corresponding to an area on the logical disk. The performance category 301c indicates whether the area on the logical disk is included in the high-speed disk pool 102 or the low-speed disk pool 112. The storage pool identification information 301d and the storage address information 301e indicate the location of the physical block corresponding to the area on the logical disk represented by the logical disk ID 301a and the assigned address 301b.

  Here, when the storage pool identification information 301d is “P1”, the physical partition is included in the high-speed disk pool 102, so the performance classification 301c is “high”. When the storage pool identification information 301d is “P2”, the physical block is included in the low-speed disk pool 112, so the performance classification 301c is “low”. There are areas on the logical disk to which no physical block is allocated. In that case, the performance classification 301c, the storage pool identification information 301d, and the storage address information 301e corresponding to the area are blank.

  FIG. 4 is a conceptual diagram showing the data structure of the read frequency table 302 shown in FIG. The read frequency table 302 is a table for managing the read frequency of the logical disk. The read frequency table 302 reads the logical disk ID 302a that is information for identifying the logical disk 120, the address 302b corresponding to the area on the logical disk, and the address. Each piece of data such as read discrimination information 302c indicating the presence or absence of the data is included.

  FIG. 5 is a conceptual diagram showing the data structure of the write frequency table 303 shown in FIG. The write frequency table 303 is a table for managing the write frequency of the logical disk. The logical disk ID 303a, which is information for identifying the logical disk 120, the address 303b corresponding to the area on the logical disk, and the write to the address Each piece of data, such as write discrimination information 303c indicating the presence or absence of.

  FIG. 6 is a conceptual diagram showing the data structure of the no access status record table 304 shown in FIG. The non-access status recording table 304 is a table for recording the presence / absence of access to the logical disk. The logical disk ID 304a, which is information for identifying the logical disk 120, the address 304b corresponding to the area on the logical disk, and the address On the other hand, each piece of data includes a no-access frequency 304c indicating the number of no-access states.

  FIG. 7 is a flowchart showing the operation of the data reading unit 201 shown in FIG. When the data reading unit 201 starts the operation in response to a read request from the host computer 10, first, based on the logical disk ID 302a for identifying the requested logical disk, the requested address, and the data size, a logical for the data read request. The corresponding address 302b on the disk is calculated, and the logical disk ID 302a and the read determination information 302c at the address 302b are read and recorded as “present” on the read frequency table 302 and updated (step S401).

  Subsequently, the data reading unit 201 refers to the logical disk allocation table 301 and obtains storage pool identification information 301d and storage address information 301e of the physical block allocated to the address 301b of the logical disk 301a (step S402). ), Data is read from the physical block, stored in the cache memory 140, and a response is returned to the host computer 10 (step S403).

  Thereafter, the data reading unit 201 determines whether the physical block is allocated to the high speed disk pool 102 or the low speed disk pool 112 from the performance classification 301c (step S404). If it is the high-speed disk pool 102, the processing is terminated as it is, but if it is assigned to the low-speed disk pool 112, the processing from step S405 is performed to replace the physical block with the high-speed disk pool 102.

  In the process of replacing the physical block with the high-speed disk pool 102, first, an unallocated physical block in the high-speed disk pool 102 is secured as a newly allocated area (step S405), and the read data stored in the cache memory 140 is stored. Is written in the area secured in step S405 (step S406).

  In the logical disk allocation table 301, the ID and address of the newly secured high-speed disk pool are recorded and updated in the storage pool identification information 301d and storage address information 301e as the allocation area for the address 301b of the logical disk 301a. (Step S407). Finally, the original physical block allocated to the low-speed disk pool 112 is released as an unallocated area (step S408). This completes the process of replacing physical blocks with the high-speed disk pool 102.

  FIG. 8 is a flowchart showing the operation of the data writing means 202 shown in FIG. When the data writing unit 202 starts its operation in response to a write request from the host computer 10, first, based on the logical disk ID 303a for identifying the requested logical disk, the requested address, and the data size, the logical for the data write request. The corresponding address 303b on the disk is calculated, and the logical disc ID 303a and the write discriminating information 303c at the address 303b are recorded as “write” on the write frequency table 303 and updated (step S501).

  Subsequently, the data writing means 202 stores the requested data in the cache memory 140 and returns a response to the host computer 10 (step S502), referring to the logical disk allocation table 301 and corresponding address 301b of the logical disk 301a. The storage pool identification information 301d and the storage address information 301e of the physical block allocated to is acquired (step S503), and the physical block is allocated to either the high speed disk pool 102 or the low speed disk pool 112 from the performance classification 301c. It is determined whether it has been performed (step S504).

  If it is the high-speed disk pool 102 that is assigned to the physical block, the data stored in the cache memory 140 is directly written to the physical block (step S509), and the process ends, but the data is assigned to the low-speed disk pool 112. If so, the processing from step S505 is executed to replace the physical block with the high-speed disk pool 102.

  In the process of replacing the physical block with the high-speed disk pool 102, first, an unallocated physical block in the high-speed disk pool 102 is secured as an area to be newly allocated (step S505), and the data stored in the cache memory 140 is changed to the step Write in the area secured in S505 (step S506).

  In the logical disk allocation table 301, the ID and address of the newly secured high-speed disk pool 102 are recorded and updated in the storage pool identification information 301d and the storage address information 301e as the allocation area for the address 301b of the logical disk 301a. (Step S507). Finally, the original physical block allocated to the low-speed disk pool 112 is released as an unallocated area (step S508). This completes the process of replacing physical blocks with the high-speed disk pool 102.

  In the processes according to FIGS. 7 and 8 described above, the process of storing data in the cache memory 140 and returning a response to the host computer 10 is a process normally performed in a conventional storage apparatus. Since it can be said that only data relocation processing from the low-speed disk pool 112 to the high-speed disk pool 102 is added to this, the load on the storage device and the physical disk related to this data relocation is very small.

  FIG. 9 is a flowchart showing the operation of the data rearrangement unit 203 shown in FIG. When the data rearrangement unit 203 starts to operate periodically by a timer (not shown) or the like, first, the data relocation unit 203 refers to the read frequency table 302 and the write frequency table 303 (step S601), and each of the data residing on the storage apparatus 100 exists. The address of each block assigned to the logical disk 120 is checked.

  The data rearrangement means 203 checks whether or not one address of one logical disk is accessed, that is, either a read request or a write request is made within one cycle in which the data relocation means 203 operates (see FIG. Step S602).

  More specifically, the read determination information 302c at the logical disk ID 302a and address 302b is “read” on the read frequency table 302, or the write determination information at the logical disk ID 303a and address 303b on the write frequency table 303. If 303c is “write”, it is determined that there has been access to the address of the logical disk in the past. When both the read determination information 302c and the write determination information 303c are “none”, it is determined that there is no access in the past.

  If it is determined in step S602 that no access has existed in the past, 1 is added to the value of the no access frequency 304c at the address 304b of the logical disk ID 304a in the no access status recording table 304 (step S603). ), The process proceeds to step S605.

  On the other hand, if it is determined in step S602 that access has existed in the past, the value of the no-access frequency 304c at the address 304b of the logical disk ID 304a is set to “0”, and the read determination of the address in the read frequency table 302 is further performed. When the information 302c is read “None”, the write determination information 303c at the address on the write frequency table 303 is also changed to “None” (Step S604), and the process proceeds to Step S605.

  That is, if the operation of the data reading unit 201 shown in FIG. 7 is not performed within one cycle in which the data rearrangement unit 203 operates, the read determination information 302c remains “none”. Similarly, if the operation of the data writing unit 202 shown in FIG. 8 is not performed within one cycle, the write determination information 303c remains “none”. Thereby, the determination of the presence or absence of access shown in step S602 is established.

  Subsequently, it is determined whether or not the value of the no access frequency 304c at the address 304b of the logical disk ID 304a satisfies the movement condition (step S605). Here, the movement condition is that the value of the no-access frequency 304c exceeds a specific threshold value. This threshold value may be set in advance as a fixed value, or may be a value that can be changed from the external environment. If the movement condition is not satisfied, the process proceeds to step S612 described later.

  If the migration condition is satisfied in step S605, the data relocation unit 203 reserves an unallocated physical block as a newly allocated area on the low-speed disk pool 112 (step S606). Subsequently, the data recorded in the physical block on the high-speed disk pool 102 currently assigned to the address 304b of the logical disk ID 304a is read and stored in the cache memory 140 (step S607).

  Then, the data relocation unit 203 writes the data stored in the cache memory 140 in step S607 to the physical block on the low speed disk pool 112 secured in step S606 (step S608).

  In the logical disk allocation table 301, the ID and address of the newly secured low-speed disk pool 112 are recorded and updated in the storage pool identification information 301d and the storage address information 301e as the allocation area for the address 301b of the logical disk 301a. (Step S609). Further, the original physical block allocated to the high-speed disk pool 102 is released as an unallocated area (step S610).

  Finally, the value of the no access frequency 304c at the address 304b of the logical disk ID 304a is set to “0” (step S611). As a result, when the address 304b of the logical disk ID 304a is assigned to the low speed disk pool 112, the next time the data relocation unit 203 operates, the no access frequency 304c = “0”. A meaningless operation of moving a physical block allocated to the disk pool 112 to the low-speed disk pool 112 does not occur.

  This completes the determination and processing related to data relocation in a specific block of a specific logical disk. It is confirmed whether or not all the blocks of all the logical disks to be checked are checked, and if not completed, the processing from step S601 is continued with the next block as the check target (step S612).

  In the present embodiment, the frequency of access is high by relocating data from the low-speed disk pool 112 to the high-speed disk pool 102 using the cache memory 140 with respect to the area of the logical disk that has actually been read / written. A logical disk area is allocated to the high-speed disk pool 102. Since the data rearrangement from the low-speed disk pool 112 to the high-speed disk pool 102 is performed in the data read / write process, the load on the storage device and the physical disk is very small.

  Then, the access frequency is measured for each block, a block with low access frequency is determined as a relocation target, and relocation from the high speed disk pool 102 to the low speed disk pool 112 is performed. An area is allocated to the low speed disk pool 112. Thereby, data can be rearranged appropriately between the high-speed disk pool 102 and the low-speed disk pool 112 according to the access frequency.

  Data relocation from the low-speed disk pool 112 to the high-speed disk pool 102 is performed in the data read / write process as described above, and as a routine periodically executed by a timer or the like, the high-speed disk pool 102 to the low-speed disk pool 112 are routinely executed. Only relocation processing is added. Accordingly, in this case as well, the load on the storage device and the physical disk can be reduced.

  Although the present invention has been described with the specific embodiments shown in the drawings, the present invention is not limited to the embodiments shown in the drawings, and is known so far as long as the effects of the present invention are achieved. It goes without saying that any configuration can be adopted.

  It can be used in an apparatus in which a plurality of high-speed physical disks and low-speed physical disks are combined. It does not have to be a dedicated storage device.

10 host computer 100 storage device 101, 101a, 101b high speed physical disk 102 high speed disk pool 103a, 103b, 113a, 113b physical block 111, 111a, 111b low speed physical disk 112 low speed disk pool 120 logical disk 130 host adapter 140 cache memory 150 control Unit 201 Data reading means 202 Data writing means 203 Data relocation means 301 Logical disk allocation table 302 Logical disk reading frequency table 303 Logical disk writing frequency table 304 No access status recording table

Claims (14)

A first disk pool including one or more first physical disk devices;
A second disk pool including one or more second physical disk devices having a data access speed lower than that of the first physical disk device;
Cache memory for temporarily storing data,
Corresponding to a read request and a write request from the host computer to the logical disk using a logical disk allocation table that associates physical blocks on the first disk pool and the second disk pool with addresses on the logical disk. And a control unit for reading and writing data to and from the physical block,
In response to the read request for the address on the logical disk, the control unit reads data from the first physical block and stores it in the cache memory, and the first physical block is stored in the second disk. When the data is on the pool, the data that releases the first physical block and updates the logical disk allocation table after writing the data on the cache memory to the second physical block on the first disk pool A storage apparatus comprising a reading unit.   In response to the write request for the address on the logical disk, the control unit stores the data received from the host computer on the cache memory and corresponds to the address on the logical disk. If the third physical block is on the second disk pool, the third physical block is released after the data on the cache memory is written to the fourth physical block on the first disk pool. The storage apparatus according to claim 1, further comprising data writing means for updating the logical disk allocation table. The data reading means records in the read frequency table provided in advance in the control unit that data has been read for the address on the logical disk from which data was read in response to the read request,
The data writing means records in the write frequency table provided in advance in the control unit that data has been written to the address on the logical disk to which data has been written in response to the write request. The storage apparatus according to claim 2, wherein:   The control unit periodically inspects the read frequency table and the write frequency table, and sets the no-access frequency for the addresses on the logical disk that have not been read and written to data within a predetermined interval. The storage apparatus according to claim 3, further comprising a data rearrangement unit that records in a no-access frequency table provided in advance in the control unit.   If the no-access frequency at the address on the logical disk assigned to the fifth physical block on the first disk pool exceeds a predetermined threshold, the data rearrangement means exceeds the second disk pool. After writing the data on the fifth physical block to the sixth physical block above, releasing the fifth physical block and updating the logical disk allocation table to erase the data with no access frequency The storage apparatus according to claim 4, wherein:   In the read frequency table and the write frequency table, the data rearrangement means reads and writes data to an address on the logical disk where the data is read and written within a predetermined interval. 6. The storage apparatus according to claim 4, wherein the data to the effect is erased. A first disk pool including one or more first physical disk devices, and a second disk pool including one or more second physical disk devices having a data access speed lower than that of the first physical disk device. A method of rearranging data in a storage device having a cache memory for temporarily storing data,
Upon receiving a read request for the address on the logical disk from the host computer, the logical disk allocation table prepared in advance on the first disk pool or the second disk pool corresponding to the address on the logical disk A first physical block specifying step for specifying a first physical block;
A data read step of reading data from the first physical block and storing it in the cache memory;
A first disk pool specifying step for specifying in which of the first disk pool and the second disk pool the first physical block exists;
When the first physical block is on the second disk pool, the data on the cache memory is written to the second physical block on the first disk pool, and then the first physical block is And a first data relocation step of releasing and updating the logical disk allocation table. Upon receiving a write request for the address on the logical disk from the host computer, a third on the first disk pool or the second disk pool corresponding to the address on the logical disk is determined by the logical disk allocation table. A second physical block specifying step for specifying a physical block of
A data cache step of storing data received from the host computer on the cache memory;
A second disk pool specifying step for specifying in which of the first disk pool and the second disk pool the third physical block exists;
When the third physical block is on the second disk pool, after writing the data on the cache memory to the fourth physical block on the first disk pool, the third physical block The data relocation method according to claim 7, further comprising a second data relocation step of releasing and updating the logical disk allocation table. The first data rearrangement step records in the read frequency table provided in advance that data has been read from the address on the logical disk from which data was read in response to the read request. Including a reading frequency recording step,
In the second data rearrangement step, the fact that data has been written to the address on the logical disk where data was written in response to the write request is recorded in a write frequency table provided in advance. The data rearrangement method according to claim 8, further comprising a writing frequency recording step. A frequency inspection step of periodically inspecting the read frequency table and the write frequency table;
In the read frequency table and the write frequency table, no access frequency is recorded in the no access frequency table provided in advance for the addresses on the logical disk that have not been read or written in the data within a predetermined interval. The data relocation method according to claim 9, further comprising an access frequency recording step. The non-access frequency table is periodically inspected to detect a fifth physical block corresponding to an address on the logical disk that is assigned to the first disk pool and whose non-access frequency exceeds a predetermined threshold. An access-free inspection process,
After the data on the fifth physical block is written to the sixth physical block on the second disk pool, the fifth physical block is released to update the logical disk allocation table and the no-access frequency The data rearrangement method according to claim 10, further comprising a third data rearrangement step of erasing the data.   In the third data rearrangement step, in the read frequency table and the write frequency table, data read and write are performed with respect to addresses on the logical disk where the data read and the data are written within a predetermined interval. 12. The data rearrangement method according to claim 10 or 11, further comprising a frequency erasing step of erasing data indicating that the data exists. A first disk pool including one or more first physical disk devices, and a second disk pool including one or more second physical disk devices having a data access speed lower than that of the first physical disk device. A computer for controlling a storage device having a cache memory for temporarily storing data;
Upon receiving a read request for the address on the logical disk from the host computer, the logical disk allocation table prepared in advance on the first disk pool or the second disk pool corresponding to the address on the logical disk A first physical block specifying process for specifying a first physical block;
A data read process for reading data from the first physical block and storing it in the cache memory;
A first disk pool specifying process for specifying in which of the first disk pool and the second disk pool the first physical block exists;
When the first physical block is on the second disk pool, the data on the cache memory is written to the second physical block on the first disk pool, and then the first physical block is A data rearrangement program that executes a first data rearrangement process that releases and updates the logical disk allocation table. Upon receiving a write request for the address on the logical disk from the host computer, a third on the first disk pool or the second disk pool corresponding to the address on the logical disk is determined by the logical disk allocation table. A second physical block specifying process for specifying the physical block;
Data cache processing for storing data received from the host computer on the cache memory;
A second disk pool specifying process for specifying in which of the first disk pool and the second disk pool the third physical block exists;
When the third physical block is on the second disk pool, after writing the data on the cache memory to the fourth physical block on the first disk pool, the third physical block 14. The data rearrangement program according to claim 13, wherein a second data rearrangement process for releasing and updating the logical disk allocation table is executed.

Images