JP4862841B2 - Storage apparatus, system, method, and program - Google Patents

Description

  The present invention relates to a storage apparatus, system, method, and computer program provided with a disk array unit.

  Computer systems have become larger and the amount of data to be stored has become enormous. In general, storage devices are used for data storage, but in order to reduce initial costs, storage devices with a small storage capacity are initially introduced, and additional storage devices are installed as the system scale increases. Sometimes.

  For this reason, a scalable storage device has been developed that can be implemented at a low price at a low cost, and can realize a large-scale configuration by dynamically adding resources later.

  The following is an analysis of the related art according to the present invention.

  As described above, in a system that enables scalable configuration expansion and stores data in a distributed manner for each disk array unit, the disk array unit that has received a write / read request from the host computer and the data corresponding to the request In a case of access across a plurality of disk array units, such as when the disk array unit storing the data is different, communication between the disk array units is required, so that a response delay occurs.

  Accordingly, it is an object of the present invention to provide an apparatus for reducing an access delay that spans a plurality of disk array units.

  The invention disclosed in the present application is generally configured as follows in order to solve the above problems.

  According to the present invention, there is provided a storage apparatus including a plurality of disk array units whose units are connected to each other, wherein the disk array unit temporarily stores write data to the disk and / or read data from the disk. And a means for moving data between the cache memory of one disk array unit and the cache memory of another disk array unit based on the respective access statuses from the plurality of disk array units, Is provided.

  In the present invention, the disk array unit is a disk that manages a correspondence between a disk number that uniquely identifies a plurality of disks of the plurality of disk array units and a unit number of a cache memory that holds data of the disk number. A correspondence table and access monitoring means for monitoring the access status of the disks from the plurality of disk array units, wherein the number of accesses per unit time from other disk array units is predetermined in the disk array unit. If there is a disk number exceeding the specified threshold, the unit number of the cache memory of the disk correspondence table is changed to the unit number of the cache memory of the other disk array unit, and the cache memory of the disk array unit is changed. The data corresponding to the disk number, moved to the cache memory of the other disk array unit.

  In the present invention, the disk correspondence table includes a logical disk number that uniquely identifies a plurality of disks of the plurality of disk array units corresponding to a logical address space, and a physical disk number of a disk corresponding to the logical disk number. And the correspondence relationship between the unit number of the cache memory holding the data corresponding to the logical disk number.

  In the present invention, the disk array unit manages an access statistic that manages a disk number, an access status of each of the plurality of disk array units with respect to the disk number, and a threshold value of an access count that is a reference for cache memory change. An information table is provided, and each access status from the plurality of disk array units is managed using the access statistical information table.

  In the present invention, the access statistical information table includes a logical disk number uniquely identified corresponding to a logical address space of a disk of the plurality of disk array units, and the plurality of disk arrays for the data of the logical disk number. The correspondence relationship between the number of accesses from a unit per unit time and the threshold value of the number of accesses per unit time serving as a reference for changing the cache memory may be stored.

  In the present invention, when the disk number data is transferred to the cache memory of the other disk array unit in the one disk array unit and the disk number data exists in the cache memory, the disk number data May be transferred to the cache memory of the other disk array unit, and the data of the disk number may be deleted from the cache memory of the one disk array unit.

  In the present invention, when a write request to the disk number is received while moving the data of the disk number from the cache memory of the migration source disk array unit to the cache memory of the migration destination disk array unit, In the disk array unit, when the data of the disk number is not transferred between the cache memories, the data of the disk number is transferred from the cache memory of the source disk array unit to the cache of the destination disk array unit. The disk number data is copied to the memory, and the disk number data is deleted from the cache memory of the source disk array unit to move the disk number data between the cache memories, and then the write data is transferred to the corresponding disk. Aleille Stored in the cache memory in the dot and write data on the cache memory may be configured to write to the disk corresponding to the disk number.

In the present invention, when the disk number data is being transferred from the cache memory of the migration source disk array unit to the cache memory of the migration destination disk array unit, when the disk number data read request is received, the migration source If there is data to be read in the cache memory of the disk array unit,
When the migration of the data between the cache memories has not been performed, the cache disk of the migration source disk array unit is copied from the cache memory of the migration source disk array unit to the cache memory of the migration destination disk array unit. By deleting the data of the disk number from the memory, the disk number data is moved between cache memories,
When there is no data to be read in the cache memory of the migration source disk array unit, the data is read from the disk of the migration destination disk array unit and stored in the cache memory of the migration destination disk array unit, Data stored in the cache memory may be returned to the request source as read data.

  In the present invention, a logical disk number is used as the disk number.

  According to the present invention, a disk memory unit that includes a cache memory that temporarily holds data to be written to the disk and / or data to be read from the disk, and is interconnected with one or another disk array unit, There is provided a disk array unit comprising means for moving data between the cache memory of the disk array unit and the cache memory of another disk array unit based on the access status from the disk array unit.

  According to the present invention, a computer having a disk array unit that includes a cache memory that temporarily holds data to be written to the disk and / or data to be read from the disk, and is interconnected with one or another disk array unit. A program for executing a process of moving data between the cache memory of the disk array unit and the cache memory of another disk array unit based on the access status from each disk array unit is provided.

According to the present invention, there is provided a disk array unit that includes a cache memory that temporarily holds write data to the disk and / or read data from the disk, and is interconnected with one or more other disk array units. ,
There is provided a storage control method for moving data between the cache memory of the disk array unit and the cache memory of another disk array unit based on the access status from each disk array unit.

  According to the present invention, a plurality of disk array units are connected as in the case where a disk array unit that has received a write / read request from a host computer is different from a disk array unit that stores data corresponding to the request. Reduce access delays that span.

The above-described present invention will be described below with reference to the accompanying drawings in order to explain in more detail.
In the present invention, the status of access to the logical disk from each of the plurality of disk array units is monitored, and the logical disk number and the logical disk number for the logical disk whose number of accesses per unit time exceeds a predetermined threshold value. The unit number of the cache memory in the logical disk correspondence table that manages the correspondence with the unit number of the cache memory is changed to instruct the movement of data between the cache memories. By moving the data stored in the cache memory in the disk array unit to the cache memory of the disk array unit that accesses the data most frequently, the disk array units can be Arrange the data so that communication is unnecessary.

  The storage apparatus includes a plurality of disk array units (201) including a host control unit (210), an intra-unit connection switch (220), a cache memory (230), a disk control unit (240), and a physical disk (250). The disk array units (201) can be connected to each other by an inter-unit connection switch (202).

  When the cache memory of one disk array unit is frequently accessed from the other disk array unit side via the inter-unit connection switch (202), the cache corresponding to the logical disk number in the logical disk correspondence table Change the unit number of the memory to the unit number of the cache memory of the other disk array unit, and move the data corresponding to the logical disk number of the cache memory of one disk array unit to the cache memory of the other disk array unit Let Thereafter, in the other disk array unit, the data of the path logical disk number is accessed to the cache memory in the other disk array unit, and access between the disk array units (201) occurs. Do not.

In one aspect of the present invention, the cache memory unit number in the logical disk correspondence table is changed by monitoring the access status to each logical disk from the host control unit (210) of each disk array unit (201), and accessing statistics. You may make it perform automatically based on information. Alternatively, in another aspect of the present invention, the access status is displayed on the maintenance terminal (300), and the correspondence table between the cache memory (230) and the physical disk (250) is indicated by an instruction from the maintenance terminal (300). May be changed.
Examples will be described below.

  FIG. 1 is a diagram showing a system configuration of an embodiment of the present invention. The storage apparatus 200 includes a plurality of disk array units 201. Each disk array unit 201 receives and analyzes a request (command) from the host computer (101 to 104), issues an access request, and responds with an access result. From the host control unit 210 to be returned to the host computers (101 to 104), the intra-unit connection switch 220 for switching control between the host control unit 210 and the cache memory 230, the write data to the physical disk 250, and the physical disk 250 A cache memory unit 230 that temporarily holds the read data, a disk control unit 240 that controls writing and reading of data to and from the physical disk 250, and a physical disk unit 250 including a plurality of disks.

  In FIG. 1, each disk array unit 201 has a host control unit 210, a cache memory 230, and a disk control unit 240 in a double redundant configuration, one of which is an active system and the other is a standby system. When a failure occurs, switching to the standby system is performed. The intra-unit connection switch 220 connects the active host control unit 210 to the active cache memory 230. Each of the duplicated cache memories 230 can access any of the duplicated disk control units 240. The host control units 210 of the disk array units 0, 1, 2, 3 (201) are connected to the host computers 101, 102, 103, 104, respectively.

  The host control unit 210 in one disk array unit 201 is communicatively connected to the host control unit 210 in another disk array unit 201. The host control unit 210 in one disk array unit 201 is also communicatively connected to the host control unit 210 (standby system) in one disk array unit 201.

  The disk array units 201 are connected by an inter-unit connection switch 202. That is, the intra-unit connection switch 220 of the disk array unit 0 (201) and the disk array unit 1 (201) is connected to the inter-unit connection switch 202, and the disk array unit 2 (201) and the disk array unit 3 (201). The intra-unit connection switch 220 is connected to the inter-unit connection switch 202, and the inter-unit connection switches 202 of the disk array units 0 to 3 (201) can be connected to each other.

  In FIG. 1, four disk array units are used. However, in the present invention, the number of disk array units is not limited to four, and may be three or less or five or more. Of course.

  FIG. 2 is a diagram showing an example of a detailed configuration of the disk array unit 201 of FIG. Referring to FIG. 2, the host control unit 210 receives a command from the host computer and analyzes the host command control unit 211, a data transfer control unit 212 that controls data (write / read data) transfer, and a host computer. A host access monitoring unit 213 that monitors access, and an inter-control unit communication control unit 214 that controls communication between other unit host control units 210 (inter-control unit communication control unit 214). ing.

  The cache memory unit 230 includes a logical disk correspondence table 231 that holds correspondences between logical disk numbers, physical disk unit numbers, and cache memory unit numbers, a cache memory 232, and host access monitored by the host access monitoring unit 213. A host access statistical information table 233 for managing the statistical information.

  The disk control unit 240 receives a command issued by the host command control unit 211, and controls the transfer of write data to the physical disk 250 and read data read from the physical disk 250, which controls writing / reading. A data transfer control unit 242 that controls the copying of data in the cache memory between the cache memory of another disk array unit, an inter-control unit communication control unit 214 of the host control unit 210, And an inter-control unit communication control unit 244 that controls communication with the disk control unit 240 of another disk array unit.

  FIG. 3 is a diagram schematically showing the contents of the logical disk correspondence table 231 in the cache memory unit 230 of FIG.

  The logical disk correspondence table 231 is a table showing correspondences between logical disk numbers, physical disk unit numbers, and cache memory unit numbers.

  In this embodiment, the logical disk number is assigned to the physical disk units of the physical disk units 250 of the disk array units 0, 1, 2, and 3, for example, by serial numbers. The correspondence between the logical address number and the unit number of the physical disk is common among the disk array units 0, 1, 2, and 3. Therefore, the contents of the logical disk correspondence table 231 include the disk array units 0, 1, 2, 3 is the same. Although not particularly limited, the unit number of the cache memory corresponds to the disk array unit number (the unit numbers 0, 1, 2, and 3 of the cache memory are disk array units 0, 1, 2, and 3). In the example shown in FIG. 3, the physical address space of the physical disk 250 (physical disk number 0) of the disk array unit 0 is a logical address space composed of four logical disks of logical disk numbers 0000, 0001, 0002, and 0003. Correspondingly, the data blocks in the logical address space of the logical address numbers 0000 to 0003 are respectively held in the data areas of the cache memory (unit number 0) of the disk array unit 0. Similarly, the physical address space of the physical disk (physical disk number 1) of the disk array unit 1 corresponds to a logical address space composed of four logical disks of logical disk numbers 0004, 0005, 0006, and 0007. Data blocks in the logical address space of logical address numbers 0004 to 0007 are held in the data areas of the cache memory (unit number 0) of the disk array unit 1, respectively. The same applies to logical disk numbers 0008 and after. Although not particularly limited, in the example shown in FIG. 3, the physical disk number is an identification number for uniquely identifying the physical disk 250 (the physical disk numbers 0, 1, 2, and 3 are the disk array units 0, 1, 2 and 3), the logical disk number uniquely identifies a partial space of the entire logical address space formed by a plurality of units of physical disks 250.

  The host command control unit 211 of the host control unit 210 that has received the write / read request (access request command) from the host computer refers to the logical disk correspondence table 231 from the access target address to determine which logical disk to be accessed. It is checked whether it corresponds to the physical disk (unit number) and the cache memory (unit number).

  FIG. 4 is a diagram schematically illustrating an example of the host access statistical information table 233. The host access statistical information table 233 is a table indicating the number of accesses per unit time from the host control unit 210 of each disk array unit for each logical disk number. The host access monitoring unit 213 monitors the status of write / read requests from the host computer and updates it periodically.

  The host access statistical information table 233 includes a logical disk number, the number of accesses per unit time from the host control unit 210 of the disk array units 0, 1, 2, and 3, and a cache change threshold (a cache memory for storing data blocks is changed. The threshold of the number of accesses per unit time used as a reference) is held as one entry.

  FIG. 5 is a diagram showing an operation flow when a write request is received from the host computer in one embodiment of the present invention.

  Receiving the write request from the host computer, the host control unit 210 searches the unit number of the cache memory 232 corresponding to the logical disk number with reference to the logical disk correspondence table 231 (step S101).

  Next, the host control unit 210 acquires the access right of the data block for writing to the corresponding logical disk (step S102). The host control unit 210 stores the data (write data) received from the host computer in the cache memory 232 in the corresponding disk array unit (step S103).

  Next, the host control unit 210 releases the access right of the data block for writing to the corresponding logical disk (step S104), and transmits a response to the host computer (step S105).

  Thereafter, the disk control unit 240 writes unwritten data in the cache memory 232 to the physical disk 250 asynchronously with the write request from the host computer (step S106).

  In FIG. 1, when the host computer 101 accesses a logical disk configured from the physical disk 250 of the disk array unit 0 (201), the host control unit 210, the intra-unit connection switch 220 in the disk array unit 0, The data flow is the cache memory 230.

  On the other hand, when the host computer 101 accesses a logical disk composed of the physical disks 250 of the disk array unit 2 (201), the host control unit 210 in the disk array unit 0, the intra-unit connection switch 220, between units The data flows of the connection switch 202, the intra-unit connection switch 220 in the disk array unit 2, and the cache memory 230 in the disk array unit 2.

  For this reason, a plurality of switches are inserted in between, and an access delay occurs as compared with the case of only in the disk array unit.

  FIG. 6 is a flowchart showing an operation flow when a read request is received from the host computer in the embodiment of the present invention.

  Receiving the read request from the host computer, the host control unit 210 searches for the unit number of the cache memory 232 corresponding to the logical disk number (step S201).

  Next, the host control unit 210 acquires the access right of the data block for reading the corresponding logical disk (step S202), and checks whether the corresponding data block exists in the cache memory 232 (cache hit / cache miss) ( Step S203).

  When the data block to be read does not exist in the cache memory 232 (No in step S203: cache miss), the host control unit 210 instructs the disk control unit 240 in the corresponding unit to read data from the physical disk 250. Upon receiving the instruction (Step S204), the disk controller 240 reads data from the physical disk 250 and transfers the data to the cache memory 232 (Step S205).

  Next, the host control unit 210 transfers the data stored in the cache memory 232 to the host computer (step S206), releases the access right to the data block of the corresponding logical disk (step S207), and reads it to the host computer. A response to the request is transmitted (step S208).

  In FIG. 1, when the host computer 101 accesses a logical disk configured from the physical disk 250 of the disk array unit 0 and a cache hit occurs, the cache memory 230 in the disk array unit 0, the intra-unit connection switch 220, The host controller 210 has a data flow.

  On the other hand, when the host computer 101 accesses a logical disk composed of the physical disks 250 of the disk array unit 2 and a cache hit occurs, the cache memory 230 in the disk array unit 2 and the units in the disk array unit 2 The data flow includes the connection switch 220, the inter-unit connection switch 202, the intra-unit connection switch 220 in the disk array unit 0, and the host control unit 210 in the disk array unit 0.

  For this reason, a plurality of switches are inserted in between, and an access delay occurs as compared with the case of only in the disk array unit.

  FIG. 7 is a diagram schematically illustrating another example of the host access statistical information table 233. In the host access statistical information table 233, the access (4000) from the disk array unit 0 to the logical disk 0008 exceeds the cache change threshold (3000), and the access from other disk array units (units 1, 3) Access from 0 and 1000 from unit 2 are below the cache change threshold.

  In this case, it is possible to speed up the writing / reading operation if the data of the logical disk 0008 is stored in the cache memory 232 in the disk array unit 0.

  FIG. 9 is a flowchart showing an operation flow of access status monitoring processing by the host access monitoring unit 213 in an embodiment of the present invention.

  The host access monitoring unit 213 checks the access status from the host control unit 210 of each unit of the logical disk with reference to the host access statistical information table 233 (step S301), and the number of accesses per unit time is a certain time or more. It is checked whether or not the threshold (cache change threshold) is exceeded (step S302).

  If the number of accesses per unit time exceeds a threshold value for a certain time or longer (Yes branch in step S302), the unit number of the corresponding cache memory in the logical disk correspondence table 231 is changed as shown in FIG. 8, the unit number of the cache memory is changed from 2 to 0), and the data in the cache memory 232 is moved to the inter-cache data copy control unit 253 in the disk control unit 240 (the data in the cache memory 232 of a certain unit is separated). (In step S304), the processes in steps S301 to S304 are repeated until all logical disks have been checked (No branch in step S305).

  In this embodiment, the access status monitoring process by the host access monitoring unit 213 is executed at predetermined time intervals. Alternatively, it may be executed by using a predetermined event occurrence as a trigger.

  FIG. 10 is a flowchart showing an operation flow of data movement processing between cache memories in one embodiment of the present invention.

  The cache memory of the source unit is checked (step S401).

  If there is data on the relevant logical disk (Yes in step S402), the right to access the relevant data block on the relevant logical disk is acquired (step S403), and the relevant data block on the relevant logical disk is copied to the cache memory of the destination unit. (Step S404).

  At this time, the cache memory unit 230 of the migration source disk array unit, the intra-unit connection switch 220 of the migration source disk array unit, the inter-unit connection switch 202, the intra-unit connection switch 220 of the migration destination disk array unit, and the migration destination disk array unit The data flow is the cache memory 230.

  After the data copy between the cache memories is completed, the corresponding data block is deleted from the cache memory 232 of the migration source disk array unit (step S405), and the access right to the data block of the corresponding logical disk is released (step S406).

  This process is repeated until the data block of the relevant logical disk disappears from the cache memory of the migration source unit.

  In this embodiment, by exclusively controlling the access right to the data block being moved in this way, consistency can be maintained even when the corresponding data block is accessed during data movement.

  FIG. 11 is a flowchart showing an operation flow when a write request to the target logical disk is received from the host computer during data movement between cache memories in one embodiment of the present invention.

  Receiving the write request from the host computer, the host controller 210 refers to the logical disk correspondence table 231 and searches for the unit number of the cache memory corresponding to the logical disk number (step S501).

  Next, the access right of the data block to be written to the logical disk is acquired (step S502), and it is checked whether or not the data movement of the data block has been performed (step S503).

  If migration has not been performed (Yes in step S503), the relevant data block of the relevant logical disk is copied from the cache memory of the migration source unit to the cache memory of the migration destination unit (step S504), and the relevant data is transferred from the cache memory of the migration source unit. The block is deleted (step S505).

  Next, the data received from the host computer is stored in the cache memory in the corresponding unit (step S506). If it is determined in step S503 that the movement is performed (No), the process branches to step S506.

  The access right of the data block for writing to the relevant logical disk is released (step S507), and a response is transmitted to the host computer (step S508).

  Thereafter, the disk control unit 240 writes disk unwritten data in the cache memory to the physical disk 250 asynchronously with the write instruction from the host computer (step S509).

  FIG. 12 is a flowchart showing an operation flow when a read request to the target logical disk is received from the host computer during data movement between cache memories in one embodiment of the present invention.

  Receiving the read request from the host computer, the host control unit 210 searches for the unit number of the cache memory corresponding to the logical disk number (step S601).

  Next, the access right of the data block for reading the corresponding logical disk is acquired (step S602), and it is checked whether the corresponding data block exists in the cache memory (step S603).

  If there is a target data block in the cache memory (Yes in step S603; cache hit), and the data movement of the data block has not been performed (Yes in step S604), the corresponding data block on the corresponding logical disk is cached in the source unit. The data is copied from the memory to the cache memory of the migration destination unit (step S605), and the corresponding data block is deleted from the cache memory of the migration source unit (step S606).

  If there is no target data block in the cache memory (No in step S603; cache miss), the disk controller 240 in the migration destination unit is instructed to read data from the physical disk 250 (step S607).

  Upon receiving the instruction, the disk controller 240 in the migration destination unit reads the data from the physical disk 250 and transfers the data to the cache memory of the migration destination unit (step S608).

  Next, the host control unit 210 transfers the data stored in the cache memory to the host computer (step S609), releases the access right to the data block of the corresponding logical disk (step S610), and sends a response to the host computer. (Step S611).

  After the inter-cache memory data movement processing is completed, even when accessing a logical disk composed of physical disks under other disk array units, if there is a cache hit, the host is not connected via the inter-unit connection switch. Write / read processing from a computer can be performed. For any logical disk, a write / read operation from the host computer can be executed at high speed.

  According to this embodiment, the cache memory is rearranged so that the access between the disk array units does not occur in response to a write / read request from the host port, so that the access delay between the disk array units is reduced. Therefore, it is possible to provide a high-performance and scalable storage device.

  In the above embodiment, the host controller 210 monitors the access status to each logical disk, and automatically changes the correspondence table between the cache memory and the physical disk and moves the cache data based on the statistical information. However, the access status may be displayed on the maintenance terminal 300 in FIG.

  It should be noted that the embodiments and examples can be changed and adjusted within the scope of the entire disclosure (including claims) of the present invention and based on the basic technical concept. Various combinations and selections of various disclosed elements are possible within the scope of the claims of the present invention. That is, the present invention of course includes various variations and modifications that could be made by those skilled in the art according to the entire disclosure including the claims and the technical idea.

It is a figure which shows the structure of one Example of this invention. It is a figure which shows the structure of the disk array unit in one Example of this invention. It is a figure which shows an example of the logical disk correspondence table in one Example of this invention. It is a figure which shows an example of the host access statistical information table in one Example of this invention. It is a flowchart which shows the procedure of the write-in process in one Example of this invention. It is a flowchart which shows the procedure of the read-out process in one Example of this invention. It is a figure which shows an example of the host access statistical information table in one Example of this invention. It is a figure which shows an example of the logical disk corresponding | compatible table in one Example of this invention. It is a flowchart which shows the procedure of the access condition monitoring process in one Example of this invention. It is a flowchart which shows the procedure of the cache data movement process in one Example of this invention. It is a flowchart which shows the procedure of the write processing in the cache data movement in one Example of this invention. It is a flowchart which shows the procedure of the read process in the cache data movement in one Example of this invention.

Explanation of symbols

101-104 Host computer 200 Storage device 201 Disk array unit 202 Inter-unit connection switch 210 Host control unit 211 Host command control unit 212 Data transfer control unit 213 Host access monitoring unit 214 Inter-control unit communication control unit 220 Inter-unit connection switch 230 Cache Memory unit 231 Logical disk correspondence table 232 Cache memory 233 Host access statistical information table 240 Disk control unit 241 Disk command control unit 242 Data transfer control unit 243 Inter-cache copy control unit 244 Inter-control unit communication control unit 250 Physical disk unit 300 For maintenance Terminal

Claims (19)

A storage device comprising a plurality of disk array units connected between units,
The disk array unit is
A cache memory that temporarily holds data to be written to the disk and / or data read from the disk;
A disk correspondence table for managing a correspondence between a disk number that uniquely identifies a plurality of disks of the plurality of disk array units and a unit number of a cache memory that holds data of the disk number;
Access monitoring means for monitoring the access status of the disks from the plurality of disk array units;
In the disk array unit, when there is a disk number in which the number of accesses from other disk array units per unit time exceeds a predetermined threshold, the unit number of the cache memory in the disk correspondence table is set to the other number. Changing to the unit number of the cache memory of the disk array unit, and moving the data corresponding to the disk number of the cache memory of the disk array unit to the cache memory of the other disk array unit ;
A storage apparatus comprising: The disk correspondence table includes a logical disk number that uniquely identifies a plurality of disks of the plurality of disk array units corresponding to a logical address space;
The physical disk number of the disk corresponding to the logical disk number;
The logic for storing a correspondence between the unit number of the cache memory for holding data corresponding to the disk number, the storage device according to claim 1, wherein a. The disk array unit includes the disk number, each of the access status from the plurality of disk array units for the disk number, the access statistical information table for managing a threshold number of accesses as a reference of the cache memory changes, the Prepared,
The storage apparatus according to claim 1 or 2 , wherein access statuses from the plurality of disk array units are managed using the access statistical information table. The access statistical information table includes a logical disk number uniquely identified corresponding to a logical address space of a disk of the plurality of disk array units, and a unit time from the plurality of disk array units for the data of the logical disk number. 4. The storage apparatus according to claim 3 , wherein a correspondence relationship between the number of hits and the threshold value of the number of accesses per unit time serving as a reference for changing the cache memory is stored. In the disk array unit,
When moving the disk number data to the cache memory of the other disk array unit, if the disk number data exists in the cache memory, the disk number data is transferred to the cache memory of the other disk array unit. and, wherein the cache memory of one disk array units, wherein deleting the data disk number, the storage device according to any one of claims 1 to 4, characterized in. When a disk number data is being transferred from the cache memory of the source disk array unit to the cache memory of the destination disk array unit and a write request to the disk number is received,
In the source disk array unit,
If the migration of the disk number data between the cache memories is not performed, the disk number data is copied from the migration source disk array unit cache memory to the migration destination disk array unit cache memory, By deleting the data of the disk number from the cache memory of the source disk array unit, the data of the disk number is moved between the cache memories,
Then, store the write data in the cache memory of the corresponding disk array unit, writes the write data on the cache memory to the disk corresponding to the disk number, any one of claims 1 to 5, characterized in that 1 The storage device according to item. When the disk number data is being transferred from the cache memory of the migration source disk array unit to the cache memory of the migration destination disk array unit, when a read request for the data of the disk number is received,
When there is data to be read in the cache memory of the migration source disk array unit,
When the migration of the data between the cache memories has not been performed, the cache disk of the migration source disk array unit is copied from the cache memory of the migration source disk array unit to the cache memory of the migration destination disk array unit. By deleting the data of the disk number from the memory, the disk number data is moved between cache memories,
When there is no data to be read in the cache memory of the migration source disk array unit, the data is read from the disk of the migration destination disk array unit and stored in the cache memory of the migration destination disk array unit, as read the data stored in the cache memory data is returned to the request source, the storage device according to any one of claims 1 to 6, characterized in that. The use of logical disk number as the disk number, the storage device according to any one of claims 5 to 7, characterized in that. System comprising a storage system according the one or more host computers to any one of claims 1 to 8. A disk array unit that includes a cache memory that temporarily holds write data to the disk and / or read data from the disk and that is interconnected with one or more other disk array units,
A disk correspondence table for managing a correspondence between a disk number that uniquely identifies a plurality of disks of the plurality of disk array units and a unit number of a cache memory that holds data of the disk number;
Access monitoring means for monitoring the status of access to the disk in the disk array unit from the other disk array unit;
In the disk array unit, when there is a disk number in which the number of accesses per unit time from the other disk array unit exceeds a predetermined threshold, the unit number of the cache memory in the disk correspondence table is Means for changing to the unit number of the cache memory of the other disk array unit and moving the data corresponding to the disk number of the cache memory of the disk array unit to the cache memory of the other disk array unit ;
A disk array unit with A cache memory for temporarily holding the read data from the write data and / or disk to disk, one or more interconnected with other disk array unit, a plurality of disks of the plurality of disk array units A disk array unit computer having a disk correspondence table that manages the correspondence between the disk number that uniquely identifies the disk number and the unit number of the cache memory that holds the data of the disk number ;
A process of monitoring the access status of the disk in the disk array unit from the other disk array unit;
In the disk array unit, when there is a disk number in which the number of accesses per unit time from the other disk array unit exceeds a predetermined threshold, the unit number of the cache memory in the disk correspondence table is change the unit number of the cache memory of another disk array unit, the data corresponding to the disk number of the cache memory of the disk array unit, a processing you move in the cache memory of the other disk array units,
A program that executes When moving the disk number data to the cache memory of the other disk array unit, if the disk number data exists in the cache memory, the disk number data is transferred to the cache memory of the other disk array unit. The program according to claim 11 , wherein the computer executes a process of deleting the data of the disk number from the cache memory of the one disk array unit. A disk array unit that includes a cache memory that temporarily holds write data to the disk and / or read data from the disk and that is interconnected with one or more other disk array units,
Managing the correspondence between the disk number uniquely identifying the plurality of disks of the plurality of disk array units and the unit number of the cache memory holding the data of the disk number;
Monitoring the access status of the disk in the disk array unit from the other disk array unit;
In the disk array unit, when there is a disk number in which the number of accesses per unit time from the other disk array unit exceeds a predetermined threshold, the unit number of the cache memory in the disk correspondence table is It is changed to the unit number of the cache memory of another disk array unit, and the data corresponding to the disk number of the cache memory of the disk array unit is moved to the cache memory of the other disk array unit. Storage control method. The disk correspondence table is:
A logical disk number that uniquely identifies a plurality of disks of the plurality of disk array units corresponding to a logical address space; and
The physical disk number of the disk corresponding to the logical disk number;
Wherein storing the correspondence between the unit number of the cache memory for holding data corresponding to the logical disk number, the storage control method according to claim 1 3, wherein a. The disk array unit includes an access statistical information table for managing a disk number, an access status of each of the plurality of disk array units with respect to the disk number, and a threshold value of the number of accesses as a reference for changing the cache memory. 15. The storage control method according to claim 13 , wherein the access status from each of the plurality of disk array units is managed using the access statistical information table.   The access statistical information table includes a logical disk number uniquely identified corresponding to a logical address space of a disk of the plurality of disk array units, and a unit time from the plurality of disk array units for the data of the logical disk number. 16. The storage control method according to claim 15, wherein a correspondence relationship between the number of hits and the threshold value of the number of accesses per unit time serving as a reference for changing the cache memory is stored. In moving the disk number data to the cache memory of the other disk array unit in the one disk array unit, if the disk number data exists in the cache memory, the disk number data is transferred to the other disk array unit. was transferred to the cache memory of the disk array units, the cache memory of the one of the disk array units, and deletes the data of the disk number, according to any one of claims 1 3 to 11 6, characterized in that Storage control method. When a disk number data is being transferred from the cache memory of the source disk array unit to the cache memory of the destination disk array unit and a write request to the disk number is received,
The migration source disk array unit transfers the data of the disk number from the cache memory of the migration source disk array unit to the migration destination disk when the migration of the data of the disk number between cache memories has not been performed. Copy the data of the disk number to the cache memory of the array unit and delete the data of the disk number from the cache memory of the source disk array unit. stored in the cache memory of the corresponding disk array unit, according to any one of the cache write data on the memory write to the disk corresponding to the disk number, claim 1 3 or 17, characterized in that Storage control method. When the disk number data is being transferred from the cache memory of the migration source disk array unit to the cache memory of the migration destination disk array unit, when a read request for the data of the disk number is received,
When there is data to be read in the cache memory of the migration source disk array unit, and when the migration of the data between the cache memories is not performed, the migration destination disk is transferred from the cache memory of the migration source disk array unit. Copying to the cache memory of the array unit, and deleting the data of the disk number from the cache memory of the source disk array unit, to move the data of the disk number between the cache memory,
When there is no data to be read in the cache memory of the migration source disk array unit, the data is read from the disk of the migration destination disk array unit, the data is transferred to the cache memory of the migration destination disk array unit, and the cache the storage control method according to any one of claims 1 3 to 18 as a data read out data stored in the memory and returns to the requester, wherein the.

Images