JP4466530B2 - Disk array device, logical address control method, and program - Google Patents

Description

  The present invention relates to a disk array device including a plurality of disk devices, and more particularly to a disk array device capable of reducing input / output across the disk devices.

  A disk array device provided with a plurality of magnetic disk devices, in which data is distributed and allocated to each magnetic disk device and redundant data is added, can obtain high input / output performance and reliability with an economical configuration. In recent years, it has been actively used. FIG. 12 is a block diagram showing a configuration example of a conventional general disk array device 100, which includes a plurality of disk devices DISK1 to DISK5 such as a magnetic disk device, an instruction control unit 101, a cache control unit 102, A logical / physical disk control unit 103 and a logical / physical correspondence table 104 are provided.

  Incidentally, the host computer 200 connected to the disk array device 100 recognizes that one or more host logical disks 300 as shown in FIG. 13 exist on the disk array device 100. Data for informing the BIOS (basic input / output systm) where the OS (operating system) used area (OS used area) 302 is located at the head of the host logical disk 300. A management area 301 for storing the OS use area 302 start address and size) is provided. This management area 301 is architecture-dependent and has one sector (512 bytes) in the case of a PC / AT compatible machine. Therefore, since the management area 301 is not specific to the OS, it is not affected by the difference between OSs such as Windows (registered trademark) and Linux.

  On the other hand, the disk array device 100 recognizes that the logical disk 400 shown in FIG. 13 exists as a disk corresponding to the host logical disk 300. If the host computer 200 recognizes a plurality of host logical disks, it recognizes the logical disk corresponding to each host logical disk. As shown in the figure, the logical address spaces assigned to the host logical disk 300 and the logical disk 400 are the same. Further, as shown in FIG. 14, the logical disk 400 is divided into stripes of a predetermined size from the start address “address 0”, and is recorded in each of the disk devices DISK1 to DISK5.

  Here, the data in the management area 301 is stored at the head of the logical disk 400, as shown in FIG. 15, and in the disk array device 100, the stripe number of the disk device DISK1, as shown in FIG. Stored in the leading portion of the ST1 stripe (also referred to as stripe ST1). Therefore, when the host computer 200 performs stripe size input / output from the start address of the OS use area 302, the input / output performance is degraded because the input / output crosses the two disk devices DISK1, DISK2. For example, assuming that the size of the management area 301 is 512 bytes and the stripe size is 2048 bytes, and the host computer 200 inputs / outputs the stripe size from the start address “512” of the OS use area 302, the disk device DISK1 Since input / output is performed for addresses 512 to 2047 of the stripe ST1 and addresses 0 to 511 of the stripe ST1 of the disk device DISK2, the input / output performance is degraded.

  Therefore, in order to solve such problems, the following technique has been conventionally proposed (see, for example, Patent Document 1). This conventional technique will be described with reference to FIG.

  In the conventional technique described in Patent Document 1, when an I / O command is issued from the host computer 200 to the disk array device 100, first, an address indicating an access destination included in the I / O command. Is the address of the management area 301. When the address in the I / O instruction is the address of the management area 301, the I / O instruction is issued to the disk array device 100 as it is. On the other hand, if the address in the I / O instruction is not the address of the management area 301, the difference between the stripe size and the size of the management area 301 is added to the address in the I / O instruction. An I / O instruction including a later address is issued to the disk array device 100.

  Now, for example, it is assumed that the size of the management area 301 is 512 bytes and the stripe size is 2048 bytes. In addition, it is assumed that the I / O instruction inputs / outputs a stripe size from the start address “512” in the OS use area 302 and includes an address “512” and a size “2048 bytes”. In this example, since the address included in the I / O instruction is “512 address” and not the address of the management area 301, the host computer changes the address “512” included in the I / O instruction. {512+ (2048−512)} = converted to address 2048, and an I / O instruction including the converted address is issued to the disk array device. By issuing an I / O instruction with such address conversion to the disk array device 100, the disk array device 100 performs input / output only for the stripe ST1 of the disk device DISK2. Output performance is improved.

JP 2001-125754 A

  According to the conventional technique described in Patent Document 1 described above, input / output across the disk device can be reduced and input / output performance can be improved. However, in the conventional technique described in Patent Document 1, when the address included in the I / O instruction is not the address of the management area 301, the address in the I / O instruction is changed to {in the I / O instruction Address + (stripe size−management area size)}, an unused area is formed in the disk device in the disk array device 100, and the capacity efficiency of the disk array device 100 is poor. There is a problem of becoming. In the above example in which the size of the management area 301 is 512 bytes and the stripe size is 2048 bytes, the addresses 512 to 2047 of the stripe ST1 of the disk device DISK1 become unused areas, and the capacity efficiency of the disk array device is poor. There is a problem of becoming.

(Object of invention)
Accordingly, an object of the present invention is to reduce input / output across disk devices without forming an unused area in the disk device of the disk array device.

A first disk array device according to the present invention includes:
In a disk array device having a plurality of disk devices,
Shift that shifts the logical address space allocated to the logical disk recognized by the disk array device backward by the size of the management area provided at the beginning of the host logical disk recognized by the host computer An implementation department;
Of the data on the logical disk stored in the disk device, data other than the data in the management area is shifted forward by the size of the management area on the disk device, and the data in the management area is transferred to the disk. And a logical / physical disk control unit that is moved immediately after the last data of the logical disk on the apparatus.

A second disk array device according to the present invention is the first disk array device,
If the logical address in the I / O instruction issued from the host computer is less than the start address of the logical address space assigned to the logical disk, the logical address in the I / O instruction is changed to the logical address. It is characterized in that a shift calculation unit for converting the final logical address of the disk into an address obtained by subtracting the size of the management area is provided.

A third disk array device according to the present invention is the second disk array device,
A logical / physical correspondence table in which a correspondence relationship between a logical address assigned to the logical disk and a physical address on the disk device is registered;
The shift execution unit has a configuration in which a value corresponding to the size of the management area is added to a logical address registered in the logical / physical correspondence table.

A first logical address control method according to the present invention includes:
A logical address control method in a disk array device comprising a plurality of disk devices,
The logical address space allocated to the logical disk recognized by the disk array device is shifted backward by the size of the management area provided at the beginning of the host logical disk recognized by the host computer,
Further, the logical disk data other than the management area data stored in the disk device is shifted forward by the size of the management area on the disk device, and the management area data is also transferred. The disk device is moved immediately after the last data of the logical disk.

A second logical address control method according to the present invention is the first logical address control method,
If the logical address in the I / O instruction issued from the host computer is less than the start address of the logical address space assigned to the logical disk, the logical address in the I / O instruction is changed to the logical address. The conversion is made to an address obtained by subtracting the size of the management area from the final logical address of the disk.

A third logical address control method according to the present invention is the second logical address control method,
A logical / physical correspondence table in which a correspondence relationship between a logical address assigned to the logical disk and a physical address on the disk device is registered;
The host computer recognizes the logical address space allocated to the logical disk by adding a value corresponding to the size of the management area to the logical address registered in the logical / physical correspondence table. It is characterized by shifting backward by the size of the management area provided at the beginning of the host logical disk.

The first program according to the present invention is:
A program for causing a computer having a plurality of disk devices to function as a disk array device,
The computer,
Shift that shifts the logical address space allocated to the logical disk recognized by the disk array device backward by the size of the management area provided at the beginning of the host logical disk recognized by the host computer Implementation department,
Of the data on the logical disk stored in the disk device, data other than the data in the management area is shifted forward by the size of the management area on the disk device, and the data in the management area is transferred to the disk. It functions as a logical / physical disk controller that moves immediately after the last data of the logical disk on the apparatus.

The second program according to the present invention is the first program,
The computer,
If the logical address in the I / O instruction issued from the host computer is less than the start address of the logical address space assigned to the logical disk, the logical address in the I / O instruction is changed to the logical address. It is made to function as a shift calculation unit for converting to an address obtained by subtracting the size of the management area from the final logical address of the disk.

A third program according to the present invention is the second program,
The computer includes a logical / physical correspondence table in which a correspondence relationship between a logical address assigned to the logical disk and a physical address on the disk device is registered; and
The shift execution unit has a configuration in which a value corresponding to the size of the management area is added to a logical address registered in the logical / physical correspondence table.

[Action]
The shift execution unit in the disk array device is a management area provided at the head of the host logical disk recognized by the host computer, with the logical address space assigned to the logical disk recognized by the disk array device. Shift backward by the size of. As a result, the management area does not exist at the head of the logical disk recognized by the disk array device, so the disk device that was generated because the management area existed at the head of the logical disk. It is possible to reduce input / output across the network.

  In addition, the logical / physical disk control unit in the disk array device manages on the disk device data other than the management area data among the logical disk data stored in each disk device included in the disk array device. The area is shifted forward by the size of the area, and the data in the management area is moved immediately after the final data of the logical disk on the disk device. As a result, all data on the logical disk can be stored on the disk device without forming an unused area on the disk device.

  According to the present invention, it is possible to reduce input / output across disk devices without forming an unused area in the disk device of the disk array device.

  The reason why the I / O across the disk device can be reduced is because the host computer recognizes the logical address space assigned to the logical disk recognized by the disk array device by the shift execution unit. This is because the size is shifted backward by the size of the management area provided at the head portion of. In other words, by performing the above processing in the shift execution unit, the management area does not exist at the head part of the logical disk recognized by the disk array device, so the management area exists at the head part of the logical disk. Therefore, it is possible to reduce the input / output across the disk devices that has occurred.

  Also, the reason why the unused area is not formed in the disk device of the disk array device is that the logical / physical disk control unit manages the data of the logical disk stored in each disk device provided in the disk array device. This is because data other than the data in the area is shifted forward by the size of the management area on the disk device, and the data in the management area is moved immediately after the final data of the logical disk on the disk device.

  Next, embodiments of the present invention will be described in detail with reference to the drawings.

[Description of Configuration of Embodiment]
FIG. 1 is a block diagram showing a configuration example of an embodiment of a disk array device according to the present invention, which includes a disk array device 1, a management device 2, and a host computer 200.

  Similarly to the host computer 200 shown in FIG. 12, the host computer 200 recognizes that one or more host logical disks 300 as shown in FIG. 13 are present on the disk array device 1. The host computer 200 has a function of issuing an I / O command to the disk array device 1. This I / O instruction includes the number of the host logical disk (host logical disk number) to be input / output, the logical address of the head part for input / output, and the size of the data to be input / output.

  The management device 2 has a function of issuing to the disk array device 1 a shift command including a logical disk number (logical disk number) to be logically shifted and a shift amount in accordance with an instruction from the administrator.

  The disk array device 1 includes an instruction control unit 11, a shift information storage unit 12 having a shift information table 13, a shift execution unit 14, a logical / physical correspondence table 15, a logical / physical disk control unit 16, and a disk correspondence A shift calculation unit 17 having a table 18, a cache control unit 19, and a plurality of disk devices DISK 1 to DISK 5 such as magnetic disk devices are included. It is assumed that disk device numbers # DISK1 to # DISK5 are assigned to the disk devices DISK1 to DISK5, respectively. Further, as shown in FIG. 13, the disk array device 1 recognizes that the logical disk 400 corresponding to each host logical disk 300 recognized by the host computer 200 exists on the own device 1.

  When a shift command is input from the management device 2, the command control unit 11 has a function of passing the logical disk number and shift amount included in the shift command to the shift information storage unit 12 as shift information, or from the host computer 200. When a / O command is input, the host logical disk number, logical address, and size included in the I / O command are transferred to the shift calculation unit 17 as I / O information.

  In the shift information table 13 in the shift information storage unit 12, the shift amount of the logical address space allocated to the logical disk is associated with the logical disk number of each logical disk recognized by the disk array device 1. It is registered. FIG. 2 shows an example of the contents of the shift information table 13. The example of FIG. 5 indicates that the logical disk with the logical disk number L1 has a shift amount of 0 bytes, and the logical disk with the logical disk number L2 has a shift amount of 512 bytes. In the initial state, 0 byte is set as the shift amount.

  The shift information storage unit 12 has the following functions.

When the shift information is passed from the instruction control unit 11, it is included in the shift information among the shift amount included in the shift information and the shift amount of each logical disk stored in the shift information table 13. This function obtains the difference from the shift amount specified by the logical disk number.
A function of passing the obtained difference and the logical disk number included in the shift information to the shift execution unit 14.
A function of replacing the shift amount of the corresponding logical disk stored in the shift information table 13 with the shift amount in the shift information after obtaining the difference.
When a shift amount inquiry including a logical disk number is sent from the shift calculation unit 17, the shift amount registered corresponding to the logical disk number is searched from the shift information table 13 and returned to the shift calculation unit 17. Function to do.

  The logical / physical correspondence table 15 registers the correspondence between the host logical disk number of each host logical disk recognized by the host computer 200 and the logical disk number of the logical disk recognized by the disk array device 1. ing. Further, the logical / physical correspondence table 15 registers the correspondence between the logical addresses assigned to the logical disks 400 recognized by the disk array device 1 and the physical addresses on the disk devices DISK1 to DISK5. .

  FIG. 3 shows an example of the contents of the logical / physical correspondence table 15. The example in the figure shows that the host logical disk of host logical disk number HL1 recognized by the host computer 200 and the logical disk of logical disk number L1 recognized by the disk array device 1 correspond to each other. Yes. Further, in the example of the figure, the data corresponding to the stripe size (2048 bytes) from the logical address “0 address, 2048 address, 4096 address, 6144 address” of the logical disk of the logical disk number L1 is the disk devices DISK1, DISK2, DISK3, respectively. , Stored in the stripe of stripe number ST1 of DISK4, and the data corresponding to the stripe size from "address 8192, address 10240, address 12288, address 14336" is stored in the stripe of stripe number ST2 of disk devices DISK5, DISK1, DISK2, and DISK3, respectively. Indicates that it is stored. In the following description, the host logical disk with the host logical disk number HLX may be referred to as the host logical disk HLX, and the logical disk with the logical disk number LY may be referred to as the logical disk LY.

  The shift execution unit 14 has the following functions.

When a shift amount difference and a logical disk number of a logical disk to be shifted are passed from the shift information storage unit 12, it is checked whether the difference is “0”. If the difference is not “0”, A function of shifting the logical address space allocated to the logical disk of the logical disk number backward by the difference.
A function of passing the difference and logical disk number passed from the shift information storage unit 12 to the logical / physical disk control unit 16.

  In the disk correspondence table 18 in the shift calculation unit 17, the correspondence between the host logical disk recognized by the host computer 200 and the logical disk recognized by the disk array device 1 is registered. FIG. 4 is a diagram showing an example of the contents of the disk correspondence table 18. The example in the figure shows that the host logical disks with host logical disk numbers HL1, HL2,... Correspond to the logical disk numbers L1, L2,.

  The shift calculation unit 17 has the following functions.

A function of retrieving a logical disk number corresponding to the host logical disk number from the disk correspondence table 18 when I / O information including the host logical disk number, logical address, and size is passed from the instruction control unit 11.
A function for sending a shift amount inquiry including the logical disk number retrieved from the disk correspondence table 18 to the shift information storage unit 12.
A function of comparing the shift amount (corresponding to the start address of the corresponding logical disk) returned from the shift information storage unit 12 in response to the shift amount inquiry with the logical address in the I / O information.
If the logical address in the I / O information is greater than or equal to the start address of the logical disk, the I / O information is passed to the cache control unit 19 as is, and if the logical address is less than the start address of the logical disk, the I / O information A function of replacing the logical address in the information with a value obtained by subtracting the difference from the final address of the logical disk, and passing the replaced I / O information to the cache control unit 19.

  The cache control unit 19 has a buffer (not shown) made of a semiconductor memory therein, and when the data indicated by the I / O information is present on the buffer, it performs input / output to / from the buffer and does not exist. In this case, it has a function of passing I / O information to the logical / physical disk control unit 16.

  The logical / physical disk control unit 16 has the following functions.

When I / O information is input from the cache control unit 19, the logical / physical correspondence table 15 is used to input / output data to / from the disk devices DISK1 to DISK5. At this time, it also has a function of performing a parity check and a parity change as necessary. This function is a well-known function.
When the logical disk number and the difference of the logical disk to be shifted in the logical address space are passed from the shift execution unit 14, the data other than the difference data from the head of the data of the corresponding logical disk is divided into stripes. The function of storing the divided stripes in order in each of the disk devices DISK1 to DISK5.
A function for adding difference data (management area data) from the beginning of the corresponding logical disk to the last stripe.

  The disk array device 1 having the above-described functions can be realized by a computer. When the disk array device 1 is realized by a computer, for example, the following is performed. A disk, a semiconductor memory, and other recording media on which a program for causing the computer to function as the disk array device 1 is recorded are prepared, and the program is read by the computer. The computer controls its own operation according to the read program, so that the instruction control unit 11, the shift information storage unit 12, the shift execution unit 14, the logical / physical disk control unit 16, and the shift calculation unit 17 are provided on the computer. Realize.

[Description of Operation of Embodiment]
Next, the operation of the present embodiment will be described in detail.

  Now, for example, the host computer 200 recognizes the host logical disk 300 of FIG. 15 as the host logical disk of the host logical disk number HL1, and the disk array device 1 recognizes the logical disk 400 of FIG. 15 as the logical disk of the logical disk number L1. Assuming that At this time, it is assumed that the contents of the shift information table 13, the logical / physical correspondence table 15, and the disk correspondence table 18 are as shown in FIG. 2, FIG. 3, and FIG. 4, respectively. Therefore, as shown in FIG. 15, the data in the management area 301 is stored at the head of the stripe ST1 of the disk device DISK1 at this time, so the host computer 200 starts the address “512” in the OS use area 302. If an input / output with a stripe size (2048 bytes in this embodiment) is performed from the “address”, the input / output across the disk devices DISK1, DISK2 is performed, and the input / output performance is degraded.

  In such a case, the administrator of the disk array device 1 issues a shift command to the disk array device 1 using the management device 2. This shift command includes the logical disk number L1 of the logical disk 400 corresponding to the host logical disk 300, and includes the size (512 bytes) of the management area 301 as the shift amount.

  When a shift command is sent from the management device 2, the command control unit 11 in the disk array device 1 stores the logical disk number L1 and the shift amount “512 bytes” included therein in the shift information storage unit 12 as shift information. hand over.

  When the shift information including the logical disk number L1 and the shift amount “512 bytes” is transferred from the instruction control unit 12 to the shift information storage unit 11, the shift information table 13 changes the logical disk number L1 from the shift information table 13 as shown in the flowchart of FIG. The shift amount registered in association is searched (step S51). At this time, the contents of the shift information table 13 are as shown in FIG. 2, so that “0 byte” is retrieved as the shift amount.

  After that, the shift information storage unit 12 obtains a difference “512 bytes” between the shift amount “0 byte” searched in step S51 and the shift amount “512 bytes” in the shift information (step S52). Furthermore, the shift amount registered in association with the logical disk number L1 in the shift information table 13 is updated to the shift amount “512 bytes” in the shift information (step S53). As a result, the contents of the shift information table 13 are as shown in FIG. Finally, the shift information storage unit 12 notifies the shift execution unit 14 of the difference “512 bytes” obtained in step S52 and the logical disk number L1 in the shift information (step S54).

  When the logical disk number L1 and the difference “512 bytes” are passed from the shift information storage unit 12, the shift execution unit 14 determines whether or not the difference is “0” as shown in the flowchart of FIG. S71). In this example, the difference is “512 bytes”, and the determination result in step S71 is NO. Therefore, the shift execution unit 14 sets the logical address space allocated to the logical disk 400 with the logical disk number L1 as the difference “512 bytes”. Shift backward by the minute (step S72). More specifically, the shift execution unit 14 calculates the difference between the logical address of the logical disk 400 specified by the logical disk name L1 among the logical addresses of the logical disks registered in the logical / physical correspondence table 15. By increasing by “512 bytes”, the logical address space allocated to the logical disk 400 is shifted backward by “512 addresses”. As a result, the contents of the logical / physical correspondence table 15 are updated from those shown in FIG. 3 to those shown in FIG.

  Thereafter, the shift execution unit 14 passes the logical disk number L1 and the difference “512 bytes” to the logical / physical disk control unit 16 (step S73). In addition, also when the determination result of step S71 is YES, the shift execution unit 14 performs the process of step S73.

  When the difference “512 bytes” and the logical disk number L1 are passed from the shift execution unit 14, the logical / physical disk control unit 16 includes the data in the logical disk 400 of the logical disk number L1, as shown in the flowchart of FIG. Then, data excluding data corresponding to the difference “512 bytes” from the head (data whose logical address of the logical disk 400 is data after the 512 address and data other than the management area 301) is divided into stripes (step S91). In step S91, data at addresses 0 to 511 of the logical disk 400 (data in the management area 301) is read from the disk device DISK1 and stored therein. Here, in which stripe of which disk device the data of each logical address of the logical disk 400 is stored is the difference passed from the logical / physical correspondence table 15 shown in FIG. It can be calculated based on “512 bytes”. More specifically, first, the logical addresses “512, 2560” of the logical disk 400 registered in association with the logical disk number L1 in the information of each logical disk registered in the logical / physical correspondence table 15. ,..., Logical disk numbers “# DISK1, # DISK2,...” And stripe numbers “ST1, ST1,. Then, the logical address “0, 2048,...” Obtained by subtracting the difference “512” from the input logical address “512, 2560,. From the logical address “0, 2048,...”, The logical disk number “# DISK1, # DISK2,...” Input from the logical / physical correspondence table 15 and the stripe number “ST1, ST1,. It can be determined in which stripe of which disk device the data of each logical address exists.

  Thereafter, the logical / physical disk control unit 16 sequentially stores the stripes divided in step S91 in each stripe of each of the disk devices DISK1 to DISK5 (step S92). More specifically, based on the logical / physical correspondence table 15 having the contents shown in FIG. 8, the logical / physical disk control unit 16 has one stripe starting from the logical address “address 512, address 2560, address 4608, address 6656”. Are stored in the stripes ST1 of the disk devices DISK1, DISK2, DISK3, and DISK4, respectively, and the data for one stripe starting from the logical addresses “8704, 10752, 12800, and 14848” are respectively stored in the disk devices DISK5 and DISK1. , DISK2, and DISK3 are stored in the stripe ST2. The same processing is performed for data of logical addresses after this.

  Thereafter, the logical / physical disk control unit 16 adds the data (data in the management area 301) of the logical address “addresses 0 to 511” stored inside in step S91 to the end of the final data of the final stripe. (Step S93). As a result, as shown in FIG. 10, the data in the management area 301 is moved to the final stripe (moved to the stripe ST5 of the disk device DISK5). Through the above processing, the start position of the logical disk 400 matches the start position of the OS use area 302, and the data of the OS use area 302 is stored from the head portion of the stripe ST1 of the disk device DISK1. Accordingly, when the host computer 200 inputs / outputs the stripe size from the start address “512 address” of the OS use area 302 thereafter, the input / output is performed only to the disk device DISK1.

  Next, an operation when the host computer 200 issues an I / O command to the host logical disk 300 with the host logical disk number HL1 after the above shift processing is completed will be described.

  When performing input / output to / from the host logical disk 300, the host computer 200 includes the host logical disk number HL1 of the host logical disk 300, the logical address of the head portion for input / output, and the size of the input / output data. An I / O command is issued to the disk array device 1.

  Upon receiving an I / O command from the host computer 200, the instruction control unit 11 in the disk array device 1 uses the host logical disk number HL1, logical address, and size included therein as I / O information, and the shift calculation unit 17 To pass.

  When the I / O information is passed, the shift calculation unit 17 is registered in association with the host logical disk number HL1 in the I / O information from the disk correspondence table 18 as shown in the flowchart of FIG. The disk number is searched, and a shift amount inquiry including the searched logical disk number is transmitted to the shift information storage unit 12 (step S111). Assuming that the contents of the disk correspondence table 18 are as shown in FIG. 4, the shift calculation unit 17 transmits a shift amount inquiry including the logical disk number L1 to the shift information storage unit 12.

  When the shift information storage unit 12 receives a shift amount inquiry including the logical disk number L1 from the shift calculation unit 17, the shift information storage unit 12 searches the shift information table 13 for the shift amount registered in association with the logical disk number L1. The searched shift amount is returned to the shift calculation unit 17. At this time, since the contents of the shift information table 13 are as shown in FIG. 6, the shift information storage unit 12 returns the shift amount “512 bytes” to the shift calculation unit 17. This shift amount corresponds to the start address “512 address” of the logical address space allocated to the logical disk 400.

  When the shift amount “512 bytes” is sent from the shift information storage unit 12, the shift calculation unit 17 recognizes that the start address of the logical address space assigned to the logical disk 400 is “512 address” (step 512). S112). Thereafter, the shift calculation unit 17 compares the start address “address 512” of the logical address space with the logical address included in the I / O information (the logical address of the leading portion for input / output) (step S113). .

  When the logical address in the I / O information is equal to or higher than the start address “512 address” of the logical address space allocated to the logical disk 400 (NO in step S115), the instruction control unit 11 passes the logical address. The I / O information is directly passed to the cache control unit 19 (step S115). On the other hand, when the logical address in the I / O information is less than the start address “512 address” of the logical address space allocated to the logical disk 400 (YES in step S114), the I / O information is in the I / O information. Is converted into an address obtained by subtracting the shift amount “512 bytes” from the final address of the logical address space allocated to the logical disk 400, and the converted I / O information is passed to the cache control unit 19 (step S116). ). That is, when the logical address in the I / O information corresponds to the OS use area 302 (NO in step S114), the process of step S115 is performed, and the logical address in the I / O information is changed to the management area 301. (Step S114 is YES), the process of step S116 is performed.

  When the data indicated by the I / O information is present in the buffer (when hit), the cache control unit 19 performs input / output on the buffer, and when it does not exist (when missed), the logical / physical disk I / O information is passed to the control unit 16.

  When the logical / physical disk control unit 16 receives the I / O information from the cache control unit 19, the logical / physical disk control unit 16 refers to the logical / physical correspondence table 15 to determine which disk device has which data to be input / output. Input / output data.

[Effect of the embodiment]
According to the present embodiment, it is possible to reduce input / output across the disk devices DISK1 to DISK5 without forming unused areas in the disk devices DISK1 to DISK5 of the disk array device 1.

  The reason why the input / output across the disk devices DISK1 to DISK5 can be reduced is that the host computer 200 determines the logical address space assigned to the logical disk 400 recognized by the disk array device 1 by the shift execution unit 14. This is because the size is shifted backward by the size of the management area 301 provided at the top of the recognized host logical disk 300. That is, by performing the above-described processing in the shift execution unit 14, the management area does not exist in the head part of the logical disk 400 recognized by the disk array device 1, and therefore there is a management area in the head part. Therefore, it is possible to reduce the input / output across the disk devices that has occurred.

  The reason why unused areas are not formed in the disk devices DISK1 to DISK5 of the disk array device 1 is that the logical / physical disk control unit 16 is stored in each of the disk devices DISK1 to DISK5 provided in the disk array device 1. Of the data on the logical disk 400, data other than the data in the management area 301 is shifted forward by the size of the management area 301 on the disk devices DISK1 to DISK5, and the data in the management area 301 is logically moved on the disk devices DISK1 to DISK5. This is because the data is moved immediately after the final data on the disk 400.

  The present invention can be applied to a disk array device that includes a plurality of disk devices and reads / writes data associated with the host logical disk recognized by the host computer in association with the logical disk recognized by the disk array device.

1 is a block diagram showing a configuration example of an embodiment of a disk array device according to the present invention. It is a figure which shows the example of the content of the shift information table 13 before performing the shift process of a logical address space with respect to the logical disk L1. It is a figure which shows the example of the content of the logical / physical correspondence table 15 before performing the logical address space shift process with respect to the logical disk L1. 4 is a diagram showing an example of contents of a disk correspondence table 18. FIG. 5 is a flowchart illustrating a processing example of a shift information storage unit 12. It is a figure which shows the example of the content of the shift information table 13 after performing the shift process of a logical address space with respect to the logical disk L1. 5 is a flowchart illustrating a processing example of a shift execution unit 14; It is a figure which shows the example of the content of the logical / physical correspondence table 15 after performing the shift process of a logical address space with respect to the logical disk L1. 4 is a flowchart showing a processing example of a logical / physical disk control unit 16. It is a figure for demonstrating the reason that the input / output across a disk apparatus reduces. 5 is a flowchart illustrating a processing example of a shift calculation unit 17. 1 is a block diagram showing a configuration example of a conventional general disk array device 1. FIG. 3 is a diagram showing a correspondence relationship between a host logical disk 300 recognized by a host computer 200 and a logical disk 400 recognized by a disk array device 1. FIG. It is a figure which shows the correspondence of the disk device DISK1-DISK5 which is a logical disk 400 and a physical disk. FIG. 10 is a diagram illustrating a state in which input / output across a disk device occurs when input / output of a stripe size is performed. It is a figure for demonstrating the prior art described in patent document 1. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1,100 ... Disk array apparatus 11, 101 ... Instruction control part 12 ... Shift information storage part 13 ... Shift information table 14 ... Shift execution part 15, 104 ... Logical / physical correspondence table 16, 103 ... Logical / physical disk control part 17 ... shift calculation unit 18 ... disk correspondence tables 19, 102 ... cache control units DISK1 to DISK5 ... disk device 2 ... management device 200 ... host computer 300 ... host logical disk 301 ... management area 302 ... OS use area 400 ... logical disk

Claims (3)

In a disk array device having a plurality of disk devices,
A logical / physical correspondence table in which a correspondence relationship between a logical address of a logical address space allocated to a logical disk recognized by the disk array device and a physical address on the disk device is registered;
When a shift instruction including the size of the management area provided at the head of the host logical disk recognized by the host computer as a shift amount is input from the management device, it is registered in the logical / physical correspondence table. A shift execution unit that adds a value corresponding to the shift amount included in the shift instruction to a logical address ;
After the shift execution unit updates the contents of the logical / physical table in accordance with the shift instruction , the data on the logical disk stored in the disk device is stored based on the contents of the updated logical / physical table . A logical / physical disk that shifts data other than the management area data forward on the disk device by the shift amount and moves the data in the management area immediately after the last data of the logical disk on the disk device. A control unit;
The logical address in the I / O instruction issued from the host computer after the shift processing and migration processing by the logical / physical disk control unit is completed is the logical address of the management area recognized by the host computer. Whether the logical address in the I / O instruction is compared with the start address of the logical address space allocated to the logical disk recognized from the shift amount in the shift instruction, and the management If it is a logical address of the area, the logical address in the I / O instruction is moved from the final logical address of the data in the management area moved immediately after the final data of the logical disk on the disk device. Shift to convert to an address obtained by adding the logical address in the I / O instruction to the address subtracted by the size of the management area The disk array apparatus being characterized in that a calculation unit. A logical address control method in a disk array device having a plurality of disk devices and a logical / physical correspondence table ,
In the logical / physical correspondence table, a correspondence relationship between a logical address in a logical address space allocated to a logical disk recognized by the disk array device and a physical address on the disk device is registered, and
When the shift execution unit receives a shift instruction including the size of the management area provided at the head part of the host logical disk recognized by the host computer as a shift amount from the management device, the logical / physical correspondence table A first step of adding a value corresponding to the shift amount included in the shift instruction to the logical address registered in
The logical / physical disk control unit updates the contents of the logical / physical table in the first step, and then stores the logical / physical table stored in the disk device based on the contents of the updated logical / physical table . The logical disk data other than the management area data is shifted forward by the shift amount on the disk device, and the management area data is shifted immediately after the final data of the logical disk on the disk device. A second step of moving ;
The management area in which the host computer recognizes the logical address in the I / O instruction issued from the host computer after the second step by the logical / physical disk control unit is completed by the shift calculation unit By comparing the logical address in the I / O instruction with the start address of the logical address space assigned to the logical disk recognized from the shift amount in the shift instruction. If the logical address of the management area is determined, the logical address in the I / O instruction is moved immediately after the final data of the logical disk on the disk device. Change to an address obtained by adding the logical address in the I / O instruction to the address obtained by subtracting the size of the management area from the final logical address. Logical address control method characterized by comprising a third step of. A program for causing a computer having a plurality of disk devices and a logical / physical correspondence table to function as a disk array device,
In the logical / physical correspondence table, a correspondence relationship between a logical address in a logical address space allocated to a logical disk recognized by the disk array device and a physical address on the disk device is registered, and
The computer,
When a shift instruction including the size of the management area provided at the head of the host logical disk recognized by the host computer as a shift amount is input from the management device, it is registered in the logical / physical correspondence table. A shift execution unit that adds a value corresponding to the shift amount included in the shift instruction to a logical address ;
After the shift execution unit updates the contents of the logical / physical table according to the shift instruction, the logical disk data stored in the disk device is stored based on the contents of the updated logical / physical table . A logical / physical disk that shifts data other than the management area data forward on the disk device by the shift amount and moves the data in the management area immediately after the last data of the logical disk on the disk device. control,
The logical address in the I / O instruction issued from the host computer after the shift processing and migration processing by the logical / physical disk control unit is completed is the logical address of the management area recognized by the host computer. Whether the logical address in the I / O instruction is compared with the start address of the logical address space allocated to the logical disk recognized from the shift amount in the shift instruction, and the management If it is a logical address of the area, the logical address in the I / O instruction is moved from the final logical address of the data in the management area moved immediately after the final data of the logical disk on the disk device. Shift to convert to an address obtained by adding the logical address in the I / O instruction to the address subtracted by the size of the management area Program for functioning as a calculation unit.

Images