JP2013073409A - Disk array device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a disk array device for efficiently using a storage capacity by releasing a physical extent that is releasable when a release request is received from a host device.SOLUTION: The disk array device constitutes a logical disk by combining physical extent whose storage areas are divided into many areas and is provided with: a storage part for storing a prescribed number of ranges of write request and ranges of a release request with respect to logical extent from an external device; and a management part for determining whether or not the physical extent corresponding to the logical extent that is an object of the release request is releasable on the basis of the range of the write request and range of the release request stored in the storage part when the write requests and release requests from the external device exceed the prescribed number, and the release request is received from the external device.

Description

  Embodiments described herein relate generally to a disk array device.

Conventionally, in a disk array device comprising a plurality of storage devices such as hard disk drives (HDD) and constituting an array, a logical volume accessed by the host device,
Correspondence with an array made up of one or more storage media that make up the logical volume is fixed, and depending on the system, the percentage of the area that is not actually used as a logical volume is increased, which is wasteful. large.

For such problems, thin provisioning (Thin) reduces the physical capacity by managing the physical area of the array by dividing it into a certain size called an extent, and assigning extents virtually during write access to the logical volume. Provisionin
g).

In thin provisioning, physical extents are allocated upon writing from the host device. The physical extent that is no longer required is released by a release command from the host device. The release command is, for example, a SCSI command WRITE SAME 16 or an UNMAP command.

An ideal method for releasing the physical extent is a method in which the physical extent allocated to the storage area of the logical volume indicated by the UNMAP command is released with an accuracy of one sector. However, such a method must implement either a physical extent and a logical extent that are associated with each other in units of one sector or whether all written ranges are managed in units of one sector. However, there is a problem that a large storage area for management information is required.

To deal with such a problem, as shown in Patent Document 1, 0 data is written to the storage area specified by the UNMAP command, and the physical extent is moved to ALL0 by periodic processing.
There is a technique for releasing the physical area when the physical extent of ALL0 is found. However, with this method, write data exists only in one physical extent, and when an UNMAP request is received, even if the UNMAP range is the range of the write data, it cannot be immediately released and reused. . Thus, the above-described method may be inefficient.

JP 2008-217689 A

A problem to be solved by an embodiment of the present invention is to make it possible to efficiently use the storage capacity of a storage device by enabling a physical extent to be released when a release request is received from a host device. Is to provide a device.

The disk array device of the embodiment is a disk array device that constitutes a logical disk by combining physical extents obtained by dividing a storage area into a plurality of storage areas. The range of write requests and the range of release requests from an external device to a logical extent The range of write requests stored in the storage unit and the release request when the release request is received from the external device when a predetermined number of storage units and the write request and release request from the external device exceed the predetermined number And a management unit that determines whether or not a physical extent corresponding to a logical extent that is a target of a release request can be released based on the range.

1 is a diagram showing an overall configuration of a disk array device according to a first embodiment. 1 is a diagram showing a functional configuration of a disk array device according to a first embodiment. 1 is a diagram showing a concept of a logical disk according to a first embodiment. FIG. 5 is a diagram showing an example of a logical extent management data structure according to the first embodiment. The figure which shows an example of the physical extent management data structure which concerns on 1st Embodiment. FIG. 5 is a diagram showing an example of a write / erase range management data structure according to the first embodiment. 6 is a flowchart showing an example of the operation of the disk array device according to the first embodiment. 6 is a flowchart showing an example of the operation of the disk array device according to the first embodiment. FIG. 3 is a diagram showing a concept of zero write processing of the disk array device according to the first embodiment. The figure which shows an example of the physical extent management data structure which concerns on 2nd Embodiment. FIG. 10 is a diagram showing an example of a write / erase range list according to the second embodiment. The figure which shows an example of the free list which concerns on 2nd Embodiment 2. 9 is a flowchart showing an example of the operation of the disk array device according to the second embodiment. 9 is a flowchart showing an example of the operation of the disk array device according to the second embodiment.

Embodiments of the present invention will be described with reference to the drawings. In addition, the same code | symbol is provided to the same structure.

(First embodiment)
FIG. 1 is a block diagram showing the overall configuration of the disk array device 100 of the present embodiment.
As shown in FIG. 1, the disk array device 100 includes a recording device 20 and a control unit 10 that controls the recording device 20. The control unit 10 is provided between the host device connected via the data communication path 30 and the recording device 20.

The recording device 20 includes a plurality of HDDs (Hard Disk Drives) 21 and 22 and a RAID (Redundant Arrays of Inexpensive Disc).
ks) 1 and RAID2. The HDDs 21 and 22 are SSD (So
lid State Drive). In the present embodiment, n HDDs included in RAID 1 are denoted as HDDs 21 (# 1) to (#n). Similarly, n included in RAID2
The HDDs are denoted as HDDs 22 (# 1) to (#n).

Here, the number of HDDs included in RAID 1 and RAID 2 is the same, but they may be different. Further, the number of RAIDs included in the recording device 20 may not be two. Further, the recording device 20 does not constitute a RAID, and the HDD 21 (# 1) to the HDD 21 (#n) and the HDD 22 are not included.
The storage areas (# 1) to (#n) may be configured as one continuous RAID.

The control unit 10 includes a host interface unit (HOST I / F) 11, a CPU 12, a program memory 13, a cache memory 14, and an HDD interface unit (HDD I / F).
F) 15 is included.

The HOST I / F unit 11 controls communication with an externally connected host device. That is, the HOST I / F unit 11 controls communication with the data communication path 30 side. HDD
The I / F unit 15 controls communication on the recording device 20 side.

The CPU 12 implements functions in the control unit 10 by executing various programs stored in the program memory 13. The program memory 13 has an area for storing data used by the CPU 12 in addition to the area for storing the above-described program.

The cache memory 14 is written or read in a write / read request issued from the host device or other data storage device connected via the data communication path 30 to the recording device 20 in the disk array device 100. Store data temporarily.

In the disk array device 100, a write process in response to an input request from an external device such as a host device is performed as follows. First, input data (hereinafter referred to as write data) designated by an input request (hereinafter referred to as write request) is transferred to the cache memory 14.
When the entry is entered, the completion of the writing process is notified to the requesting host apparatus.
Thereafter, actual data writing from the cache memory 14 to the recording device 20 is executed.

In addition, a read process in response to a read request from the host device is performed as follows. First, it is determined whether or not the corresponding data is entered in the cache memory 14, and if it is entered, the data is returned from the cache memory 14 to the host device via the data communication path 30. On the other hand, if the corresponding data is not entered in the cache memory 14, the data is read from the recording device 20 to the cache memory 14. Then, the read data is transmitted via the data communication path 30 to the requesting host device.

FIG. 2 shows a functional block diagram of the disk array device 100 of the present embodiment. The function shown in FIG.
2 is realized by reading and executing.

As shown in FIG. 2, the control unit 10 includes a HOST I / F control unit 111, a cache memory management unit 112, a configuration management unit 113, a logical / physical address management unit 114, and a RAID control unit 1.
15 and an HDD I / F control unit 116, and is connected to a logical disk 200 realized by the recording device 20. That is, the logical disk 200 includes HDDs 21 (# 1) to (#
n) and HDDs 22 (# 1) to (#n).

In the present embodiment, the number of logical disks 200 is one, but a plurality of logical disks 200 may be configured.

The HOST I / F control unit 111 is in charge of interface processing between the disk array device 100 and the host device. The cache memory management unit 112 manages access to the cache memory from the host device.

The configuration management unit 113 manages mapping between logical disks and RAID configurations. For example, which combination of HDDs configures a RAID, which RAID on which logical disk
It manages whether the ID belongs. The configuration management unit 113 manages mapping information for converting physical addresses and logical addresses. The mapping information is, for example, HDD 21
(1) is stored in advance as a mapping information table (not shown) and updated by the configuration management unit 113.

The logical / physical address management unit 114 is a logical / physical extent management data storage unit 120.
Connected with. Based on the mapping information managed by the configuration management unit 113, the logical / physical address management unit 114 determines the addresses in the logical disk 200 (hereinafter referred to as logical addresses) and the addresses in RAID 1 and 2 (hereinafter referred to as physical addresses). Perform conversion.

A RAID mapping unit, that is, a physical mapping unit is referred to as a physical extent. A mapping unit of a logical disk, that is, a logical mapping unit is called a logical extent.

In other words, the logical / physical address management unit 114 includes the RAID 1 and the HD configured by the logical disk 200 accessed from the host device and the HDDs 21 (# 1) to (#n).
Logical / physical extent management data for managing the correspondence relationship with RAID 2 configured by D22 (# 1) to (#n) in units of extents is created and stored in the logical / physical extent management data storage unit 120.

The HDD I / F control unit 116 controls access from the host device to the logical disk 200.

Here, the concept of the logical disk 200 will be described with reference to FIG. As shown in FIG. 3, the logical disk 200 according to the present embodiment has RAID 1 or HDD 22 (# 1) that defines the storage areas of HDDs 21 (# 1) to (#n) as one continuous storage area.
A storage area of RAID 2 that defines the storage areas of (#n) as one continuous storage area is configured by logical extents 211 that are divided into the same size as the physical extent 201 that is divided into a fixed size.

The physical extent 201 is secured by the logical / physical extent management unit 114 when the logical disk 200 receives a write request from the host device. The logical extent 211 included in the write request and the reserved physical extent 201 are associated with each other and stored in the logical / physical extent management storage unit 120.

As shown in FIG. 3, in this embodiment, RAID1 has m physical extents 201 (0
) To physical extent 201 (m−1). RAID2 is k physical extents 2
01 (0) to physical extent 201 (k-1). The logical disk 200 includes h logical extents 211 (0) to (h-1).

The logical / physical extent management unit 114 uses RAID 1 physical extents 201 (0) to
The correspondence relationship between the physical extents 201 (0) to (k-1) of (m-1) and RAID2 and the logical extents 211 (0) to (h-1) is managed.

In other words, the logical / physical extent management unit 114 manages the correspondence (logical extent management data 130 and physical extent management data 140).
Is created and stored in the logical / physical extent management data storage unit 120.

The structure of the logical extent management data 130 stored in the logical / physical extent management data storage unit 120 and the structure of the physical extent management data 140 will be described with reference to FIGS.

As shown in FIG. 4, the logical extent management data 130 includes a logical disk number 131, a logical extent number 132, and a physical extent pointer 133. The logical extent number 132 stores the logical extent number. The physical extent pointer 133 stores the address of physical extent management data 140 described later.

The logical extent 211 (0) included in the logical disk 200 shown in FIG.
This corresponds to the physical extent 201 (1) included in D1. In this embodiment, since there is one logical disk, “1” is stored as the logical disk number 131. Logical extent number 1
32 stores “0”, and the physical extent pointer 133 stores “1, 1”. A RAID number is stored on the left of the physical extent pointer 133, and a physical extent number is stored on the right. Alternatively, the physical extent pointer 133 may store an address indicating the RAID1 physical extent 201 (1).

Further, the logical extent 211 (h-3) included in the logical disk 200 shown in FIG.
Corresponds to the physical extent 201 (m−1) of RAID1. In this case, the logical disk number 131 is “1”, the logical extent number 132 is “h-3”, and the physical extent pointer 133 is “1, (m−1)”.

Also, the logical extent 211 (h-1) included in the logical disk 200 shown in FIG.
Corresponds to the physical extent 201 (k-3) of RAID2. In this case, the logical disk number 131 is “2”, the logical extent number 132 is “h−1”, and the physical extent pointer 133 is “2, (k−3)”.

That is, when the physical extent 201 is allocated to the logical extent 211, a memory address indicating physical extent management data 140 (to be described later) is held in the physical extent pointer 133. Physical extent 201 is logical extent 21
If it is not assigned to 1, the physical extent pointer 133 holds “NULL”.

The logical / physical address management unit 114 acquires physical extent management data 140 corresponding to a specific logical extent 211 based on the physical extent pointer 133 of the logical extent management data 130.

When no physical extent is allocated to the logical disk, all physical list pointers 133 included in the logical extent management data 130 are “NULL”. When receiving a write request from the host device, the configuration management unit 113 allocates the physical extent 201 to the logical extent 211.

Next, the physical extent management data 140 will be described with reference to FIG. As shown in FIG. 5, the physical extent management data 140 includes a RAID number 141, a physical extent number 142, a logical extent pointer 143, and a physical extent management flag 14.
4 and write / erase range management data 145.

The RAID number 141 indicates the number of each RAID constituting the recording device 20. The physical extent number 142 indicates the number of the physical extent included in each RAID. The logical extent pointer 144 indicates a logical extent allocated to the physical extent indicated by the physical extent number 142.

The physical extent 201 uniquely specified based on the RAID number 141 and the physical extent number 142 corresponds to the logical extent 211 stored in the logical extent pointer 143. When the physical extent 201 is not allocated to the logical extent 211, the logical extent pointer 143 is “NULL”.

That is, the logical / physical address management unit 114 acquires the logical extent 211 corresponding to the specific physical extent 201 based on the logical extent pointer 143 of the physical extent management data 140.

The physical extent management flag 144 is write / erase range management data 145 described later.
Is a flag indicating whether or not the number of data to be stored in the write / erase range array 146 included in the number exceeds the number of write / erase range arrays 146 included in the write / erase range management data 145.

The write / erase range management data 145 is composed of write / erase range arrays 146 [0] to [s-1] which are s pieces of array data. s is an arbitrary natural number and is set in advance.

The write / erase range management data 145 includes a write request range or a release request range from the host device, and the write / erase range array 146 [0] to [s-
1].

In this embodiment, when the physical extent management flag 144 is “1”, all the write / erase range arrays 146 included in the write / erase range management data 145 are in use (
Data is stored). When the physical extent management flag 144 is “0”,
It is assumed that a write / erase range array 146 that is not in use exists in the write / erase range management data 145.

That is, the logical / physical address management unit 114 writes the write / erase range management data 145.
Is used to manage the range of write requests or release requests from the host device. Specifically, when the logical / physical address management unit 114 receives a write request or release request from the host device, the write / erase range array 146 [0] to [s−1] included in the write / erase range management data 145. ], The received write request or release request range is stored in an unused array.

The data structure of the write / erase range array 146 [0] to [s-1] constituting the write / erase range management data 145 will be described with reference to FIG.

As shown in FIG. 6, the write / erase range array 146 constituting the write / erase range management data 145 stores a write / erase flag 147, an offset 148, and a size 149.

The write / erase flag 147 indicates whether the request from the host device stored in the write / erase range array 146 including the write / erase flag 147 is a write request or an open request. For example, in the case of a write request, “0” is stored in the write / erase range flag 146. In the case of the release request, “1” is stored in the write / erase range flag 146.

The offset 148 is a write / erase range array 146 including the offset 148.
Range of the write request stored in the write / erase range array 146 including the offset 148 in the physical extent 201 from the start address of the physical extent 201 specified by the physical extent number 142 of the physical extent management data 140 having Alternatively, it indicates the distance to the start address of the release request range. That is, the offset 148 indicates the address of the start position in the physical extent 201 of data to be written or erased from the host device.

The size 149 indicates the size of data written or erased from the host device.
For example, when the disk array device 100 receives a write request from the host device, the offset 148 stores the address of the write start position in the physical extent 201 indicated by the physical extent number 142, and the size 149 is written. Stores the size of the data.

When the disk array device 100 receives a release request from the host device, the offset 148 stores the position where the erasing process starts, and the size 149 stores the size of the data to be deleted.

With reference to FIG. 7, the processing operation of the disk array device 100 of the present embodiment when a write request is received from the host device will be described.

When the HOST I / F control unit 111 of the disk array device 100 receives a write request from the host device, the logical / physical address management unit 114 receives a logical extent corresponding to the range of the write request from the logical / physical extent management data storage unit 120. The logical extent management data 130 of 211 is extracted (step S10). The search is performed based on, for example, a hash or a tree.

Subsequently, the logical / physical address management unit 114 checks whether or not the physical extent pointer 133 included in the physical extent management data 140 corresponding to the logical extent management data 130 extracted in step S10 is NULL (step S11). Note that the physical extent pointer 133 being NULL means that the physical extent 201 has not been allocated to the logical extent 211 that is the target of the write request.

When the physical list pointer 133 is not NULL (Yes in step S11), that is, the physical extent 2 is already added to the logical extent 211 to be written from the host device.
If 01 is already assigned, the logical / physical address management unit 114 checks the physical extent management flag 144 (step S12).

If the physical extent management flag 144 is “1” (No in Step S12), the control unit 10 performs a write process on the recording device 20 based on the range of the write request from the host device (Step S15) and ends. .

When the physical extent management flag 144 is “0” (step S12 is Yes),
The logical / physical address management unit 114 determines that the logical / physical address management unit 114 uses an unused write / erase range array 146 for the management data 140 of the allocated physical extent 201.
It is confirmed whether there is any (step S13).

When the write / erase range array 146 is all in use (No in step S13), the logical / physical address management unit 114 is recorded in the write / erase range array 146 included in the write / erase range management data 145. Data initialization is performed, and the physical extent management flag 144 of the logical extent management data 140 is set to 1 (step S14). Thereafter, the control unit 10 performs a write process on the recording apparatus 20 based on the write request received from the host apparatus (step S15), and ends the process.

When there is an unused write / erase range array 146 (Yes in step S13), the logical / physical address management unit 114 stores the write request range from the host device in the unused write / erase range array 146. (Step S16). If there are a plurality of unused write / erase range arrays 146, they are recorded in the youngest write / erase range array 146, for example.

In other words, the logical / physical address management unit 114 sets the address of the start position where writing is performed in the physical extent 201 assigned to the logical extent 211 of the write request destination from the host device to the offset 148 of the write / erase range array 146. Store. The size 149 stores the data size of the write request from the host device.

Thereafter, the logical / physical address management unit 114 sets the write / erase flag 147 of the write / erase range array 146 to “0” indicating that the data stored in the write / erase array 146 is data for the write process. Set (step S17), the process ends.

When the physical extent pointer 133 is NULL (No in step S11), the configuration management unit 113 secures one physical extent from unused physical extents managed by a free list (not shown) or the like (step S18). The configuration management unit 113 sets the address of the physical extent secured in step S13 to the physical extent pointer 133 confirmed in step S12 (step S19).

At that time, the logical / physical address management unit 114 sets the physical extent management flag 144 of the logical extent management data 140 to “0” (step S20). Subsequently, the control unit 1
0 performs a write process on the recording device 20 based on the write request received from the host device (step S21). The writing process in step S21 is the same as that in step S15.

Thereafter, the logical / physical address management unit 114 proceeds to the processing of step S16 and step S17, and ends the processing.

Next, with reference to FIG. 8, the operation of the disk array device 100 of this embodiment when a data release request is received from the host device will be described. Note that the data release request of the present embodiment is made by, for example, an UNMAP command.

Upon receiving the UNMAP command from the host device, the logical / physical address management unit 114 receives the UN
The logical extent management data 130 in the MAP range is acquired (step S30).

The logical / physical address management unit 114 checks whether or not the physical extent management flag 144 included in the physical extent management data 140 corresponding to the acquired logical extent management data 130 is 0 (step S31).

If the physical extent management flag 144 is “1” (No in step S31), that is, if all the write / erase range arrays 146 included in the write / erase range management data 145 are in use, logical / physical address management The unit 114 performs zero write processing on the UNMAP range in the physical extent 211 assigned to the logical extent 211 of the UNMAP request destination (step S39). Zero write processing refers to the physical extent 211 allocated to the logical extent 211 of the UNMAP request destination received from the host device.
In this process, “0” is written in all the UNMAP ranges in FIG.

When the physical extent management flag 144 is “0” (Yes in step S31), the logical / physical address management unit 114 is unused in the physical extent management data 140 corresponding to the logical extent management data 130 acquired in step S30. It is confirmed whether or not there is a write / erase range array 146 (step S32).

When there is an unused write / erase range array 146 (Yes in step S32), logical /
The physical address management unit 114 stores the physical extent management data 14 corresponding to the logical extent management data 130 acquired in step S30 in the write / erase range array 146.
The UNMAP range in the physical extent 201 indicated by the physical extent number 133 of 0 is stored (step S33). At this time, the logical / physical address management unit 114 sets the write / erase flag 147 of the corresponding write / erase range array 146 to “1” (step S34).

After recording the UNMAP range in step S33, or when there is no unused write / erase range array 146 (No in step S32), the logical / physical address management unit 114 uses the physical extent 201 by the UNMAP command received in step S30. It is confirmed whether or not can be opened (step S35).

FIG. 9 shows a conceptual diagram of a method for confirming whether the physical extent 201 can be opened. As shown in FIG. 9, the logical / physical address management unit 114 creates a bitmap having the number of bits corresponding to the number of sectors of the physical extent, for example, and initializes all of the bitmaps to 1.

The logical / physical address management unit 114 writes / writes the write / erase range management data 145.
The erase range array 146 is confirmed in ascending order, and if the write / erase flag 147 is 0, that is, if the process stored in the target write / erase range is a write process, the corresponding bit is changed to 0.

If the write / erase flag 147 is 1, the corresponding bit is set to 1. Confirmation is performed up to the last array, and if the resulting bitmap becomes ALL1, it is determined that the physical extent can be released.

For example, it is assumed that the write / erase range array 146 [0] to [2] as shown in FIG. 9 exists in the write / erase range management data 145. Write / erase range array 146 as described above
When the bitmap is changed based on the write / erase flag 147, the offset 148, and the size 149 in order from [0], “0” exists in the bitmap after checking the write / erase range array 146 [2]. doing. For this reason, in the case of the arrangement shown in FIG. 9, the physical extent is not releasable.

Returning to the description of FIG. As a result of confirmation, when it is determined that the physical extent 201 can be released (
In step S36, the logical / physical address management unit 114 determines that the physical extent 20
1 is released (step S37).

The physical extent 201 is released when the physical / logical address management unit 114 sets the physical list pointer in the logical extent management data 130 to NULL and the physical extent management data 140 of the released physical extent 201 is implemented as a free list. This is done by returning to the free state by the unallocated physical extent management method. At this time, the physical / logical address management unit 114 transmits a physical extent release request to the configuration management unit 113. Based on the received release request, the configuration management unit 113 executes the physical extent 201.
Is released. That is, the configuration management unit 113 stores the configuration in which the physical extent 201 is released in the mapping information table based on the release request. That is, it is mapped that the physical extent 201 is unused.

If the physical extent cannot be released (No in step S36), the logical / physical address management unit 114 sets “1” in the physical extent management flag 144 (step S38). Thereafter, the logical / physical address management unit 114 performs zero write processing (step S39). Zero write processing is performed from the array [0] of the write / erase range management data 145 [
s-1] is checked in ascending order, and the erasure range other than the portion overwritten by subsequent writing in time series is zero-written. This zero write process creates bitmaps for the number of sectors of the physical extent in the same manner as the extent release check in step S35.
The initial value is set to a bit pattern of all 0s, and the range in which 1 remains is written to zero.

As described above, when the disk array device 100 of this embodiment receives a write request or a release request from the host device, the management of the physical extent 201 allocated to the logical extent 211 of the write request or release request destination. If there is an unused write / erase range array 146 in the data 140, the received write request range or erase request range is stored in the write / erase range array 146.

Further, when receiving a release request from the host device, the disk array device 100 according to the present embodiment determines whether or not the physical extent 201 allocated to the release request destination logical extent 211 can be released.

As a result of the determination, if the release is possible, the configuration management unit 113 releases the physical extent 201 assigned to the logical extent 211 of the release request destination.

In other words, in the present embodiment, the physical extent 201 indicates the range of the received write request or release request until the s-th write request or release request, which is a predetermined number preset for each physical extent 201, is received. Management is performed using the management data of the physical extent. As a result, the disk array device 100 according to the present embodiment performs s
Until the first write request or release request, the physical extent can be released in synchronization with the release request. For this reason, the disk array device 100 of this embodiment can improve the capacity efficiency.

(Second Embodiment)
A second embodiment of the present invention will be described with reference to the drawings. In addition, the same code | symbol is provided to the structure same as 1st Embodiment.

The disk array device 100 of this embodiment has the same functional configuration as that of the first embodiment.
The disk array device 100 of this embodiment manages the range of write requests and the range of release requests by a management method using a free list. That is, the logical / physical address management unit 114 of this embodiment secures and releases the logical extent 211 and the physical extent 201 using a free list management method. That is, the logical / physical address management unit 1
14 has an allocator function.

Therefore, the structure of the physical extent management data stored in the logical / physical extent management data storage unit 120 is different from that of the first embodiment.

  FIG. 10 shows the structure of the physical extent management data 240 in the second embodiment.

As shown in FIG. 10, the structure of the physical extent management data 240 is RAID number 1
41, a physical extent number 142, a logical extent pointer 143, a physical extent management flag 144, and a write / erase range list pointer 245.

RAID number 141, physical extent number 142, logical extent pointer 143
Since the physical extent management flag 144 is the same as that of the first embodiment, description thereof is omitted.

The write / erase range list pointer 245 is the physical extent 2 allocated to the logical extent 211 indicated by the write request or release request from the host device.
Reference numeral 01 denotes a pointer indicating a write / erase range list 246 for managing a write range or an erase range. The data structure of the write / erase range list 246 is shown in FIG.

As shown in FIG. 11, the write / erase range list 246 has a data structure including a next pointer 247, an offset 148, and a size 149. Offset 148 and size 14
Since 9 is the same as that of the first embodiment, description thereof is omitted. Next pointer 247 is
The next write / erase range list 246 is shown.

Note that the number of write / erase range lists 246 managed in this embodiment is set in advance for each logical disk 200. Here, the number of write / erase range lists is t.

FIG. 12 shows a free list 251 for managing unused ones of the write / erase range list 246.

As shown in FIG. 12, the free list 251 does not store a write request or release request range, that is, an unused write / erase range list 246 is stored with the list head 250 at the head by a pointer.

When the logical / physical address management unit 114 receives a write request or a release request from the host device, the logical / physical address management unit 114 secures a head write / erase range list 246 pointed to by the pointer of the list head 250 of the free list 251. Then, the logical / physical address management unit 114 removes the reserved write / erase range list 246 from the free list, and the next pointer 2
47, the wrist head 250 is updated. Secured write / erase range list 24
6 is linked to the tail of the list starting from the write / erase range list pointer of the physical extent management data 240.

NULL is stored in the next pointer 247 of the last write / erase range list 246 stored in the free list 251. That is, the logical / physical address management unit 11
4 determines that there is no more unused write / erase range list 246 when the next pointer 247 stored in the list head 250 of the free list 251 is NULL.

Here, with reference to FIG. 13, processing when the disk array device 100 of the second embodiment receives a write request from the host device will be described.

Of the processes shown in FIG. 13, steps that are the same as the processes shown in FIG. Therefore, step S113 and step S in FIG.
The process 116 will be described.

The logical / physical address management unit 114 determines whether there is an unused write / erase range list 246 based on the write / erase range free list 251 (step S113).
.

When there is an unused write / erase range list 246 (Yes in step S113), the logical / physical address management unit 114 removes the unused write / erase range list from the free list and adds it to the list of physical extent management data 240. Connect. The write request range received from the host device is stored in the write / erase range list 246 (step S116).
), The writing process is terminated.

Next, the operation of the disk array device 100 when receiving a release request from the host device will be described with reference to FIG. Similar to FIG. 13, the processing of step S <b> 132 and step S <b> 133 will also be described for FIG.

The logical / physical address management unit 114 determines whether there is an unused write / erase range list 246 based on the free list 251 (step S132). If there is an unused write / erase range list 246 (Yes in step S132), the logical / physical address management unit 114 receives the UNMAP range in the physical extent 201 allocated to the release request destination logical extent 211 received from the host device. Is acquired from the free list and stored in the write / erase range list 246 linked to the physical extent management data 240 (step S133).

Subsequently, the logical / physical address management unit 114 performs an extent release check (step S35). Since the process after step S35 is the same as the process after step S35 of FIG. 8, description is abbreviate | omitted.

Even when there is no unused write / erase range list 246 (No in step S132).
), The logical / physical address management unit 114 performs an extent release check (step S).
35).

The allocation of the write / erase range list 246 per physical extent 201 may be unlimited. However, when a large number of writes are requested, t number of logical extents set per logical disk 200 with a small number of physical extents. The write / erase range list 246 may be used up. Therefore, an upper limit value may be set for the number of write / erase range lists 246 that can be allocated for each physical extent.

As described above, in the present embodiment, the physical extent 211 can be released in synchronization with the release request. In the present embodiment, the write request range and the release request range from the host device are recorded and managed by the free list method, thereby managing a predetermined number of write requests and release requests for each logical disk. It becomes possible. Therefore, the write / erase range list can be managed even if the write request or release request of the host device is local. As a result, it is possible to provide a disk array device with high use efficiency of the storage area.

As mentioned above, although embodiment of this invention was described, these embodiment was shown as an example to the last and is not intending limiting the range of invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

DESCRIPTION OF SYMBOLS 10 ... Control part 11, HOST I / F part, 12 ... CPU, 13 ... Program memory, 1
4 ... Cache memory, 15 ... HDD I / F unit, 20 ... Recording device, 21, 22 ... HDD
30 ... Connection path, 100 ... Disk array device

Claims (4)

A disk array device that configures a logical disk by combining physical extents obtained by dividing a storage area into a plurality of storage areas,
A storage unit for storing a predetermined number of write request ranges and release request ranges for the logical extents from the external device;
When the write request and the release request from the external device exceed the predetermined number,
When the release request is received from the external device, the physical extent corresponding to the logical extent that is the target of the release request is released based on the range of the write request and the range of the release request stored in the storage unit. A management unit for determining whether or not it is possible;
A disk array device comprising: 2. The disk array device according to claim 1, wherein the storage unit stores a flag indicating whether or not the write request range and the release request range stored in the write / erase range array exceed the predetermined number. . A disk array device that configures a logical disk by combining physical extents obtained by dividing a storage area into a plurality of storage areas,
A free management list that stores a predetermined number of write request ranges and release request ranges for the logical extents from the external device, and a management list that does not store the write request range and release request range in the management list. A storage unit for storing the list;
When an access from the external device exceeds the number of management lists stored in the free list, and when a release request is received from the external device, based on the management data, the logical extent corresponding to the release target A management unit that determines whether a physical extent can be released; and
A disk array device comprising: The disk array device according to any one of claims 1 to 3, further comprising: an opening unit that makes the physical extent unused when the determination result of the management unit is openable.

Images