JP2011150433A - Storage device and storage area management method in the storage device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To manage a storage area which is not used any more by a host device without writing any prescribed data pattern in the storage area by the host device itself. <P>SOLUTION: In a storage device 1, a data pattern generating and responding part 13 responds a prescribed data pattern to the host device 2 when a data read from an area in a logical extent to which any physical extent has not been assigned is requested in response to the command received by a host interface IF1. When the command received by the host interface IF1 is an area release command for notifying a range in the logical disk of the storage area which is not used any more by the host device 2, a data pattern write-in part 14 writes the prescribed data pattern into an area in an array 20 corresponding to the range notified by the area release command. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a storage apparatus including an array, and more particularly to a storage apparatus that manages storage areas that are no longer used by a host apparatus and a storage area management method in the storage apparatus.

  In a storage system provided with a storage device, the storage capacity (required capacity) required for the operation of the system may increase more than the required capacity assumed before the operation of the system is started. For this reason, when constructing a storage system, it is assumed that the storage capacity actually required will increase after the start of operation of the system, and the storage capacity is larger than the storage capacity required at the time of starting the operation. In general, a storage device is prepared in advance.

  However, as described above, when an excess capacity storage device is applied on the premise of an increase in required capacity in the future, the storage area that is not used increases and the management cost increases. Therefore, in order to reduce the management cost, the following extent method has been conventionally applied. In this extent method, a logical volume accessed from a host device that uses a storage device is managed in units of a certain size called a logical extent. Similarly, the physical area of the storage device is managed in a unit of a certain size called a physical extent. According to the extent method, when a write access occurs to a logical extent in a logical volume, a physical extent is dynamically allocated to the logical extent.

  In the storage apparatus, when a file is deleted on the file system of the host apparatus, for example, a storage area in a logical volume that is not used by the host apparatus, a so-called unused area is generated. When the entire logical extent is the unused area, if the physical extent allocated to the logical extent is released from the state allocated to the logical extent, the physical area of the storage device to which the extent method is applied It can be used efficiently. However, a non-use area generated by deleting a file on the file system cannot be recognized from the storage apparatus side. For this reason, even if the entire logical extent is an unused area, that is, even if the entire physical extent allocated to the logical extent is an unused area, the physical extent cannot be released. The utilization efficiency of the physical area deteriorates.

  Thus, for example, Patent Document 1 discloses a mechanism for enabling a storage apparatus to identify a storage area from which data is deleted when the host apparatus deletes data such as a file. In other words, the host device described in Patent Literature 1 writes a predetermined data pattern to the storage area from which data is deleted, thereby indicating that the area is an unused area that is no longer used by the host apparatus. It has a function for identifying.

  On the other hand, the storage device described in Patent Document 1 has a function of responding to a predetermined data pattern in response to a read request from a host device to an area to which a physical extent is not allocated. The storage device described in Patent Document 1 not only receives such a read request from the host device, but also issues a read request to an area where a predetermined data pattern is written by the host device when data is deleted. Even when received from the device, it has a function of responding to the host device with a predetermined data pattern. The storage device further checks the data stored in the physical extent, and if it detects that the entire physical extent matches the predetermined data pattern, releases the physical extent from the state allocated to the logical extent. It has a function to do.

JP 2008-217689 A

  As described above, in the prior art described in Patent Document 1, when the host device deletes data such as a file, the storage area from which data is deleted is an unused area that is no longer used by the host device. In order to make the storage apparatus identify, it is necessary that the host apparatus itself has a function of explicitly writing a predetermined data pattern to the storage area.

  The present invention has been made in consideration of the above circumstances, and its purpose is to manage the storage area without writing a predetermined data pattern to the storage area that is no longer used by the host apparatus. And a storage area management method in the storage apparatus.

  According to one aspect of the present invention, when a physical area of an array is divided into a plurality of physical extents, and a physical extent is not allocated to a logical extent in a logical disk specified by a write request from the host device. A storage apparatus is provided that allocates a physical extent that is not allocated to another logical extent among the plurality of physical extents to the designated logical extent. In this storage device, when a data read from an area in a logical extent to which a physical extent is not allocated is requested by a host interface that receives a command from the host device and a command received by the host interface, Data pattern generation response means for responding a predetermined data pattern to the host device, and area release for notifying the range in the logical disk of the storage area where the command received by the host interface is no longer used by the host device In the case of a command, it comprises data pattern writing means for writing the predetermined data pattern in an area in the array corresponding to the notified range.

  According to the present invention, when the host interface of the storage device receives an area release command for notifying the range in the logical disk of the storage area that is no longer used by the host device, the notification is given. Data is deleted when the host device deletes data such as files by applying a configuration in which a predetermined data pattern is written by the storage device data pattern writing means to the storage area in the array corresponding to the range. This eliminates the need for the host device itself to have a function of writing a predetermined data pattern in the storage area.

1 is a block diagram showing a configuration of a storage system including a storage device according to an embodiment of the present invention. The figure which shows the relationship between a logical disk and the physical area of an array. The figure which shows the data structure example of an extent management table. The figure which shows the example of a data structure of an unused last address management table. 6 is a flowchart for explaining processing of the storage apparatus from when an UNMAP command is received from the host apparatus to when a response indicating completion of reception of the command is returned to the host apparatus. The flowchart for demonstrating a data pattern matching process. 6 is a flowchart for explaining processing of an unused final address management unit when a write request is received from a host device. The figure for demonstrating the transition of the unused last address on a logical extent, and the write of a predetermined | prescribed data pattern in the 1st condition at the time of UNMAP command reception. The figure for demonstrating the transition of the unused last address on a logical extent, and the write of a predetermined | prescribed data pattern in the 2nd condition at the time of UNMAP command reception. The figure for demonstrating the transition of the unused last address on a logical extent, and the write of a predetermined | prescribed data pattern in the 3rd condition at the time of UNMAP command reception. The figure for demonstrating the transition of the unused last address on a logical extent, and the write of a predetermined | prescribed data pattern in the 4th condition at the time of UNMAP command reception. The figure for demonstrating the transition of the unused last address on a logical extent, the write of a predetermined | prescribed data pattern, and the release of the physical extent allocated to the logical extent under the 5th condition at the time of UNMAP command reception. . The figure which shows the state of the extent management table after releasing a physical extent.

Embodiments of the present invention will be described below with reference to the drawings.
FIG. 1 is a block diagram showing the configuration of a storage system including a storage apparatus according to an embodiment of the present invention. In the present embodiment, the storage system includes a storage device 1 and a host device 2 that uses the storage device 1. Note that the storage system may include a plurality of storage devices including the storage device 1. In addition, the storage system may include a plurality of host devices including the host device 2.

  The host device 2 performs data communication including data transfer with the storage device 1 via a host interface IF1 described later. That is, the host device 2 performs read access and write access to the storage device 1.

  The storage device 1 includes a storage controller 10 and an array 20. The array 20 includes a plurality of storage media, for example, n HDDs 21-1 (# 1) to 21-n (#n). Here, the array 20 combines the HDDs 21-1 to 21-n and defines a set of a part or all of the storage areas of the HDDs 21-1 to 21-n as one physical area with continuous physical addresses. Shall be composed. It is also possible to configure a plurality of arrays by combining the HDDs 21-1 to 21-n.

  FIG. 2 shows the relationship between the logical disk (LDISK) 210 and the physical area of the array 20. The physical area of the array 20 is managed by being divided into a plurality of physical extents of a certain size including the physical extents 201 (A) and 202 (B). The logical disk 210 is managed by being divided into a plurality of logical extents of a certain size including the logical extents 211 (A) and 212 (B). In general, the logical extent and the physical extent are the same size. In the present embodiment, the size of the logical extent and the physical extent is 600 sectors (blocks). A sector (block) is a minimum access unit of the HDDs 21-1 to 21-n.

  A physical extent in the array 20 is allocated to a logical extent in the logical disk 210 as necessary. For example, when a physical extent is not allocated to a logical extent specified by a write request from the host device 2, a physical extent that is not allocated to another logical extent is allocated to the specified logical extent. Here, whether a physical extent has not been allocated to the designated logical extent can be determined by referring to an extent management table 120 described later.

  In the example of FIG. 2, the physical extents 201 (A) and 202 (B) in the array 20 are respectively shown in the logical extents 211 (A) and 212 (B) in the logical disk 210 as indicated by arrows 221 and 222. Is assigned. In FIG. 2, the numerical value assigned to the physical extent start position in the array 20 indicates a physical block address (PBA), and the numerical value assigned to the logical extent start position in the logical disk 210 is a logical block address ( LBA). The logical disk 210 may be composed of one logical extent.

  For example, a logical disk (LDISK) number 0 is assigned to the logical disk 210 as identification information for identifying the logical disk 210. In general, the host device 2 recognizes the logical disk 210 but not the array 20.

  Referring to FIG. 1 again, the storage controller 10 includes a host interface IF1, a disk interface IF2, a physical extent securing unit 11, an extent management table storage unit 12, a data pattern generation response unit 13, a data pattern writing unit 14, and a data pattern matching unit. 15, an unused final address storage unit 16, an unused final address management unit 17, and a physical extent release unit 18. These interfaces IF1 and IF2 and the units 11 to 18 are connected to each other by a bus 19.

  The host interface IF1 receives a command from the host device 2. When the command from the host device 2 is an access request, the host interface IF1 transfers read data accompanying read access and write data accompanying write access between the host device 2 and the storage device 1. The disk interface IF2 transfers read data and write data between the storage controller 10 and the HDD 21-i in the array 20 in accordance with a request from the storage controller 10, for example.

  The physical extent securing unit 11 manages the correspondence between the logical extent in the logical disk 210 accessed from the host device 2 and the physical extent in the array 20 allocated to the logical extent. In particular, in response to a write request from the host device 2, the physical extent securing unit 11 secures a physical extent that is allocated to a logical extent in the logical disk 210 specified by the write request, and the logical extent and the physical extent. Are managed in the extent management table storage unit 12.

The extent management table storage unit 12 stores an extent management table 120. Each entry of the extent management table 120 holds extent management information indicating the correspondence between a logical extent and a physical extent allocated to the logical extent. FIG. 3 shows an example of the data structure of the extent management table 120. The extent management information held in each entry of the extent management table 120 includes logical extent information that is management information of a logical extent and physical extent information that is management information of a physical extent allocated to the logical extent.

  The logical extent information includes the logical disk (LDISK) number of the logical disk to which the corresponding logical extent belongs, the start logical block address (LBA) in the logical disk (LDISK) of the logical extent, and the size (for example, sector) of the logical extent. Number) of each piece of information. The physical extent information includes information on the start physical block address (PBA) in the array of the corresponding physical extent and the size (for example, the number of sectors) of the physical extent.

  In the example of FIG. 3, as is clear from FIG. 2, the logical extent 211 (A) has the LDISK number of the logical disk 210 to which the logical extent 211 (A) belongs, and the logical extent 211 (A) has a logical extent. This is managed by logical extent information indicating that the start LBA in the disk 210 is 0 and the sector size of the logical extent 211 (A) is 600. The physical extent 201 (A) allocated to the logical extent 211 (A) has a start PBA of 3000 in the array 20 of the physical extent 201 (A) and a sector size of the physical extent 201 (A) of 600. It is managed by physical extent information indicating this.

Similarly, in the logical extent 212 (B), the LDISK number of the logical disk 210 to which the logical extent 212 (B) belongs is 0, the start LBA in the logical disk 210 of the logical extent 212 (B) is 600, and the logical extent The extent 212 (B) is managed by logical extent information indicating that the sector size is 600. The physical extent 202 (B) allocated to the logical extent 212 (B) has a start PBA of 6600 in the array 20 of the physical extent 202 (B) and a sector size of the physical extent 202 (B) of 600. It is managed by physical extent information indicating this. ,
When a physical extent is newly allocated to a logical extent, the physical extent securing unit 11 adds an entry (row) for defining a correspondence relationship between the logical extent and the physical extent to the extent management table 120, and corresponds to the entry. Set the extent management information to be used.

  The data pattern generation response unit 13 has a function of generating a predetermined data pattern in which all bits are 0, for example. The data pattern generation response unit 13 also sends a read request to the storage device 1 from the host device 2 to the logical device to which the physical extent is not allocated, and the read request is received by the host interface IF1 in the storage device 1. In this case, a function of responding to the host apparatus 2 via the host interface IF1 with the predetermined data pattern as read data is provided. Whether the physical extent is an unallocated logical extent is determined by referring to the extent management table 120 stored in the extent management table storage unit 12.

  The data pattern writing unit 14 notifies a specific SCSI (Small Computer System Interface) command (hereinafter referred to as an UNMAP command) as an area release command that notifies the host apparatus 2 of the unused area range for releasing the unused area. Is operated by the host interface IF1. The UNMAP command belongs to the SCSI communication command set and is currently being standardized.

  The UNMAP command includes area information for designating a range of an area (unused area) in the logical disk that is not used by the host device 2. A range designated by the area information is referred to as a designated range. The area information includes, for example, information on the logical disk number of the logical disk to which the specified range belongs, the logical block address (LBA) at the start of the specified range, and the size of the specified range. The size of the specified range is represented by the number of blocks (sectors). The logical block address at the start (that is, the start position) of the specified range is called a start address, and the logical block address at the end (that is, the end position) of the specified range is called an end address. The data pattern writing unit 14 writes a predetermined data pattern in an area in the array 20 corresponding to the designated range indicated by the UNMAP command.

  The data pattern matching unit 15 uses data stored in a block area and a predetermined data pattern in units of a predetermined size area (hereinafter referred to as a block area) in the physical extent secured by the physical extent securing unit 11. A matching process is performed to compare with. In the present embodiment, the size of the block area is assumed to be 100 blocks (sectors), for example. If the data stored in the block area matches the predetermined data pattern as a result of the matching process, the data pattern matching unit 15 performs the matching process on the block area at the next address of the matched block area. It is assumed that the matching process by the data pattern matching unit 15 is performed independently regardless of the reception timing of the UNMAP command.

  The unused final address storage unit 16 stores an unused final address management table 160. The unused final address management table 160 is stored in the logical disk corresponding to the position of the end of the block area continuous from the start end of the physical extent among the block areas determined by the data pattern matching unit 15 to match the predetermined data pattern. Are stored as unused final addresses. The unused final address management table 160 holds one unused final address for each logical disk recognized by the host device 2.

  FIG. 4 shows an example of the data structure of the unused final address management table 160. As shown in FIG. 4, the unused final address management table 160 holds a pair of an LBA as an unused final address and the LDISK number of the logical disk to which the unused final address belongs. A pair of LBA and LDISK number can be defined as an unused final address. In the following description, for the sake of simplicity, the logical disk is only the logical disk 210, and the unused final address management table 160 does not necessarily have to hold the LDISK number of the logical disk to which the unused final address belongs. Shall. The unused final address is a pointer that manages an area in which physical addresses that are not used in the entire array 20 of the storage apparatus 1 are continuous.

  If the UNMAP command is received by the host interface IF1 and the specified range indicated by the UNMAP command includes an area whose end is the position of the unused final address, the unused final address management unit 17 The matching logical block address of the designated range, that is, the end address of the designated range, is registered in the unused final address management table 160 as an unused final address without performing the matching process. In addition, the unused final address management unit 17 is configured so that the host interface IF1 is located before the position of the unused final address or the position of the unused final address from the host device 2 (that is, the position whose address is smaller than the unused final address). When a write request for requesting a write to an area ending with (that is, an area not behind the position of the unused final address) is received, the unused final address held in the unused final address management table 160 is The address is updated to the address immediately before the start address specified by the write request.

  When the unused final address held in the unused final address management table 160 is the end of the physical extent or an address after the end, the physical extent releasing unit 18 uses the entire physical extent as an unused area. It is determined that there is, and the physical extent is released.

Next, with reference to the flowchart of FIG. 5, the storage apparatus 1 having the configuration shown in FIG. 1 receives the UNMAP command from the host apparatus 2 and responds to the host apparatus 2 with the completion of reception of the command. I will explain.
Now, it is assumed that an UNMAP command is transmitted from the host apparatus 2 to the storage apparatus 1 and the UNMAP command is received by the host interface IF1 in the storage apparatus 1 (step S100). In this case, the unused final address management unit 17 uses the unused final address in which the position of the start address in the specified range indicated by the UNMAP command is held in the unused final address management table 160 in the unused final address storage unit 16. It is determined whether it is behind the position of the next address (step S101).

  If the determination in step S101 is Yes, the data pattern writing unit 14 operates to write a predetermined data pattern in the area in the array 20 corresponding to the specified range indicated by the UNMAP command (step S102). In the following description, for simplification, an area in the array 20 corresponding to the specified range may be simply expressed as an area of the specified range.

  When the data pattern writing unit 14 writes a predetermined data pattern in the area of the specified range indicated by the UNMAP command (step S102), the host interface IF1 responds to the host device 2 with the completion of reception of the UNMAP command. (Step S103). As a result, the processing at the time of receiving the UNMAP command in the storage apparatus 1 ends.

  On the other hand, if the determination in step S101 is No, the unused final address management unit 17 determines whether the end address position of the specified range indicated by the UNMAP command is behind the unused final address position. (Step S104). If the determination in step S104 is No, the host interface IF1 responds to the host apparatus 2 with the completion of reception of the UNMAP command (step S103). As a result, the processing at the time of receiving the UNMAP command in the storage apparatus 1 ends.

  As described above, in this embodiment, when the position of the end address of the specified range indicated by the UNMAP command is not behind the position of the unused final address (steps S101 and S104 are No), that is, the area of the specified range. When the entirety is included in a checked area, which will be described later, regarding writing of a predetermined data pattern, the writing of the predetermined data pattern to the area of the specified range is omitted. Thereby, the load of the storage apparatus 1 can be reduced.

  On the other hand, if the determination in step S104 is Yes, the data pattern writing unit 14 operates to operate a predetermined data pattern in an area from the address next to the unused final address to the end address of the designated range indicated by the UNMAP command. Is written (step S105). Here, if the position of the start address of the specified range indicated by the UNMAP command is not behind the position of the unused final address, the predetermined data pattern of the area from the start address of the specified range to the unused final address The light is omitted.

  As described above, in the present embodiment, when the third area that is not behind the position of the unused final address is included in the designated range indicated by the UNMAP command, the array 20 in the array 20 corresponding to the third area is included. Writing of the predetermined data pattern to the fourth area is omitted. Thereby, the load of the storage apparatus 1 can be reduced.

  If the determination in step S104 is Yes, the unused final address management unit 17 indicates the unused final address held in the unused final address management table 160 in the unused final address storage unit 16 by the UNMAP command. Update to the end address of the specified range (step S106). As a result of the processing in step S106, the area in the array 20 corresponding to the position of the unused final address before the update and the position of the unused final address after the update is handled as checked with respect to writing of a predetermined data pattern. The data pattern matching process by the data pattern matching unit 15 for the region is omitted.

  Next, the unused final address management unit 17 determines the position of the end address of the designated range indicated by the UNMAP command from the position of the end address of the logical extent corresponding to the unused final address before being updated in the latest step S106. It is determined whether it is before (step S107). In step S107, the unused final address management unit 17 refers to the unused final address management table 160 in the unused final address storage unit 16 and the extent management table 120 in the extent management table storage unit 12. Done.

  In the following description, the end address of the logical extent (or physical extent) is referred to as the final address of the logical extent (or physical extent). In addition, the start address of the logical extent (or physical extent) is referred to as the start address of the logical extent (or physical extent). Similarly, the start address and end address of an area such as a check area, which will be described later, are referred to as the start address and end address of the area, respectively.

  When the determination in step S107 is No, the physical extent release unit 18 operates and the physical corresponding to the updated unused final address from the physical extent corresponding to the unused final address before being updated in the latest step S106. The area just before the extent is released (step S108). This step S108 is realized by the physical extent releasing unit 18 deleting the entry defining the logical extent corresponding to the physical extent to be released from the extent management table 120 in the extent management table storage unit 12.

  As described above, in this embodiment, the position of the end address in the specified range indicated by the UNMAP command is not earlier than the position of the end address of the logical extent corresponding to the unused final address before being updated in the latest step S106. (No in step S107) Therefore, when the area (fifth area) from the position of the unused final address to the end of the logical extent is included in the designated range, the array corresponding to the fifth area The matching process for detecting a match between the data stored in the area 20 (sixth area) and a predetermined data pattern is omitted, and the physical extent to which the sixth area belongs is released (that is, the first area The allocation of the physical extent to which the sixth area belongs to the logical extent to which the fifth area belongs is released). The physical extent release process can be simplified by releasing the physical extent associated with the omission of the matching process.

  When step S108 is executed, the host interface IF1 responds to the host apparatus 2 with the completion of reception of the UNMAP command (step S103). As a result, the processing at the time of receiving the UNMAP command in the storage apparatus 1 ends.

  On the other hand, if the determination in step S107 is Yes, the host interface IF1 responds to the host device 2 that the UNMAP command has been received (step S103). As a result, the processing at the time of receiving the UNMAP command in the storage apparatus 1 ends.

  Next, data pattern matching processing centered on the data pattern matching unit 15 in the storage apparatus 1 will be described with reference to the flowchart of FIG. In the present embodiment, it is assumed that the data pattern matching process is repeatedly executed at a constant cycle.

  First, the data pattern matching unit 15 includes a block area indicated by the address next to the unused final address held in the unused final address management table 160 in the unused final address storage unit 16 (more specifically, the unused final address). The second area in the array 20 corresponding to the first area, which is a block area in the logical disk indicated by the address next to the address, is determined as a check area used for matching with a predetermined data pattern (step S201). Next, the data pattern matching unit 15 performs a comparison process for matching the data stored in the determined check area with a predetermined data pattern, and confirms that the matching is obtained, that is, the data is stored in the check area. It is determined whether the existing data matches a predetermined data pattern (step S202). If there is a mismatch (No in step S202), the data pattern matching unit 15 ends the data pattern matching process.

  On the other hand, if they match (Yes in step S202), the data pattern matching unit 15 confirms that a predetermined data pattern is written in the check area. Then, the unused final address management unit 17 operates to set the checked area in which the match is detected as the checked area, and the unused final address held in the unused final address management table 160 as the final of the checked area. Update to address (step S203). Next, the unused final address management unit 17 determines whether the updated unused final address matches the final address of the logical extent corresponding to the unused final address (step S204). The determination in step S204 is performed by the unused final address management unit 17 referring to the unused final address management table 160 and the extent management table 120.

  If the determination in step S204 is Yes, the physical extent release unit 18 operates and is assigned to the logical extent corresponding to the unused final address by referring to the unused final address management table 160 and the extent management table 120. The specified physical extent is specified, and the specified physical extent is released (step S205). This step S205 is realized by the physical extent releasing unit 18 deleting the entry for the logical extent corresponding to the physical extent to be released from the extent management table 120. When the physical extent is released by the physical extent release unit 18 (step S205), the data pattern matching unit 15 switches the check area by updating the address of the check area to the next address (step S206). Return to S202.

  On the other hand, if the determination in step S204 is No, the check area is immediately switched by the data pattern matching unit 15 (step S206). Then, the data pattern matching unit 15 returns to step S202.

Next, the processing of the unused final address management unit 17 when the storage device 1 receives a write request from the host device 2 will be described with reference to the flowchart of FIG.
Now, it is assumed that a write request is transmitted from the host apparatus 2 to the storage apparatus 1 and the write request is received by the host interface IF1 in the storage apparatus 1. This write request designates an area in the logical disk in which data is to be written (hereinafter referred to as a write area). In order to specify the write area, the write request includes the logical disk number of the logical disk to which the write area belongs, the logical block address at the start of the write area (that is, the start address of the write area), and the size information of the write area. Consists of

  When a write request is received by the host interface IF1, the unused final address management unit 17 indicates that the position of the logical block address at the end of the write area specified by the write request, that is, the position of the end address of the write area is not used. It is determined whether it is behind the position of the unused final address held in the final address management table 160 (step S301). If the determination in step S301 is Yes, the unused final address management unit 17 ends the process when the write request is received.

  On the other hand, if the determination in step S301 is No, the unused final address management unit 17 uses the unused final address held in the unused final address management table 160 as the start of the write area designated by the write request. Update to the address immediately before the address (step S302). When the unused final address management unit 17 executes Step S302, the unused final address management unit 17 ends the process when the write request is received.

  Next, when the UNMAP command sent from the host apparatus 2 to the storage apparatus 1 is received by the host interface IF1, the unused final address p on the logical extent under the first to fifth situations, respectively. A specific example of the transition of data, the writing of a predetermined data pattern, and the release of a physical extent allocated to a logical extent will be described with reference to FIGS.

  First, a specific example under the first situation will be described with reference to FIG. As described below, the feature of the first situation is that the specified range of the UNMAP command is included in the area before the position of the address next to the unused final address p.

  In FIG. 8, the leading LBA of the logical extent 211 (A) is 0, and the final LBA is 599 (600-1 = 599). In the example of FIG. 8, the logical extent 211 (A) is divided into six block areas. Each block area is composed of 100 blocks (sectors). For example, the first LBA of the first block area of the logical extent 211 (A) is 0, and the final LBA is 99 (100-1 = 99). For example, the first LBA of the fifth block area of the logical extent 211 (A) is 400, and the final LBA is 499 (500-1 = 499). For example, the first LBA of the first block area of the logical extent 212 (B) next to the logical extent 211 (A) is 600, and the final LBA is 699 (700-1 = 699). The logical extents 211 (A) and 212 (B) described above are the same in the examples of FIGS.

  In the first situation, as shown in FIG. 8A, a predetermined data pattern is written in five block areas continuous from the head of the logical extent 211 (A). The final LBA (= 499) of the fifth block area of the logical extent 211 (A) matches the unused final address p. That is, in the first situation, the unused final address p points to the fifth block area of the logical extent 211 (A). The state of the unused final address management table 160 at this time is shown in FIG.

  In the first situation, as shown in FIG. 8A, the specified range of the UNMAP command is the size of the start LBA = 200 to 200 sectors, that is, the third and fourth block areas of the logical extent 211 (A). It is. In other words, the first situation is that the position of the start LBA of the specified range of the UNMAP command is not behind the position of the address next to the unused final address p and step S101 is No), and the position of the end LBA of the specified range is This is a case where it is not behind the position of the unused final address p (No in step S104).

  In such a first situation, as shown in FIG. 8A, a predetermined data pattern has already been written in a continuous block area having the block area at the position of the unused final address p as the final block area. Has been. Therefore, in the first situation in which steps S101 and S104 are both No, as shown in FIG. 8C, the specified range of the UNMAP command (more specifically, the area in the array 20 corresponding to the specified range). It is possible to omit writing of a predetermined data pattern and updating of the unused final address p. That is, under the first situation, the processing in the flowchart of FIG. 5 is executed in the order of start → S100 → S101 → S104 → S103 → end.

  FIG. 8D shows the state of the unused final address management table 160 at the end of the process in the flowchart of FIG. 5 executed under the first situation. As is clear from the comparison between FIGS. 8C and 8D and FIGS. 8A and 8B, the unused final address p is not updated by the processing in the flowchart of FIG. 5 in the first situation. .

  Next, a specific example under the second situation will be described with reference to FIG. As described below, the feature of the second situation is that an area in which a predetermined data pattern is written (an area in which a data pattern has been checked) and a predetermined data pattern are not written within the specified range of the UNMAP command. And the area is included.

  In the second situation, as shown in FIG. 9A, a predetermined data pattern is written in three block areas that are continuous from the head of the logical extent 211 (A). The final LBA (= 299) of the third block area of the logical extent 211 (A) matches the unused final address p. That is, in the second situation, the unused final address p points to the third block area of the logical extent 211 (A). The state of the unused final address management table 160 at this time is shown in FIG.

  In the second situation, as shown in FIG. 9A, the specified range of the UNMAP command is the size of the start LBA = 100 to 400 sectors, that is, the second to fifth block areas of the logical extent 211 (A). It is. That is, the second situation is that the position of the start LBA of the specified range of the UNMAP command is not behind the position of the address next to the unused final address p, and the position of the end LBA of the specified range is the unused final address p. (Step S101 is No, Step S104 is Yes).

  Moreover, in the second situation, a predetermined data pattern has already been written in a partial area within the range specified by the UNMAP command. Here, as is apparent from FIG. 9A, a predetermined data pattern is written in the second and third block areas of the logical extent 211 (A) in the range specified by the UNMAP command.

  Therefore, in the second situation, as shown in FIG. 9C, from the specified range of the UNMAP command, the address next to the unused final address p (LBA = 300) to the end LBA (= 499) of the specified range. A predetermined data pattern is written only to the area up to, that is, the fourth and fifth block areas of the logical extent 211 (A) (step S105). Also, the unused final address p is updated from LBA = 299 shown in FIGS. 9A and 9B to LBA = 499 as shown in FIGS. 9C and 9D (step S106). . That is, under the second situation, the processing in the flowchart of FIG. 5 is executed in the order of start → S100 → S101 → S104 → S105 → S106 → S107 → S103 → end.

  Next, a specific example under the third situation will be described with reference to FIG. The feature of the third situation is that, as described below, the area specified by the UNMAP command is an area after the position of the address next to the unused final address p, that is, an area that is not continuous with the unused final address p. It is in the point shown.

  In the third situation, as shown in FIG. 10A, a predetermined data pattern is written in three block areas that are continuous from the head of the logical extent 211 (A). The final LBA (= 299) of the third block area of the logical extent 211 (A) matches the unused final address p. These states are the same as in the second situation. The state of the unused final address management table 160 at this time is shown in FIG.

  The third situation is different from the second situation in that the specified range of the UNMAP command is 5 from the start LBA = 400 to the size of 200 sectors, that is, the logical extent 211 (A), as shown in FIG. The sixth and sixth block areas are located outside the area where the predetermined data pattern is written. In this third situation, the position of the start LBA of the specified range is behind the position of the address next to the unused final address p (Yes in step S101).

  In such a third situation, as shown in FIG. 10C, a predetermined data pattern is set in the fifth and sixth block areas of the logical extent 211 (A), which is the area specified by the UNMAP command. Is written (step S102). That is, under the third situation, the processing in the flowchart of FIG. 5 is executed in the order of start → S100 → S101 → S102 → S103 → end.

  FIG. 10D shows the state of the unused final address management table 160 at the end of the processing in the flowchart of FIG. 5 executed under the third situation. As is clear from the comparison between FIGS. 10C and 10D and FIGS. 10A and 10B, the unused final address p is not updated by the processing in the flowchart of FIG. 5 under the third situation. .

  Next, a specific example under the fourth situation will be described with reference to FIG. As described below, the feature of the fourth situation is that the UNMAP command is sandwiched between the position of the unused final address p and the area where a predetermined data pattern that is not continuous with respect to the unused final address p is written. This is in that the sandwiched area is designated as a non-use area so as to fill the area.

  In the fourth situation, as shown in FIG. 11A, a predetermined data pattern is written in three block areas that are continuous from the head of the logical extent 211 (A). The final LBA (= 299) of the third block area of the logical extent 211 (A) matches the unused final address p. A predetermined data pattern is also written in the fifth and sixth block areas of the logical extent 211 (A) that are not continuous with the unused final address p. The state of the unused final address management table 160 at this time is shown in FIG.

  In the fourth situation, the specified range of the UNMAP command is the size of 100 sectors from the start LBA = 300, as shown in FIG. 11A, and is the fourth block area of the logical extent 211 (A). . A predetermined data pattern is not written in the fourth block area. In other words, in the fourth situation, the UNMAP command designates the fourth block area as an unused area so as to fill the fourth block area to which a predetermined data pattern is not written. For this reason, in the fourth situation, the position of the start LBA of the specified range of the UNMAP command is not behind the position of the address next to the unused final address p, and the position of the end LBA of the specified range is the unused final address. It is behind the position of p (Step S101 is No, Step S104 is Yes).

  In such a fourth situation, as shown in FIG. 11C, an area from the address next to the unused final address p (LBA = 300) to the end LBA (= 399) of the specified range of the UNMAP command, That is, a predetermined data pattern is written to the fourth block area of the logical extent 211 (A) (step S105). Here, the address (LBA = 300) next to the unused final address p matches the start LBA of the specified range of the UNMAP command. Further, the unused final address p is updated from LBA = 299 shown in FIGS. 10A and 10B to LBA = 399 (step S106). That is, under the fourth situation, the processing in the flowchart of FIG. 5 is executed in the sequence of start → S100 → S101 → S104 → S105 → S106 → S107 → S103 → end.

  In the present embodiment, the process according to the flowchart of FIG. 6 is performed independently of the process according to the flowchart of FIG. In the process according to the flowchart of FIG. 6, the data pattern matching unit 15 performs a matching process in units of block areas using the block area as a check area for the area following the block area indicated by the unused final address p (step S201). , S202). The unused final address management unit 17 sets the unused final address p to the final address of the block area while the data pattern matching unit 15 detects that the data stored in the block area matches a predetermined data pattern. (Step S203). In the example of FIG. 11, after the unused final address p is updated from LBA = 299 to LBA = 399 as shown in FIG. 11C, first in the data pattern matching process according to the flowchart of FIG. 6, LBA = 399. To LBA = 499, and then LBA = 499 to LBA = 599. FIG. 11D shows the state of the unused final address management table 160 at this time.

  When the unused final address p is updated to LBA = 599 as shown in FIGS. 11C and 11D, the unused final address p is changed to the non-used final address p in the data pattern matching process according to the flowchart of FIG. It is determined that the final address of the logical extent corresponding to the used final address p matches the final address (Yes in step S204). That is, it is determined that the logical extent 211 (A) indicated by the unused final address p is not used by the host device 2. Then, the physical extent release unit 18 releases the physical extent corresponding to the logical extent 211 (A) (step S205).

  In this way, when the unused final address p is updated from LBA = 299 shown in FIGS. 11A and 11B to LBA = 599 shown in FIGS. 11C and 11D, the flowchart shown in FIG. The process is executed in the order of start → S201 → S202 → S203 → S204 → S205 → end, and the physical extent 201 (A) corresponding to the logical extent 211 (A) is released. For this release, the entry defining the correspondence relationship between the logical extent 211 (A) and the physical extent 201 (A) is deleted from the extent management table 120. The state of the extent management table 120 at this time is shown in FIG.

  Next, a specific example under the fifth situation will be described with reference to FIG. A feature of the fifth situation is that the specified range of the UNMAP command includes an area from the position of the address next to the unused final address p to the end of the logical extent.

  In the fifth situation, as shown in FIG. 12A, a predetermined data pattern is written in three block areas that are continuous from the head of the logical extent 211 (A). The final LBA (= 299) of the third block area of the logical extent 211 (A) matches the unused final address p. These states are the same as in the second situation. The state of the unused final address management table 160 at this time is shown in FIG.

  The fifth situation is different from the second situation in that the specified range of the UNMAP command is the size of the start LBA = 300 to 400 sectors, as shown in FIG. 12A, and the logical extent 211 (A) The fourth to sixth block areas and the first block area of the logical extent 212 (B) are. That is, the specified range of the UNMAP command is an area starting from the position of the address (LBA = 300) next to the unused final address p, and crosses the boundary between the logical extent 211 (A) and the logical extent 212 (B). Indicates the area.

  In such a fifth situation, as shown in FIG. 12C, in the area from the address next to the unused final address p to the end address of the designated range indicated by the UNMAP command (here, the designated range area). A predetermined data pattern is written to the fourth to sixth block areas of the logical extent 211 (A) and the first block area of the logical extent 212 (B) (step S105). The unused final address p is updated from LBA = 299 to LBA = 699, which is the final address of the first block area of the logical extent 212 (B), as shown in FIGS. 12 (c) and 12 (d). (Step S106).

  When the unused final address p is updated from LBA = 299 to LBA = 699 (step S106). The position of the end address (LBA = 699) in the designated range indicated by the UNMAP command is the end address (LBA = 599) of the logical extent 211 (A) corresponding to the unused final address p (LBA = 299) before the update. It is determined whether the position is before the position (step S107). Here, the determination in step S107 is No.

  In this case, the physical extent release unit 18 extends from the physical extent corresponding to the unused final address p (LBA = 299) before the update to immediately before the physical extent corresponding to the updated unused final address p (LBA = 699). That is, the physical extent corresponding to the logical extent 211 (A) is released (step S108). For this release, the entry defining the correspondence relationship between the logical extent 211 (A) and the physical extent 201 (A) is deleted from the extent management table 120. FIG. 13 also shows the state of the extent management table 120 at this time.

  As described above, according to the present embodiment, in the storage apparatus 1 to which the volume management method for managing (allocating and releasing) the disk capacity in units of extents is notified by the UNMAP command from the host apparatus 2 that it is not used. A predetermined data pattern for identifying that the area is not used is written to the area (designated range) by the data pattern writing unit 14 in the storage apparatus 1. As a result, even if the host device itself does not have a function of writing a data pattern for identifying an unused area that is no longer used by the host device, the unused area can be managed. .

  According to the present embodiment, the logical address corresponding to the end of the checked area where it is confirmed that a predetermined data pattern is written is stored in the unused final address storage unit 16 as the unused final address. The data pattern matching unit 15 determines whether the data stored in the second area in the array 20 corresponding to the first area in the logical disk 210 starting from the address next to the unused final address matches the predetermined data pattern. Is detected, it is confirmed that the predetermined data pattern is written in the second area. Further, according to the present embodiment, not only the unused final address is updated in response to the coincidence detection by the data pattern matching unit 15 but also the predetermined data pattern is written by the data pattern writing unit 14. The unused final address is updated with the area where the data pattern is written as the checked area. Thereby, there is no need to separately hold and manage the management information for each unused area in the range notified by the UNMAP command, even if the management size in the storage apparatus 1 is notified by the UNMAP command. Even if it is larger than that, the storage capacity required for management (management of unused final addresses) in the storage apparatus 1 can be minimized.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment.

  DESCRIPTION OF SYMBOLS 1 ... Storage apparatus, 2 ... Host apparatus, 10 ... Storage controller, 11 ... Physical extent securing part, 12 ... Extent management table storage part, 13 ... Data pattern generation response part, 14 ... Data pattern writing part, 15 ... Data pattern matching , 16 ... Unused final address storage unit, 17 ... Unused final address management unit, 18 ... Physical extent release unit, IF1 ... Host interface, IF2 ... Disk interface, 20 ... Array, 21-1 to 21-n ... HDD 120 ... Extent management table, 160 ... Unused final address management table.

Claims (6)

When the physical area of the array is divided into a plurality of physical extents and no physical extent is allocated to the logical extent in the logical disk specified by the write request from the host device, In a storage device that allocates a physical extent that is not allocated to a logical extent to the specified logical extent,
A host interface for receiving commands from the host device;
A data pattern generation response means for responding a predetermined data pattern to the host device when a data read from an area in a logical extent to which a physical extent is not allocated is requested by a command received by the host interface;
If the command received by the host interface is an area release command for notifying the range in the logical disk of the storage area that is no longer used by the host device, the area in the array corresponding to the notified range And a data pattern writing means for writing the predetermined data pattern. An unused final address storage means for storing a logical address corresponding to the end of the checked area where it is confirmed that the predetermined data pattern is written, as an unused final address;
By detecting a match between the data stored in the second area in the array corresponding to the first area in the logical disk starting from the address next to the unused final address and the predetermined data pattern Data pattern matching means for confirming that the predetermined data pattern is written in the second area;
The unused final address is updated according to the coincidence detection by the data pattern matching unit, and the area where the predetermined data pattern is written is checked according to the writing of the predetermined data pattern by the data pattern writing unit. The storage apparatus according to claim 1, further comprising: an unused final address management unit that updates the unused final address as a used area.   When the unused end address updated in response to the coincidence detection by the data pattern matching unit matches the final address of the logical extent, the physical extent releasing unit is further provided for releasing the allocation of the physical extent to the logical extent. The storage device according to claim 2.   The data pattern writing means, when a third area that is at least a part of the storage area of the notified range is not behind the position of the unused final address, the data pattern writing means, corresponding to the third area 4. The storage apparatus according to claim 2, wherein writing of the predetermined data pattern to a fourth area in the array is omitted. If the storage area in the notified range includes a fifth area from the position of the address next to the unused final address to the end of the logical extent, the data pattern matching means Detecting the coincidence between the data stored in the sixth area in the array corresponding to the area and the predetermined data pattern;
The physical extent release means, when the storage area in the notified range includes the fifth area, deallocates the physical extent to which the sixth area belongs to the logical extent. The storage device according to claim 3. When the physical area of the array is divided into a plurality of physical extents and no physical extent is allocated to the logical extent in the logical disk specified by the write request from the host device, A storage area management method in a storage apparatus for allocating a physical extent that is not allocated to a logical extent to the specified logical extent,
A host interface included in the storage device receiving a command from the host device;
When data read from an area in a logical extent to which a physical extent has not been allocated is requested by a command received by the host interface, a data pattern generation response unit included in the storage device sends predetermined data to the host device. Responding with a pattern;
If the command received by the host interface is an area release command for notifying the range in the logical disk of the storage area that is no longer used by the host device, the area in the array corresponding to the notified range In addition, the data pattern writing means of the storage device includes the step of writing the predetermined data pattern.

Images