JP5472930B2 - Storage system provided with disk array device, disk array device, and extent size changing method - Google Patents

Description

  Embodiments described herein relate generally to a storage system including a disk array device in which a physical extent in an array is allocated to a logical extent of a logical disk, a disk array device, and an extent size changing method.

  In general, a disk array device includes one or more disk devices, for example, a plurality of hard disk drives (HDD) and an array controller connected to these HDDs. The array controller manages a plurality of HDDs by a commonly known technique such as RAID (Redundant Arrays of Independent Disks) or Redundant Arrays of Independent Disks. In other words, in response to a data read / write request from the host device (host computer), the array controller speeds up access by moving a plurality of connected HDDs in parallel and executing data read / write in a distributed manner. In addition, a redundant configuration is used to improve reliability.

  The disk array device includes an array (disk array) in which the storage areas of the plurality of HDDs (or one or more HDDs) are defined as one continuous area. The disk array device also has a logical disk (logical volume) that is recognized by the host device. The logical disk is configured by allocating a storage area (physical area) in the array to the area (logical address space).

  In such a disk array device, the correspondence between the logical disk accessed by the host device and the array providing the physical area constituting the logical disk is fixed. Therefore, depending on the system, the ratio of the area that is not actually used as a logical disk increases, and there is a case where waste is large.

  Therefore, for example, the entire area (physical area) of the array included in the disk array device is managed by dividing it into a certain size called an extent (physical extent), and when a write access to a logical disk occurs, A virtual logical disk method has been proposed in which a physical extent allocated to a certain area is actually secured.

  In other words, in the disk array device, a logical disk that is recognized by the host device is configured by combining one or more physical extents of all physical extents obtained by dividing the entire area of the array included in the disk array device. The

  Such a disk array device manages the correspondence between the logical disk accessed (recognized) from the host device and the physical extent allocated to the area of the logical disk, and receives a write request from the host device after the logical disk is constructed. Each time, a virtual logical volume function of securing a physical extent to be allocated to a corresponding area on the logical disk is provided.

  Also, a mechanism has been proposed in which once a physical extent is secured, the disk array device itself detects an unused area, thereby re-releasing the secured physical extent and efficiently using the storage capacity.

Japanese Patent No. 433320

  By the way, the extent can be managed and released more finely as the size of the extent is smaller. For this reason, generally, the extent size is reduced, and the utilization efficiency of the storage capacity tends to be improved. On the other hand, when the extent size is reduced, the number of extents for which allocation is managed increases, and management information necessary for allocation management increases. Moreover, it is necessary to convert from a logical extent to a physical extent, for example, every access. For this conversion, it is necessary to search for the corresponding management entry in the management information every time it is accessed. For this reason, the access performance tends to deteriorate.

  On the other hand, the data arrangement on the logical disk differs depending on the file system and application program (hereinafter abbreviated as application) used by the operating system operating on the host device. The nature of this data arrangement is often revealed only after the application is run. For this reason, it is generally difficult to set an appropriate size for the disk array device at the system construction stage where a logical disk is created.

  A problem to be solved by the present invention is to provide a storage system, a disk array device, and an extent having a disk array device, which can perform more appropriate extent management according to the situation of data arrangement on a logical disk by a host device. To provide a resizing method.

  According to the embodiment, the storage device includes a disk array device and a host device that uses the disk array device. The disk array device is an array controller that controls access to at least one array, and manages a physical area of the at least one array by dividing it into a plurality of physical extents, and among the plurality of physical extents An array controller that allocates one or more of the above to logical extents of the logical disk. The host device has a first size of a continuous first area used by the host device on the logical disk and a second size of a continuous second area not used by the host device on the logical disk. The second size is detected, and an extent size suitable for a situation in which the logical disk of the host device is used is detected based on the detected first size and second size, and the detected extent Data arrangement analyzing means for notifying the size to the array controller is provided. The array controller includes extent size changing means for changing the logical extent of the logical disk and the extent size of the physical extent allocated to the logical extent from the current extent size to the notified extent size.

1 is a block diagram showing a configuration of a storage system including a disk array device according to a first embodiment. The figure which shows the example of the use continuous area and non-use continuous area on a logical disk. The figure which shows the example of the frequency distribution of a continuous area size. The figure which shows the list of the result of ranking of the frequency of continuous area size. The figure which shows the example of the allocation condition of the physical extent with respect to the logical extent of a logical disk before extent size change. The figure which shows the example of the extent management table T1 before extent size change. The figure which shows the example of the extent management table T10 before extent size change. The figure which shows the example of the extent management table T11 before extent size change. The figure which shows the example of extent management table T12 before extent size change. The figure which shows the example of the extent copy management table before extent size change. The figure which shows the example of the allocation state of the physical extent with respect to the logical extent of a logical disk during extent size change execution. The figure which shows the example of the extent management table T1 in execution of extent size change execution. The figure which shows the example of the extent management table T10 during extent size change execution. The figure which shows the example of the extent management table T11 in execution of extent size change execution. The figure which shows the example of the extent management table T12 in execution of extent size change. The figure which shows the example of extent management table T20 during extent size change execution. The figure which shows the example of the allocation state of the physical extent with respect to the logical extent of a logical disk in an extent size change completion state. The figure which shows the example of the extent management table T1 in the extent size change completion state. The figure which shows the example of the extent management table T10 in the extent size change completion state. The figure which shows the example of the extent management table T11 in the extent size change completion state. The figure which shows the example of the extent management table T12 in the extent size change completion state. The figure which shows the example of the extent management table T20 in the extent size change completion state. The figure which shows the example of the condition which is copying the data of a part of physical extent of the extent size before a change to the physical extent of the extent size after a change in 2nd Embodiment. In the second embodiment, when copying of a part of the physical extent of the extent size before the change to the physical extent of the extent size after the change is completed, the physical extent of the extent size after the change is allocated. The figure which shows the example by which is eliminated. The figure which shows the example of the extent management table T1 of the extent size change completion state in 2nd Embodiment. The figure which shows the example of the extent management table T12 of the extent size change completion state in 2nd Embodiment.

Hereinafter, embodiments will be described with reference to the drawings.
[First Embodiment]
FIG. 1 is a block diagram showing the configuration of a storage system including a disk array device according to the first embodiment.

  The storage system shown in FIG. 1 includes a disk array device 1 and a host device 2. The disk array device 1 and the host device 2 are connected via, for example, a network. The host device 2 uses the disk array device 1 as its own external storage device, for example. The host device 2 performs data communication including data transfer with the disk array device 1 via a host interface IF1 described later. That is, the host device 2 performs read access and write access to the disk array device 1 (more specifically, a logical disk constructed in the disk array device 1).

  The host device 2 includes a data arrangement analysis unit 201. The data arrangement analysis unit 201 sets a first size (hereinafter referred to as a use continuous area size) of a continuous first area (hereinafter referred to as a use continuous area) on the logical disk used by the host device 2. To detect. The data arrangement analysis unit 201 also has a second size (hereinafter referred to as an unused continuous area size) of a continuous second area (hereinafter referred to as an unused continuous area) on a logical disk that is not used by the host device 2. Detected). Further, the data arrangement analysis unit 201 detects an extent size suitable for the current system operation status based on the detected frequency distribution of the used continuous area size and the unused continuous area size. In this case, the data arrangement analysis unit 201 calculates the greatest common divisor for the sizes up to a certain high rank with the highest frequency in the frequency distribution, and detects an appropriate extent size based on the greatest common divisor.

FIG. 2 shows an example of the use continuous area and the non-use continuous area on the logical disk. As shown in FIG. 2, n use continuous areas UCA ₀ , UCA ₁ ,..., UCA _k ,..., UCA _n−1 and m non-use continuous areas UUCA ₀ , UUCA ₁ are arranged on the logical disk. , ..., UUCA _k , ..., UUCA _m-1 . In the example of FIG. m = n + 1. In this case, the data allocation analyzing section 201, using continuous area _{UCA 0, UCA 1, ...,} UCA k, ..., UCA n-1 of size (using continuous area size) S [0], S [ 1], ..., S [k], ..., S and [n-1], nonuse continuous area _{UUCA 0, UUCA 1, ...,} UUCA k, ..., size UUCA _m-1 (unused continuous area size) G [0], G [1], ..., G [k], ..., G [m-1] are detected.

  Referring again to FIG. 1, the disk array device 1 includes an array controller 10 and an array unit 20. The array unit 20 includes one or more disk devices (storage devices), for example, p magnetic disk drives (HDDs) 21-1 (# 1) to 21-p (#p). Part or all of the HDDs 21-1 to 21-p are used to form one or more arrays, for example, a plurality of arrays. Each of the plurality of arrays is defined as one area in which a part or all of the storage areas of one or more HDDs among the HDDs 21-1 to 21-p are continuous. In the present embodiment, it is assumed that a group of arrays including arrays 210-0, 210-1, and 210-2 is configured by a set of different HDDs among the HDDs 21-1 to 21-p. In this embodiment, the arrays 210-0, 210-1 and 210-2 are identified by group numbers (GROUPno) 0, 1 and 2, respectively. Also, the entire areas of the arrays 210-0 and 210-1 are managed by being divided into physical extents having a size of 2 MB, and the entire area of the array 210-2 is managed by being divided into physical extents having a size of 32 KB. To do.

  The array controller 10 controls access to the above-described array group (more specifically, access to the HDDs 21-1 to 21 -p constituting the array group) and performs data transmission / reception with the host device 2. . The array controller 10 includes a host interface IF1, a disk interface IF2, and a main controller 100.

  The host interface IF1 is in charge of interface processing including data input / output processing with the host device 2. The disk interface IF2 is in charge of interface processing including data input / output processing with the HDDs 21-1 to 21-p. The main control unit 100 is the center of the array controller 10 and manages the HDDs 21-1 to 21-p by, for example, a RAID method. More specifically, the main control unit 100 distributes data read / write by moving some or all of the HDDs 21-1 to 21 -p in parallel in response to a read / write request from the host device 2. And run.

  The main control unit 100 includes an array configuration unit 101, a logical disk configuration unit 102, a management table storage unit 103, a physical extent reservation unit 104, a data copy unit 105, an extent size change unit 106, and a logical disk reconstruction unit 107. .

  The array configuration unit 101 is defined as one area in which a part or all of the storage areas of one or more HDDs among the HDDs 21-1 to 21-p are continuous, and is composed of one or more physical extents. Build an array of Here, as described above, a plurality of arrays including the arrays 210-0, 210-1, and 210-2 are constructed. The plurality of arrays are assumed to be arrays having a RAID configuration, for example.

  For example, the logical disk configuration unit 102 constructs one or more logical disks composed of one or more logical extents in response to an instruction from the host device 2. In the present embodiment, it is assumed that a plurality of logical disks composed of a plurality of logical extents are constructed. The plurality of logical disks include two logical disks having logical disk numbers (LDISKno) of 0 and 1, respectively. The management table storage unit 103 is used to store various management tables. A specific example of the management table stored in the management table storage unit 103 will be described later.

  The logical extent securing unit 104 secures, from one of the arrays 210-0 to 210-2, a physical extent to be allocated to a logical extent that is a target of the write access when a write access to the logical disk occurs.

  The data copy unit 105 performs data copy from the physical extent having the physical extent size before the change to the physical extent having the physical extent size after the change, which is necessary when the physical extent size (physical extent size) is changed. . The extent size changing unit 106 changes the physical extent size based on the analysis result by the data arrangement analyzing unit 201. More specifically, the extent size changing unit 106 changes the physical extent size to the size notified from the data arrangement analyzing unit 201. The logical disk reconfiguration unit 107 reconfigures the logical disk according to an instruction from the host device 2, for example.

  Next, the operation of the storage system shown in FIG. 1 will be described. In this embodiment, the extent size in the initial state in the disk array device 1 is 2 MB, and the manageable extent sizes are 2 MB, 1 MB, 512 KB, 256 KB, 128 KB, 64 KB, 32 KB, 16 KB, 8 KB, and 4 KB. It is assumed that either. That is, in the present embodiment, the extent size in the initial state in the disk array device 1 is set to the maximum manageable extent size. In this embodiment, the array unit 20 includes arrays having physical extent sizes of 1 MB, 512 KB, 256 KB, 128 KB, 64 KB, 16 KB, 8 KB, and 4 KB, respectively, in addition to the arrays 210-0 to 210-2.

  The data arrangement analysis unit 201 in the host device 2 sets the continuous area to be used for the logical disk used by the host device 2 as needed, periodically, or in accordance with an instruction from the user after the operation of the storage system is started. Each unused continuous area is detected. This detection is performed based on, for example, a file system used by an operating system operating on the host device 2.

  When detecting the use continuous area, the data arrangement analysis unit 201 detects the size of the detected use continuous area as the use continuous area size. Further, when detecting the unused continuous area, the data arrangement analysis unit 201 detects the size of the detected unused continuous area as the unused continuous area size.

In the case of the logical disk shown in FIG. 2, the data arrangement analyzing unit 201 detects _n use continuous areas UCA ₀ , UCA ₁ ,..., UCA _k ,. The sizes S [0], S [1], ..., S [k], ..., S [n-1] are detected. The data arrangement analyzing unit 201 detects m unused continuous areas UUCA ₀ , UUCA ₁ ,..., UUCA _k ,..., UUCA _m−1, thereby corresponding corresponding unused continuous area sizes G [0], G [1], ..., G [k], ..., G [m-1] are detected.

  The data arrangement analysis unit 201 detects the used continuous area sizes S [0], S [1],..., S [k],. Based on G [1],..., G [k],..., G [m−1], a frequency distribution regarding the continuous region size is calculated. That is, when detecting the use continuous area size or the non-use continuous area size, the data arrangement analysis unit 201 determines whether or not the frequency (that is, the number of appearances) of the size X that matches the detected size exists in the frequency distribution. If a frequency of size X exists, the data arrangement analysis unit 201 increments the frequency of the size X by 1. On the other hand, if the frequency of size X does not exist, the data arrangement analysis unit 201 adds the frequency of 1 for the size X to the frequency distribution.

  The data arrangement analysis unit 201 repeats the above calculation of the frequency distribution for all the detected use continuous area sizes and non-use continuous area sizes. Thereby, for example, it is assumed that the frequency distribution of the continuous region size as shown in FIG. 3 is acquired. The frequency distribution of the continuous area size reflects the operating state of the file system and application program used by the operating system operating on the host device 2.

  Based on the frequency distribution of the acquired continuous area size, the data arrangement analysis unit 201 ranks the continuous area size having the highest frequency in order of frequency, for example. Based on the frequency ranking result, the data arrangement analysis unit 201 detects the continuous area size from the first rank to a predetermined rank (that is, a high-order constant rank with a high frequency). Here, assuming that the continuous area size up to the third place is detected, in the example of FIG. 3, 1 MB (= 1024 KB), 32 KB, and 64 KB are detected in order from the first place. FIG. 4 shows a list of results of ranking the frequencies of the continuous area sizes.

  The data arrangement analysis unit 201 detects the detected greatest common divisor G. 1 MB, 32 KB, and 64 KB. C. M.M. {1 MB, 32 KB, 64 KB} = 32 KB is calculated. The data arrangement analysis unit 201 calculates the greatest common divisor G. C. M.M. Based on the above, an appropriate extent size suitable for the file system or application currently operating on the host device 2 (that is, the current system operation status) is detected. Here, the calculated greatest common divisor G. C. M.M. That is, 32 KB is detected as an appropriate extent size.

If the detected extent size is applied as a physical extent size, physical extents can be allocated to continuous areas having the frequency (number of appearances) from the first place to the third place without excess or deficiency. . The greatest common divisor G. C. M.M. Of the integral multiple of the largest common divisor G. C. M.M. A size in a range not exceeding the maximum value (1 MB) among the extent sizes used in the calculation of the above may be set as an appropriate extent size.
The data arrangement analysis unit 201 notifies the detected extent size “32 KB” to the array controller 10 of the disk array device 1.

  When the extent size is notified from the data arrangement analysis unit 201 to the array controller 10, the extent size changing unit 106 of the array controller 10 performs extent size changing processing for changing the extent size from 2 MB to 32 KB. 104 and the data copy unit 105 are executed in cooperation.

  FIG. 5 shows an example of the status of physical extent allocation to the logical extent of the logical disk whose logical disk number is 0 (LDISKno = 0) before the extent size is changed. In the example of FIG. 5, the area of the logical disk is divided into a plurality of logical extents having an extent size of 2 MB. Each logical extent is assigned a physical extent having the same extent size as the logical extent (that is, 2 MB).

For example, the physical extent PE0 “x” in the array 210-0 with the group number 0 (GROUPno = 0) in the logical extent LE0 [2n] _2MB with an even logical extent number (LEno) of the logical disk with LDISKno = 0 _2MB Is assigned. On the other hand, the logical extent number (LEno) of the logical disk with LDISKno = 0 is the odd numbered logical extent LE0 [2n + 1] _2MB , and the physical extent PE1 [y] _2MB in the array 210-1 with the group number 1 (GROUPno = 1) Is assigned.

In FIG. 5, the physical extent PE0 [0] _2MB of PEno = 0 in the array 210-0 is allocated to the logical extent LE0 [0] _2MB of LEno = 0, and the logical extent LE0 [1] _{2MB of} LEno = 1. In the figure, the physical extent PE1 [0] _{2 MB} of PEno = 0 in the array 210-1 is allocated.

  FIG. 6 shows an example of the extent management table T1 for the logical disk before the extent size is changed. The extent management table T1 is stored in the management table storage unit 103. Each entry in the extent management table T1 includes a logical disk number (LDISKno) field, a logical extent number (LEno) field, a group number (GROUPno) field, a physical extent number (PEno) field, and an extent size (Esize) field. And.

  The logical disk number field is used to hold the logical disk number (LDISKno) of the corresponding logical disk. The logical extent number field is used to hold the logical extent number (LEno) of the logical extent of the corresponding logical disk. The group number (GROUPno) field is used to hold the group number (GROUPno) of the array to which the physical extent allocated to the corresponding logical extent belongs. The physical extent number field is used to hold the physical extent number (PEno) of the physical extent allocated to the corresponding logical extent.

  FIG. 7 shows an example of the extent management table T10 for the array 210-0 with GROUPno = 0 before the extent size is changed. FIG. 8 shows an example of the extent management table T11 for the array 210-1 with GROUPno = 1 before the extent size is changed. FIG. 9 shows an example of the extent management table T12 for the array 210-2 with GROUPno = 2 before the extent size is changed. The extent management tables T10, T11, and T12 are stored in the management table storage unit 103.

  Each entry in the extent management tables T10, T11, and T12 includes a logical disk number field, a group number field, a logical extent number field, a physical extent number field, and an extent size field similar to the extent management table T1. Each entry of the extent management tables T10, T11, and T12 further includes a first copy management field (COPYDone) and a second copy management field (COPYYentryno).

  The first copy management field is used to hold a copy completion flag (COPYDone). The copy completion flag (COPYdone) indicates whether data copying to or from the physical extent indicated by the physical extent number field of the corresponding entry is complete (COPYdone = 1) or not (COPYdone = 0). Indicate. The second copy management field holds the entry number (COPYENTRYNO) of the entry in the extent copy management table T20 in which the physical extent indicated by the physical extent number field of the corresponding entry is registered as the copy source or the copy destination. Used.

  FIG. 10 shows an example of the extent copy management table T20 before the extent size is changed. The extent copy management table T20 is stored in the management table storage unit 103. Each entry in the extent copy management table T20 includes a copy entry number (COPYEYtryno) field, a copy source group number (GROUP_origin) field, a copy source physical extent number (PEno_origin) field, a copy source extent size (Esize_origin) field, and a copy. A destination extent size (Esize_dest) field, a copy destination group number (GROUPno_dest) field, and a copy destination physical extent number (PEno_dest) field are provided.

  The copy entry number field is used to hold the entry number of the corresponding entry as a copy entry number when valid entry information is registered in the corresponding entry. The copy source group number field is used to hold the group number of the array to which the copy source physical extent belongs as the copy source group number. The copy source physical extent number field is used to hold the physical extent number of the copy source physical extent as the copy source physical extent number. The copy source extent size field is used to hold the extent size of the copy source physical extent as the copy source extent size.

  The copy destination extent size field is used to hold the extent size of the copy destination physical extent as the copy destination extent size. The copy destination group number field is used to hold the group number of the array to which the copy destination physical extent belongs as the copy destination group number. The copy destination physical extent number field is used to hold the physical extent number of the copy destination physical extent as the copy destination physical extent number.

  In the present embodiment, the extent size is changed in the direction of reduction. In this case, data is copied from the copy source physical extent to a plurality of copy destination physical extents. Therefore, each entry in the extent copy management table T20 includes a plurality of copy destination group number fields and a plurality of copy destination physical extent number fields.

  The extent size change unit 106 instructs the data copy unit 105 to copy data that is necessary when the extent size is changed from 2 MB to 32 KB in the extent size change process. As a result, for example, copying of data from a physical extent group having an extent size of 2 MB already reserved for accessing a logical disk having LDISKno = 0 to a physical extent group having a newly allocated extent size of 32 KB is performed. Shall be done.

FIG. 11 shows an example of the status of physical extent allocation to the logical extent of the logical disk with LDISKno = 0 during the extent size change execution. In the example of FIG. 11, 64 MBs existing in the array 210-2 having a group number of 2 (GROUPno = 2) instead of the _{2 MB} physical extent PE0 [0] _2MB in the 2 _MB logical extent LE0 [0] _2MB . A 32 KB physical extent PE2 [0] _{32 KB} , PE2 [1] _{32 KB} , PE2 [2] _{32 KB} ,..., PE2 [63] _{32 KB} are allocated. In this embodiment, the physical extent numbers PE2 [0] _32KB , PE2 [1] _32KB , PE2 [2] _32KB ,..., PE2 [63] The _32KB physical extent numbers (PEno) are 0, 1, 2,. , 63. However, the 64 32 KB physical extents allocated to the logical extent LE0 [0] _2MB (more specifically, the logical extent LE0 [0] _2MB divided into 64) are not necessarily continuous in the array 210-2. You don't have to.

Similarly, instead of the 2 MB logical extent LE0 [1] _{2 MB} and the 2 MB physical extent PE1 [0] _{2 MB} , 64 32 KB physical extents PE2 [64] _{32 KB} and PE2 [65] existing in the array 210-2. ] _{32 KB} , PE2 [66] _{32 KB} ,..., PE2 [127] _{32 KB} are allocated. In this embodiment, physical extent PE2 [64] _32KB , PE2 [65] _32KB , PE2 [66] _32KB ,..., PE2 [127] _32KB physical extent numbers (PEno) are 64, 65, 66,. , 127. However, 64 32 KB physical extents allocated to the logical extent LE0 [1] _2MB (more specifically, the logical extent LE0 [1] _2MB divided into 64) are not necessarily continuous in the array 210-2. You don't have to.

Physical extent _{PE2 [0] 32KB ~PE2 [63} ] 32KB and _{PE2 [64] 32KB ~PE2 [127} ] physical extents group _32KB according inquiry from extent resizing unit 106 is ensured by the physical extent secured portion 104 . More specifically, a group of 32 KB physical extents existing in the GROUPno = 2 array 210-2 in which data should be copied from the 2 MB physical extents respectively allocated to all logical extents of the logical disk of LDISKno = 0. Is secured by the physical extent securing unit 104. Then, the extent size changing unit 106 updates the extent copy management table T20 as shown in FIG. The extent copy management table T20 shown in FIG. 16 will be described later.

Based on the extent copy management table T20 shown in FIG. 16, the extent size changing unit 106 changes the physical extent PE0 [0] _2MB to the physical extent PE2 [0] _{32KB to} PE2 [63] _32KB with respect to the data copy unit 105. Instructs data copy in units of 32 KB. Then, the data copy unit 105 divides the _{2 MB} physical extent PE0 [0] _2MB into 32 KB by 64 (2 MB / 32 KB), and for each divided 32 KB, the physical extent PE2 [i] _{32 KB} corresponding to the 32 KB data. i = 0, 1,..., 63).

The extent size changing unit 106 updates the extent management tables T10, T11, and T12 according to the progress of data copying by the data copying unit 105. Eventually, when the data copy from the physical extent PE0 [0] _{2 MB} to the physical extent PE2 [0] _{32 KB} to PE2 [63] _{32 KB} is completed, the extent size changing unit 106 transfers the data to the data copy unit 105 based on the extent copy management table T20. On the other hand, a data copy in units of _{32 KB} from the physical extent PE1 [0] _{2 MB} to the physical extent PE2 [64] _{32 KB} to PE2 [127] _{32 KB} is instructed. Then, the data copy unit 105 divides the _{2 MB} physical extent PE1 “0” _{2 MB} into 32 KB by 64, and for each 32 KB divided, the physical extent PE2 [j] _{32 KB} (j = 64, 65) corresponding to the 32 KB data. ,..., 127).

In FIG. 11, the physical extent PE2 [0] _{32 KB to} PE2 [63] _{32 KB} and the data copy to PE2 [64] _{32 KB} to PE2 [66] _{32 KB} are completed, that is, the physical extent PE2 [66] to _{32 KB.} Shows that data copy is complete.

  12, 13, 14, and 15 show examples of extent management tables T1, T10, T11, and T12, respectively, in the state in which the extent size change shown in FIG. 11 is being executed. FIG. 16 shows an example of the extent copy management table T20 when the extent size change shown in FIG. 11 is being executed.

  As is apparent from FIG. 12, during the extent size change execution, the contents of the extent management table T1 do not change from before the extent size change shown in FIG. On the other hand, for example, the extent copy management table T20 is updated when a 32 KB physical extent group to which data is to be copied is secured, as described above. FIG. 16 shows the state of the extent copy management table T20 at this time. The state of the extent copy management table T20 shown in FIG. 16 is maintained until the extent size change execution is completed.

  In the extent management table T20, as shown in FIG. 16, the first entry includes a copy entry number (COPYEYtryno), a copy source group number (GROUP_origin), a copy source physical extent number (PEno_origin), a copy source extent size (Esize_origin), and a copy. As the destination extent size (Esize_dest), COPYeytryno = 0, GROUPno_origin = 0, PEno_origin = 0, Esize_origin = 2MB, and Esize_dest = 32 KB are set, respectively.

  In the first entry of the extent management table T20, 64 copy destination physical extent numbers, that is, PEno_dest = 0, PEno_dest = 1,..., PEno_dest = 63 are set. In the first entry of the extent management table T20, the same copy destination group number (GROUP_dest = 2) indicating the array 210-2 is set in association with PEno_dest = 0, PEno_dest = 1,..., PEno_dest = 63. Has been.

  The first entry of the extent copy management table T20 is obtained by dividing the physical extent of the copy source indicated by GROUPno_origin = 0 and PEno_origin = 0 into 64, and for each of the divided 32 KB, the 32 KB data is GROUPno_dest = 2 and PEno_dest = Indicates that copying is to be made to a physical extent at the copy destination indicated by i (i = 0, 1,..., 63). Similarly, as is apparent from FIG. 16, the second entry of the extent copy management table T20 divides the copy source physical extent indicated by GROUPno_origin = 0 and PEno_origin = 1 into 64, and every 32 KB divided. The 32 KB data is to be copied to the copy destination physical extent indicated by GROUPno_dest = 2 and PEno_dest = j (j = 63, 64, 65,...).

  On the other hand, the extent management tables T11 to T12 are updated according to the progress of extent size change (data copy). That is, the contents of the extent management tables T10 to T12 change during the execution of the extent size change process.

  For example, in the extent management table T10, that is, the extent management table T10 for the array 210-0 of GROUPno = 0, as shown in FIG. 13, the copy entry number is set in the second copy management field (COPYENTRYNO) of each entry. Has been. The first copy management field of the entry used to manage the first entry of the extent management table T10, that is, the 2 MB physical extent PE0 [0] (see FIG. 11) specified by GROUPno = 0 and PEno = 0. In (COPYdone), a copy completion flag (COPYdone = 1) indicating copy completion is set.

  Next, also in the extent management table T11 for the array 210-1 with GROUPno = 1, as shown in FIG. 14, the copy entry number is set in the second copy management field (COPYENTRYNO) of each entry. . On the other hand, in the first copy management field (COPYdone) of each entry of the extent management table T11, a copy completion flag (COPYdone = 0) indicating that copying has not been completed is still set.

  Next, also in the extent management table T12 for the array 210-2 with GROUPno = 2, as shown in FIG. 15, the copy entry number is set in the second copy management field (COPYENTRYNO) of each entry. . On the other hand, the first to 66th entries of the extent management table T12, that is, 32 KB physical extents PE2 [0] to PE2 [66] indicated by PEno = 0 to PEno = 66 in the array 210-2 with GROUPno = 2 (FIG. 5). 11) is set in the first copy management field (COPYdone) of the first to 66th entries, and a copy completion flag (COPYdone = 1) indicating copy completion is set.

  It is assumed that the extent size changing process is completed as a result of the above operation being repeated. That is, the data copy from the 2 MB physical extent allocated to all the logical extents of the LDISKno = 0 logical disk to the 32 KB physical extent group existing in the GROUPno = 2 array 210-2 is completed. To do. Then, the extent size changing unit 106 updates the extent management table T1 based on the extent management table T12, so that a logical disk with LDISKno = 0 is composed of a 2 MB logical extent group and a 32 KB logical extent group from the logical disk. To a logical disk that consists of At this time, the extent size changing unit 106 allocates a 32 KB physical extent in the GROUPno = 2 array 210-2 to each of the 32 KB logical extent groups.

  As described above, in this embodiment, the extent size changing unit 106 reconfigures the logical disk in accordance with the change of the physical extent size. However, the logical disk reconstruction unit 107 may perform such logical disk reconstruction.

FIG. 17 shows an example of the status of physical extent allocation to the logical extent of the logical disk with LDISKno = 0 in the extent size change completion state. In the example of FIG. 17, the 32 KB logical extents LE0 [0] _{32KB to} LE [63] _32KB and LE0 [64] _32KB corresponding to the _{2 MB} logical extents LE0 [0] _2MB and LE0 [1] _2MB shown in FIG. the _{~LE [127] 32KB, GROUPno =} 2 of 32KB of physical extents PE0 in the array _{210-2 [0] 32KB ~PE [63} ] 32KB and _{PE0 [64] 32KB ~PE [127} ] 32KB is assigned .

  18, 19, 20, and 21 show examples of extent management tables T1, T10, T11, and T12, respectively, in the extent size change completion state shown in FIG. FIG. 22 shows an example of the extent copy management table T20 in the extent size change completion state shown in FIG.

  As described above, in the first embodiment, the data arrangement analysis unit 201 of the host device 2 targets the logical disk used by the host device 2 after starting the operation of the storage system, and the size and non-use of the continuous area used. The size of each continuous area is detected. Then, the data arrangement analysis unit 201 detects an extent size suitable for the current operation status of the storage system based on the detection result of the size of the used continuous area and the size of the unused continuous area. More specifically, the data arrangement analysis unit 201 calculates the greatest common divisor for the sizes up to a certain upper order with a high frequency based on the frequency distribution of the detected used continuous region size and the unused continuous region size, An appropriate extent size is detected based on the greatest common divisor. The data arrangement analysis unit 201 notifies the detected extent size to the array controller 10 of the disk array device 1.

  In response to this, the extent size changing unit 106 of the array controller 10 changes the logical extent of the logical disk and the extent size of the physical extent allocated to the logical extent to the notified extent size. As a result, the array controller 10 can select a more appropriate extent according to the operation status of the storage system, that is, the status of the data arrangement on the logical disk accompanying the operation of the file system or application used by the operating system operating on the host device 2. Extent management by size (operation of the disk array device 1) can be performed.

[Second Embodiment]
Next, a second embodiment will be described.
The first feature of the second embodiment is data copy in the extent size change processing. The data copy unit 105 checks whether the entire data to be copied to the copy destination physical extent is a predetermined data pattern, for example, all zeros. There is in point to do. The second feature of the second embodiment is that when the data to be copied to the copy destination physical extent is all zero, the extent size changing unit 106 eliminates the assignment of the copy destination physical extent to the corresponding logical extent. It is in.

  The basic configuration of the storage system in the second embodiment is the same as that of the first embodiment. Therefore, FIG. 1 is also used in the second embodiment. However, in the second embodiment, the functions of the data copy unit 105 and the extent size change unit 106 are different from the functions of the data copy unit 105 and the extent size change unit 106 in the first embodiment as described above.

As in the first embodiment, the extent size changing unit 106 performs the extent size changing process for changing the extent size from 2 MB to 32 KB in cooperation with the physical extent securing unit 104 and the data copying unit 105. Shall be executed. Further, in the extent size changing process, the extent size changing unit 106 uses the extent copy management table T20 shown in FIG. 16 to the data copy unit 105 to store the 2 MB physical in the array 210-1 with GROUPno = 0. Instructed to copy data in 32KB units from extent PE0 [0] _2MB to the newly secured 32KB physical extent PE2 [0] _{32KB to} PE2 [63] _{32KB in} the array 210-2 with GROUPno = 2 Shall.

Then, the data copy unit 105 divides the _{2 MB} physical extent PE0 [0] _2MB into 32 KB by 64 (2 MB / 32 KB), and for each divided 32 KB, the physical extent PE2 [i] _{32 KB} corresponding to the 32 KB data. i = 0, 1,..., 63). Here, among the divided 32KB of data to physical extent PE0 [0] _2MB 64, the physical extent PE0 [0] i-th from the head of the _{2MB (i = 0,1, ...,} 63) 32KB of data Is called the i-th data. When copying the i-th data of the physical extent PE0 [0] _{2 MB} to the physical extent PE2 [i] _{32 KB} , the data copy unit 105 checks whether the i-th data is all zero. This check is called zero data check.

FIG. 23 shows an example of a situation where the second data of the physical extent PE0 [0] _{2 MB} is copied to the physical extent PE2 [2] _{32 KB} . In the situation shown in FIG. 23, the copying of the 0th data and the first data of the physical extent PE0 [0] _2MB to the physical extent PE2 [0] _32KB and PE2 [1] _32KB is completed.

Further, in the state shown in FIG. 23, the data copy unit 105 checks whether all the predetermined bytes (for example, the size of one sector) are all zeros from the top of the second data of the physical extent PE0 [0] _{2 MB} by using the arrow 231. As shown, the second data is copied to the physical extent PE2 [2] _{32 KB} . When the predetermined byte is all zero, the data copy unit 105 updates the final address 232 with the zero data check stored in the management table storage unit 103 by a predetermined byte, for example.

Soon, it is assumed that the zero address checked final address 232 matches the final address of the physical extent PE2 [2] _{32 KB} . FIG. 24 shows the state at this time. The extent size changing unit 106 compares the final address 232 after the zero data check with the final address of the physical extent PE2 [2] _{32 KB} . If the zero address checked final address 232 matches the final address of the physical extent PE2 [2] _{32 KB} , the extent size changing unit 106 corresponds to the physical extent PE2 [2] _{32 KB} as indicated by an arrow 240. The allocation to the ₃₂ KB logical extent (here, LE0 [2] _{32 KB} ) is canceled. As described above, the physical extent whose allocation to the logical extent has been canceled is allocated to the logical extent when a new physical extent needs to be secured.

FIG. 25 and FIG. 26 show the extent management tables T1 and T12 in the state where the allocation of the physical extent PE2 [2] _{32KB to} the logical extent LE0 [2] _32KB is canceled and the extent size change is completed. An example is shown.

As is apparent from FIG. 25, a 32 KB physical extent is not allocated to a 32 _KB logical extent with LDISKno = 0 and LEno = 2, that is, logical extent LE0 [2] _{32 KB} . As is apparent from FIG. 26, the 32 KB physical extent of PEno = 2 existing in the GROUPno = 2 array 210-2, that is, the physical extent PE2 [2] _{32 KB} is not allocated to the 32 KB logical extent.

  According to the second embodiment, when it is detected that the i-th data is all zero, the physical extent in the array to which the i-th data is to be copied is released. Thereby, the utilization efficiency of the storage capacity of the array after the extent size changing process can be improved.

[Third Embodiment]
Next, a third embodiment will be described.
The feature of the third embodiment is that the extent size changing unit 106 detects the size of the use continuous area and the size of the non-use continuous area as the data arrangement analysis unit 201 performed in the first embodiment. In the point.

  The basic configuration of the storage system in the third embodiment is the same as that of the first embodiment. Therefore, FIG. 1 is also used in the third embodiment. However, in the third embodiment, the data arrangement analysis unit 201 is not necessarily required. On the other hand, unlike the first embodiment, the extent size changing unit 106 has a function of detecting the size of the used continuous area and the size of the unused continuous area, and detecting an appropriate extent size based on the detection result. doing.

  Next, detection of extent size by the extent size changing unit 106 will be described. The extent size changing unit 106 functions as a data arrangement analysis unit at any time, periodically, or in accordance with an instruction from the user after the operation of the storage system is started. Then, the extent size changing unit 106 detects a use continuous area and a non-use continuous area for each logical disk used by the host device 2. However, unlike the data arrangement analysis unit 201, it is generally difficult for the extent size changing unit 106 to use the file system for detecting the use continuous region and the non-use continuous region.

  Therefore, in the third embodiment, the extent size changing unit 106 detects, for example, a use continuous area and a non-use continuous area in a logical disk with LDISKno = 0 as follows. The extent size changing unit 106 refers to the entries in the extent management table T1 illustrated in FIG. 6 in order from, for example, entries in which LEno = 0 is set. Then, the extent size changing unit 106 determines whether or not the group number (GROUPno) and the physical extent number (PEno) are set in the reference entry, that is, the logical extent number (LEno) set in the reference entry. It is determined whether a physical extent is allocated. The extent size changing unit 106 detects a logical extent to which a physical extent is not allocated as at least a part of an unused continuous area.

The extent size changing unit 106 performs a zero data check on a logical extent to which a physical extent is allocated, as applied in the second embodiment for the physical extent, so that the used continuous area and the unused extent are not used. Detect continuous areas. Here, it is assumed that the second physical extent is allocated to the second logical extent subsequent to the first logical extent. The extent size changing unit 106 checks whether the data is zero data in order from the beginning of the second physical extent, thereby setting the area in the second logical extent corresponding to the area where the zero data is continuous as an unused continuous area. To detect. The extent size changing section 106 detects an area in the second logical extent corresponding to the area other than the unused continuous area as used continuous area.

  If at least the end side of the first logical extent is a use continuous area and at least the start end of the second logical extent is a use continuous area, the extent size changing unit 106 uses both use continuous areas as one use continuous area. The area is collected and the size of the combined use continuous area is detected. If at least the end side of the first logical extent is an unused continuous area and at least the start end of the second logical extent is an unused continuous area, the extent size changing unit 106 sets both unused continuous areas to 1 The two unused continuous areas are combined and the size of the combined unused continuous area is detected. In this way, the extent size changing unit 106 detects the size of all used continuous areas and the size of unused continuous areas in a logical disk with LDISKno = 0.

Here, as in the first embodiment, as in the example of FIG. 2, n number of uses continuous area _{UCA 0, UCA 1, ...,} UCA k, ..., the UCA _n-1, respectively use a continuous area size Assume that S [0], S [1],..., S [k],. Similarly, m unused continuous areas UUCA ₀ , UUCA ₁ ,..., UUCA _k ,..., UUCA _m−1 are respectively used continuous area sizes G [0], G [1],. ], G [m−1] are detected.

  In this case, the extent size changing unit 106 acquires the frequency distribution of the continuous region size as shown in FIG. 3 in the same manner as the data arrangement analyzing unit 201 in the first embodiment. Then, the extent size changing unit 106 detects 1 MB (= 1024 KB), 32 KB, and 64 KB as the continuous area size from the first place to the third place based on the obtained frequency distribution of the continuous area size. The extent size changing unit 106 detects the greatest common divisor G.1MB of 1 MB, 32 KB, and 64 KB detected. C. M.M. {1 MB, 32 KB, 64 KB} = 32 KB, and the calculated greatest common divisor G. C. M.M. Based on the above, an appropriate extent size is detected. Here, the calculated greatest common divisor G. C. M.M. = 32 KB is detected as an appropriate extent size.

  Then, the extent size changing unit 106 changes the extent size from 2 MB (current extent size) to 32 KB (the detected appropriate size), as in the case where the extent size is notified from the data arrangement analysis unit 201 in the first embodiment. The extent size changing process for changing to the extent size) is executed in cooperation with the physical extent securing unit 104 and the data copy unit 105.

  According to the third embodiment, the array controller 10 can independently detect the extent size suitable for the system operation status from the host device 2. In the third embodiment, the zero data check as applied in the second embodiment is performed in the data copy to the physical extent having the changed extent size, which is necessary when the extent size is changed. You may apply. In this case, when the extent size changing unit 106 detects the size of the used continuous area and the size of the unused continuous area, it corresponds to each zero data area in which zero data continues in the physical segment of the extent size before the change. The start position and end position in the array of the zero data area to be stored may be stored in the management table storage unit 103, for example. This makes it easy to determine whether the data to be copied from the physical segment of the extent size before the change to the physical segment of the extent size after the change is all zero based on the information of the start position and end position for each zero data area. Can be checked.

  According to at least one embodiment described above, a storage system including a disk array device and a disk array device capable of performing more appropriate extent management in accordance with the status of data arrangement on a logical disk by a host device And an extent size changing method can be provided.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the 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 1 ... Disk array apparatus, 2 ... Host apparatus, 10 ... Array controller, 20 ... Array part, 21-1-21-p ... HDD, 100 ... Main control part, 101 ... Array structure part, 102 ... Logical disk structure part, DESCRIPTION OF SYMBOLS 103 ... Management table memory | storage part, 104 ... Physical extent securing part, 105 ... Data copy part, 106 ... Extent size change part, 107 ... Logical disk reconstruction part, 201 ... Data arrangement analysis part, 210-0-210-2 ... array.

Claims (6)

An array controller for controlling access to at least one array, wherein a physical area of the at least one array is divided into a plurality of physical extents and managed, and one or more of the plurality of physical extents are logically managed A disk array device having an array controller to be allocated to the logical extent of the disk;
A host device using the disk array device,
The host device has a first size of a continuous first area used by the host device on the logical disk and a second size of a continuous second area not used by the host device on the logical disk. A second size is detected based on a file system used by an operating system operating on the host device, and the detected first size and second size are used to detect the detected first size. Obtain a frequency distribution indicating the number of occurrences of the size and the second size for each size, calculate the greatest common divisor of the size up to a certain level of higher rank, and calculate the greatest common divisor or the maximum a size of an integral multiple of common divisor, the size of the range not exceeding the maximum value of the size used in the calculation of the greatest common divisor before SL array controller A data arrangement analyzing means for knowledge,
The array controller
And extent size changing means for changing the extent size of the physical extents newly allocated to the logical extent and the logical extent of the logical disk from the current extent size to the notified size,
When the extent size after the change is smaller than the current extent size, the first physical extent allocated to the first logical extent of the logical disk is partitioned by the extent size after the change. Data copying means for copying the data corresponding to each of the areas to the second physical extent group having the changed extent size,
The extent size changing unit divides the first logical extent into a second logical extent group having the changed extent size, and the second logical extent group in which the corresponding data has been copied is copied to the second logical extent group. A storage system that allocates physical extents. The data copy means checks whether the entire corresponding data is a predetermined data pattern,
The extent size changing unit is configured to change the second logical extent of the second physical extent to which the corresponding data is copied from the second physical extent group when the entire corresponding data has a predetermined data pattern. the storage system of claim 1, wherein to eliminate the assignment to the corresponding second logical extents of the extent groups. The at least one array is a plurality of arrays including a first array having the first physical extent group and a second array having the second physical extent group;
The storage system according to claim 2 , wherein the array controller further includes physical extent securing means for securing the second physical extent group used as a copy destination from the second array when copying the data. The storage system according to any one of claims 1 to 3 , wherein the current extent size is a maximum manageable extent size set at the start of operation of the storage system. At least one array;
An array controller connected to a host device and controlling access to the at least one array, managing a physical area of the at least one array by dividing it into a plurality of physical extents, An array controller that allocates one or more of them to a logical extent of a logical disk;
The array controller
A first size of a continuous first area on a logical disk used by the host device and a second size of a continuous second area on the logical disk not used by the host device , By detecting an area in which a predetermined data pattern continues in a physical extent allocated to a logical extent of the logical disk, and based on the detected first size and second size, Obtain a frequency distribution indicating the number of occurrences of the detected first size and second size for each size, calculate the greatest common divisor of the size up to a certain upper rank with the highest frequency, and calculate the maximum common divisor, or a size of an integral multiple of the greatest common divisor, detects the size of the range not exceeding the maximum value of the size used in the calculation of the greatest common divisor And, the extent size of the physical extents newly allocated to the logical extent and the logical extent of the logical disk, the extent size changing means for changing the size, which is the detected current extent size,
When the extent size after the change is smaller than the current extent size, the first physical extent allocated to the first logical extent of the logical disk is partitioned by the extent size after the change. Data copying means for copying the data corresponding to each of the areas to the second physical extent group having the changed extent size,
The extent size changing unit divides the first logical extent into a second logical extent group having the changed extent size, and the second logical extent group in which the corresponding data has been copied is copied to the second logical extent group. A disk array device to which physical extents are allocated. An array controller for controlling access to at least one array, wherein a physical area of the at least one array is divided into a plurality of physical extents and managed, and one or more of the plurality of physical extents are logically managed An extent change method in a storage system comprising a disk array device having an array controller to be allocated to a logical extent of a disk, and a host device using the disk array device,
The data arrangement analyzing means of the host device has a first size of a continuous first area used by the host device on the logical disk and a continuous not used by the host device on the logical disk. Detecting a second size of the second area based on a file system used by an operating system operating on the host device;
The data arrangement analyzing means obtains a frequency distribution indicating the number of appearances of the detected first size and second size based on the detected first size and second size. Calculating the greatest common divisor of the size up to a certain high order with the highest frequency, and calculating the greatest common divisor, or a size that is an integer multiple of the greatest common divisor. a step of detect the size of the range not exceeding the maximum value of the size was,
The data arrangement analyzing means, and notifying the detected sizes to the array controller,
Extent size changing means of said array controller, and changing the extent size of the physical extents newly allocated to the logical extent and the logical extent of the logical disk, the size that is the notification from the current extent size,
When the extent size after the change is smaller than the current extent size, the data copy unit of the array controller changes the first physical extent allocated to the first logical extent of the logical disk. Partitioning by the extent size after, and copying the data corresponding to each partitioned area to the second physical extent group of the extent size after the change;
The extent size changing means divides the first logical extent into a second logical extent group having the changed extent size, and the second data in which the corresponding data has been copied to the second logical extent group. An extent changing method comprising: allocating physical extent groups.

Images