JP5534135B2 - Information processing apparatus, information processing method, and program - Google Patents

Description

  The present invention relates to a technique for logical disk allocation during file system defragmentation (hereinafter referred to as defragmentation).

If the file system is used for a long time and file creation / deletion is repeated, the used area of the file system becomes fragmented. When defragmenting a file system in order to eliminate this fragmentation, make sure that the file system has enough free space, or allocate temporary space for defragmentation to the host computer and (For example, patent document 1).
JP 2005-284632 A

  However, if a free area is always secured in the file system for defragmentation, a defragmentation area is required for each LD of each file system. As a result, it was found that the storage area utilization efficiency deteriorated.

  Further, it has been found that the method for allocating an LD, which is a temporary area from the storage, to the host computer as necessary requires extra processing for recognizing the LD to which the host computer is allocated.

  In addition, it was found that defragmentation uses a free area to save the copy source and then rearranges it, so that the number of times of copying increases and defragmentation takes time.

  The present invention has been made in view of the above problems. That is, it is an object of the present invention to perform defragmentation of a file system of a host device without preparing a file system defragmentation area for each of a plurality of host computers in a storage.

  The present invention for solving the above-mentioned problems is a memory management system, in which data in a data area to which a real area of a memory pool is allocated is defragmented so that fragmentation of the data is eliminated. And a control means for controlling to allocate a real area of the memory pool to the defragmentation area when data is written to the defragmentation area by the writing means. Features.

  The present invention for solving the above-mentioned problem is a memory management device, and when data in a data area to which a real area of a memory pool is allocated is written to the defragmentation area, the memory is stored in the defragmentation area. Control means for controlling to allocate a real area of the pool is provided.

  The present invention for solving the above-described problem is a memory management device, and writes data in a data area to which a real area of a memory pool is allocated to a defragmentation area so that data fragmentation is eliminated Write means; and control means for controlling to allocate a real area of the memory pool to the defragmentation area when data is written to the defragmentation area by the write means. .

  The present invention for solving the above-mentioned problem is a memory management method, comprising: a writing step for writing data in a data area to which a real area of a memory pool is assigned to a defragmentation area; and a defragmentation area. And a control step for controlling to allocate a real area of the memory pool to the defragmentation area when the data is written.

  The present invention for solving the above problem is a program for an information processing apparatus, wherein the program stores data in a data area to which a real area of a memory pool is allocated in the information processing apparatus as a defragmentation area. When written, control processing is executed to allocate the real area of the memory pool to the defragmentation area.

The present invention for solving the above-mentioned problem is a program for an information processing apparatus, and the program stores data in a data area to which an actual area of a memory pool is assigned to the information processing apparatus during defragmentation. Write processing to write to the defragmentation area so that data fragmentation is eliminated,
When data is written to the defragmentation area by the write process, a control process is executed to control to allocate the real area of the memory pool to the defragmentation area.

  Defragmentation can be performed without having defragment areas for the number of file systems in the storage by allocating real areas to the defragment areas of the host computer at the time of defragmentation.

The figure which showed an example of embodiment in this invention. The figure which showed an example of the structure of the file system management table 133. FIG. The figure which showed an example of the structure of the mapping table. The figure which showed the specific example of the file system management table 133 before defragmentation implementation. The figure which showed the specific example of the mapping table 251 before defragmentation implementation. 7 is a flowchart showing the operation of the defragmenter 131. The figure which showed the specific example of the file system management table 133 in execution of defragmentation. The figure which showed the specific example of the mapping table 251 in execution of defragmentation. The figure which showed the specific example of the file system management table 133 after defragmentation implementation. The figure which showed the specific example of the mapping table 251 after defragmentation implementation. 6 is a flowchart of the operation of the data copy unit 170. The flowchart of operation | movement of the area | region switching part 120. FIG. The flowchart of operation | movement of the area | region release part 246. The figure which showed another example of embodiment in this invention. The flowchart of operation | movement of the defragmenting request | requirement part 131 in another Example. The flowchart of operation | movement of the area | region copy part 270 in another Example.

  In the present invention, a logical disk (Logical Disk: LD) having a virtual capacity that is larger than an area required for normal operation by the file system of the host computer is constructed on a storage pool (memory pool). . Then, the constructed logical disk is assigned to the host computer.

  The host computer side creates a file system on the logical disk allocated to the own computer, and creates an area of the size required by the host computer on the file system (hereinafter, this area is called a normal area). . The file system uses a normal area as a data area. An area that is not used as a data area of the file system is a defragmentation area and is not used for other purposes.

  At this time, a real area is allocated from the pool to the LD having the virtual capacity in response to writing from the host computer. For this reason, the real area is assigned to the normal area. However, no real area is assigned to the defragmentation area. As a result, the real area of the pool is used for the same area as the area normally used by the file system.

  When defragmenting the file system, the data in the data area is copied to the defragmentation area so that data fragmentation is eliminated. At this time, the real area of the pool is assigned to the defragmentation area to which the real area of the pool has not been assigned.

<First embodiment>
Details of the present embodiment will be described below.

  The embodiment shown in FIG. 1 has a host computer 100 and a storage 200.

  The host computer 100 creates a file system in the LD having the virtual capacity of the allocated storage 200. At this time, an area corresponding to the size required by the host computer is created on the file system. In response to writing by the data processing unit of the host computer 100, a real area is assigned to the LD from the pool, so that the real area is assigned to the normal area of the file system.

  The host computer 100 includes a file system 110, an area switching unit 120, a data processing unit 130, a defragmenting request unit 131, a file system management table 133, and a data copy unit 170.

  The file system 110 is composed of a plurality of data blocks divided in logical units. These data blocks are divided into a normal area 112 that is an area that is normally used and a defragment area 114 that is used for defragmentation.

  The file system management table 133 is for managing the file system 110. FIG. 2 is an example of the configuration of the file system management table 133. Each column of the file system management table 133 will be described. The file number is a number that uniquely identifies a file created in the normal area 112 by the host computer. The file offset number is a number indicating the sequence order of data blocks constituting a file when one file is created across a plurality of data blocks on the file system. The data block number is identification information for uniquely identifying a data block, and indicates a data block in which file data is stored. The use is information indicating whether the data block belongs to the normal area 112 or the defragment area 114. The usage state is information indicating whether or not the data block is used (whether or not a file is created on the data block). The defragmentation state is information indicating whether or not the defragmentation of the data block has been completed.

  The defrag request unit 131 refers to and updates the file system management table 133 and instructs the data copy unit 170 to perform defragmentation. Further, the defrag request unit 131 requests the region switching unit 120 to switch the region, and requests the region release unit 246 provided in the storage 200 to release the region.

  The region switching unit 120 switches between the normal region 112 and the defragmentation region 114 in response to a request from the defragmentation request unit 131.

  The data processing unit 130 generates a file in the normal area 112 of the file system of the file system 110. That is, data is written to the LD 221 of the storage 200 (Write). Further, the data processing unit 130 updates the file system management table 133 in accordance with file generation.

  When receiving a request from the defrag request unit 131, the data copy unit 170 copies the data written in the normal area 112 to the defrag area 114. In this description, the data copy unit 170 and the data processing unit that creates the file system in the LD of the storage 200 are described as different configurations. However, the same configuration unit may be used. good.

  The storage 200 includes a physical disk group 240, a pool 230, an LD group 220, an LD 221, an area release unit 246, and a mapping table 251.

  The physical disk group 240 is composed of a plurality of physical disks 241 provided in the storage 200.

  The pool 230 is a virtual storage area composed of the physical disk group 240. The pool 230 is composed of a plurality of pool blocks obtained by dividing the pool into logical units. This pool block is divided into an allocated area 232 that is an area allocated to the LD 221 and an unallocated area 234 that is an area where allocation is not performed. The allocated area 232 is an area allocated as an actual area to the LD 221 in accordance with writing to the LD 221 by writing a file in the normal area 112 of the file system 110.

  The LD 221 is a logical disk having a virtual capacity. This logical disk (LD221) is used by the host computer 100. When a file is written in the normal area 112 of the file system 110, an LD 221 having the same capacity as the capacity in which the file is written is used. The LD group 220 is a collection of LDs used by each of a plurality of host computers.

  In response to a request from the defrag request unit 131, the area release unit 246 refers to and updates the mapping table 251. When the real area is assigned to the defragmentation area, the assigned content is reflected in the mapping table 251. Further, the area releasing unit 246 releases the allocated area 232 allocated to the LD 221 and returns the area to the unallocated area 234.

  The control unit 260 allocates a pool to the LD 221 as a real area in response to writing to the LD 221 by writing a file to the normal area 112 of the file system 110. Further, when copying from the normal area of the file system to the defragmentation area, the real area on the pool for the defragmentation area is stored in the LD 221 of the storage 200 according to the write from the host computer 100. Further assigned by 260. The assigned contents are reflected in the mapping table 251. In this description, the region opening unit 246 and the control unit 260 are described as separate components, but they may be the same.

  The mapping table 251 is for managing the pool 230. FIG. 3 is an example of the configuration of the mapping table 251. Each column in FIG. 3 will be described.

  The pool block number is a number that uniquely identifies a block (pool block) obtained by dividing the pool into logical units. The LD number is a number that uniquely identifies an LD that uses the pool block. The data block number is a number that uniquely identifies a block (data block) obtained by dividing the file system into logical units. The area size is information indicating the size of the pool block.

  The detailed operation of the embodiment of the present invention will be described with reference to the drawings.

  FIG. 4 is a diagram showing a specific example of the file system management table 133 before defragmentation in the present embodiment. Here, the contents shown in FIG. 4 will be described.

  Data blocks “1” to “12” of the file system 110 are set in the normal area 112. Further, data blocks “13” to “24” of the file system 110 are set in the defragmentation area 114. File “1” is stored in data blocks “1”, “2”, and “4”. File “2” is stored in data block “12”. File “3” is stored in data blocks “6” and “7”. File “4” is stored in data blocks “5”, “8”, “9”, and “11”.

  FIG. 5 is a diagram showing a specific example of the mapping table 251 before defragmentation in the present embodiment. Here, the contents shown in FIG. 5 and the contents shown in FIG. 4 will be described in correspondence with each other. The description will be made assuming that the LD number of the LD 221 used by the host computer 100 is “1”.

  As described above, the data blocks set as the normal area 112 of the file system 110 are the data blocks “1” to “12”. Pool blocks “1” to “12” are assigned as physical areas to the data blocks “1” to “12”, respectively. In this description, since the host computer 100 uses the LD 221 (LD number “1”), the LD numbers “1” are used for the pool blocks “1” to “12”. In the present invention, since no physical area is allocated to the defragmentation area 114 of the file system 110, the data block number of the defragmentation area 114 is not shown in the mapping table 251. Further, here, an example in which the total area size of the pool 230 is set to 256 is shown.

  The operation of the defragmenter 131 will be described with reference to the flowchart of FIG. 6 using the case where the state before defragmentation is the state shown in FIGS. 4 and 5.

  Upon receiving a defragmentation command from the user, the defragmentation requesting unit 131 searches the file system management table 133 and acquires all records whose usage status is “used” and records in the defragmentation area (S100). For example, the data block numbers “1”, “2”, “4”, “12”, “6”, “7”, “5”, “8” in the file system management table 132 shown in FIG. Records “9” and “11” are acquired.

  The defragmentation request unit 131 acquires data block numbers in order from the lowest file offset number for the records having the same file number among the records acquired in S100. For each acquired data block number, the data block numbers in the defragmentation area are combined one by one in order from the lowest number (S110). In the file system management table 133 of FIG. 4, when processing is performed on file number 1, data block numbers 1, 2, and 4 are acquired in order. Data block numbers 1 and 13, 2 and 14, 4 and 15 are combined. Similarly, when processing is performed on file number 2, data block number 12 is acquired and data block numbers 12 and 16 are combined. Further, when processing is performed on the file number 3, the data block numbers 6 and 7 are obtained in order, and the data block numbers 6 and 17, and 7 and 18 are combined. Further, when processing is performed on the file number 4, data block numbers 5, 8, 9, and 11 are acquired in order, and the data block numbers 5 and 19, 8 and 20, 9 and 21, and 11 and 22 are combined. .

  The defrag request unit 131 designates each data block number acquired in S110 as a copy source data block number, designates each combined data block number as a copy destination data block number, and sends it to the data copy unit 170. A request to start copying is made (S120).

  Upon receiving a copy completion report from the data copying unit 170 (S210 described later), the defragmenting request unit 131 reflects the defragmentation result in the file system management table 133. The record that is the copy source or the copy destination in S110 becomes “Done” (S130).

  The defragmentation requesting unit 131 checks whether or not there is a record in which the use state is “used” and the defragmentation state is “unexecuted” among the records acquired in S100. If there is a corresponding record, the operation is performed from S110 on the file number of the corresponding record. If there is no corresponding record, the file copy is terminated (S140).

  After S140, the file system management table 133 in FIG. 4 is in the state shown in FIG. Further, when copying from the normal area of the file system to the defragmentation area, the control unit 260 assigns the real area on the pool to the LD 221 of the storage 200 in accordance with the write from the host computer 100. Therefore, after S140, the mapping table 251 in FIG. 5 is in the state of FIG. As shown in FIG. 8, pool blocks “13” to “24” are assigned to the data blocks “13” to “24” which are set in the defragmentation area 114 of the file system 110 and to which data is written.

  The defragmentation requesting unit 131 requests the region switching unit 120 to switch the region between the normal region and the defragmentation region (S150).

  When the defrag request unit 131 receives a report of the completion of the area switching from the area switching unit 120 (S300 described later), the defrag request unit 131 searches the file system management table 132, acquires all the records in the normal area, and sets the data block number of the record. get. Then, the defrag request unit 131 designates the acquired data block number as an area to be released, and requests the area release unit 246 to release the area (S160).

  Upon receiving the area release completion report from the area release unit 246 (S420 described later), the defrag request unit 131 reflects the result of area switching and area release on the file system management table 133 (S170).

  After the execution of S170, the file system management table 133 in FIG. 7 is in the state shown in FIG. As shown in FIG. 9, the data blocks “13” to “24” set in the defragmentation area of the file system 110 become normal areas, and the data block “1” set in the normal area. To “12” is a defragmentation area. Further, after the execution of S170, the mapping table 251 in FIG. 8 is in the state shown in FIG. As shown in FIG. 10, since the pool blocks assigned to the defragmentation area 114 are assigned from “13” to “22” as normal areas, the pool blocks are data blocks assigned to “13” to “22”, respectively. “13” to “22” are assigned. On the other hand, since the pool blocks assigned to the normal area are assigned “1” to “12” as the defragmentation area, the data blocks are not assigned to the pool blocks “1” to “12”, and the LD The number is also unassigned.

  FIG. 11 is a flowchart showing the operation of the data copy unit 170 in this embodiment.

  Upon receiving the copy source data block number and the copy destination data block number from the defrag request unit 131 in S120, the data copy unit 170 reads the contents of the copy source data block (S200).

  The data copy unit 170 writes the data read in S200 to the copy destination data block, and reports the copy completion to the defrag request unit 131. At this time, a write to a block that is not an actual area of the LD 221 occurs along with the copy. The pool 230 allocates the real area to the LD according to the write, and reflects the allocation result in the mapping table 251 (S210).

  FIG. 12 is a flowchart showing the operation of the area switching unit 120 in this embodiment.

  Upon receiving a region switching request from the defragment request unit 131 in S150, the region switching unit 120 switches between the normal region 112 and the defragmentation region 114, updates the file system management information, and sends the region to the defragmentation unit 131. The completion of switching is reported (S300).

  FIG. 13 is a flowchart showing the operation of the area release unit 246 in the present embodiment.

  When the area release unit 246 receives the data block number and the area release request from the defragment request unit 131 in S160, the area release unit 246 searches the mapping table 251 and acquires a record corresponding to the data block number (S400).

  The area releasing unit 246 acquires the pool block number from the record acquired in S400, releases the area of the pool block number from the allocated area 232, and incorporates it into the unallocated area 234 (S410).

  The area release unit 246 reflects the result of the area release on the mapping table 251 and reports the completion of the area release to the defrag request unit 131 (S420).

  According to the present embodiment, the real area is not allocated to the defragment area during normal operation, and the real area is allocated to the defragment area during defragmentation. Therefore, the defragment area is provided for the number of file systems of each computer in the storage. Without defragmentation. In addition, since a real area having the same size as the area used by the file system is allocated to the file system, a process for recognizing the LD to which the host computer is allocated becomes unnecessary. Also, allocate an area of the same size as the data area currently used by the file system, copy the data area to the defragmentation area to eliminate fragmentation, and switch between the normal area and the defragmentation area after the copy is completed. As a result, it is possible to reduce the time required for defragmentation.

<Second Embodiment>
A second embodiment of the present invention will be described below. In the present embodiment, a case where defragmentation is performed in the storage 200 will be described.

  FIG. 14 shows an example in which defragmentation is performed in the storage 200. The difference from FIG. 1 is that there is no data copy unit 170 on the host computer side, the storage 200 has an area copy unit 270, and the defrag request unit 131 instructs the area copy unit 270 to copy. The operations of the area switching unit 120 and the area releasing unit 246 are not different from the above-described embodiment.

  FIG. 15 is a flowchart showing the operation of the defrag request unit 131 when defragmenting is performed in the storage 200. The difference from FIG. 6 is that the copy start request destination is not the data copy unit 170 but the area copy unit 270. In the present embodiment, the case where the area copy unit 270 and the control unit 260 are separate components will be described, but they may be the same.

  FIG. 16 is a flowchart showing the operation of the area copy unit 270 when defragmenting is performed in the storage 200.

  Upon receiving the copy source data block number and the copy destination data block number from the defragment request unit 131, the area copy unit 270 searches the mapping table 251 and acquires a record corresponding to the copy source. Further, an unallocated area having an area size capable of storing all copy source areas is searched, and the record is acquired as a copy destination record (S600).

  The area copy unit 270 acquires the copy source pool block number and the copy destination pool block number from the record acquired in S600, and copies the contents of the pool block from the copy source pool block to the copy destination pool block. (S610).

  The area copy unit 270 reflects the copy result in the mapping table 251 and reports the copy completion to the defrag request unit 131 (S630).

  According to the present embodiment, as in the first embodiment, the real area is not allocated to the defragmentation area during normal operation, and the real area is allocated to the defragmentation area during defragmentation. Defragmentation can be performed without having defragmentation areas for the number of systems. In addition, since a real area having the same size as the area used by the file system is allocated to the file system, a process for recognizing the LD to which the host computer is allocated becomes unnecessary. Furthermore, since copying is performed by the storage 200 instead of the host computer side, defragmentation can be performed without imposing a load on the host computer.

  According to the above embodiment, defragmentation can be performed without having free space for defragmentation for each of a plurality of file systems in the storage by performing defragmentation by temporarily using the unallocated area on the pool. Therefore, it is possible to improve the storage area utilization efficiency.

  Further, according to the above embodiment, there is no need to separately allocate a temporary area LD to the host computer, so that a procedure for separately identifying the LD is not necessary when defragmentation is performed.

  Also, allocate an area of the same size as the data area currently used by the file system, copy the data area to the defragmentation area to eliminate fragmentation, and switch between the normal area and the defragmentation area after the copy is completed. As a result, it is possible to reduce the time required for defragmentation.

  Further, according to the above embodiment, the use area on the file system is copied to the switching destination area, and all the copy source areas are released after defragmentation, so that it is possible to improve the storage area utilization efficiency.

  Note that each device of the above-described embodiment can be configured by hardware as is apparent from the above description, but can also be realized by a computer program. In such a case, functions and operations similar to those of the above-described embodiment are realized by a processor that operates according to a program stored in the program memory. Note that some of the functions of the above-described embodiments can be realized by a computer program.

100 Host computer 110 File system 112 Normal area 114 Defragment area 120 Area switching part 131 Defragment request part 133 File system management table 170 Data copy part 200 Storage 220 LD group 221 LD
230 Pool 232 Allocated area 234 Unallocated area 240 Physical disk group 241 Physical disk 246 Area release unit 251 Mapping table 260 Control unit

Claims (7)

A memory management system comprising storage and a host computer that creates a file system on a logical disk having a virtual capacity of the storage,
Writing means for writing data in the data area of the file system to which the real area of the memory pool is allocated to the defragmentation area of the file system so as to eliminate fragmentation of the data;
Control means for controlling the real area of the memory pool to be allocated to the defragmentation area when data is written to the defragmentation area by the writing means ;
When the writing means finishes writing to the defragmentation area, the host computer has switching means for switching so that the defragmentation area becomes a data area and the data area becomes a defragmentation area. A memory management system characterized by that.   2. The memory management system according to claim 1, further comprising a release unit that releases a real area allocated to the data area when the writing unit finishes writing to the defragmentation area. A memory management device that creates a file system on a logical disk having a virtual capacity of storage,
Writing means for writing data in the data area of the file system to which a real area of the memory pool is allocated to the defragmentation area of the file system so that data fragmentation is eliminated;
Control means for controlling the real area of the memory pool to be allocated to the defragmentation area when data is written to the defragmentation area by the writing means ;
Switching means for switching so that the defragmentation area becomes a data area and the data area becomes a defragmentation area when the writing means finishes writing to the defragmentation area. A memory management device. 4. The memory management device according to claim 3, further comprising: a release unit that releases a real area allocated to the data area when the writing unit finishes writing to the defragmentation area . A memory management method of a memory management device for creating a file system on a logical disk having a virtual capacity of storage,
A writing step of writing the data in the data area of the file system to which the real area of the memory pool is allocated to the defragmentation area of the file system ;
A control step for controlling the real area of the memory pool to be allocated to the defragmentation area when the data is written to the defragmentation area ;
A switching step of switching the defragmentation area to be a data area and the data area to be a defragmentation area when writing to the defragmentation area is completed in the writing step. Memory management method. 6. The memory management method according to claim 5 , further comprising a release step of releasing a real area allocated to the data area when writing to the defragmentation area is completed. A program for an information processing apparatus for creating a file system on a logical disk having a virtual capacity of storage, the program being stored in the information processing apparatus,
A writing process for writing data in the data area of the file system to which the real area of the memory pool is allocated to the defragmentation area of the file system so that data fragmentation is eliminated during defragmentation;
When data is written to the defragmentation area by the write process, a control process for controlling to allocate the real area of the memory pool to the defragmentation area ;
When writing to the defragmentation area is completed in the writing step, a switching process for switching the defragmentation area to be a data area and the data area to be a defragmentation area is executed. A featured program.

Images