JP3967080B2 - Virtual disk control system and virtual disk control program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a virtual disk control system and a virtual disk control program for randomly writing data to a disk device.
[0002]
[Prior art]
Conventionally, as shown in FIG. 7, data is written to the disk device by repeatedly positioning the head of the disk device at a predetermined position and then writing the data. Was written at random positions.
[0003]
Further, the backup is saved on an external medium such as a magnetic tape to save the data.
[0004]
[Problems to be solved by the invention]
As described above, in writing data to the conventional disk device, as shown in FIG. 7, it is necessary to repeat the operations of seek and write every time data is written, and the seek occurs every time and cannot be written at high speed. There was a problem.
[0005]
In addition, in order to back up data on the disk device, it is necessary to continuously read and write data on the disk device and write them sequentially to an external medium such as a magnetic tape, which takes a long time. there were.
[0006]
The present invention aims to solve these problems.
[0007]
[Means for Solving the Problems]
Means for solving the problem will be described with reference to FIG.
In FIG. 1, a virtual disk device 4 is a virtual disk device in which applications 2, 3 and the like try to write data at random positions.
[0008]
The random write / sequential write conversion mechanism 5 performs random write to the virtual disk device 4 in order to the real disk device 12, and is composed of a data mapping table 8, a backup program 10, and the like. .
[0009]
The data mapping table 8 stores a random write position on the virtual disk device 4 and a sequential write position on the real disk device 12 in association with each other.
[0010]
The backup program 10 backs up data on the real disk device 12.
Next, when there is a write to a random position from the applications 2 and 3 for explaining the operation to the virtual disk device 4, the random write / forward write conversion mechanism 5 writes sequentially from a predetermined position of the real disk device 12, A random writing position in the virtual disk device 4 and a position sequentially written in the real disk device 12 are associated with each other and stored in the data mapping table 8.
[0011]
At this time, when writing to the data mapping table 8 by associating the position written in the random position of the virtual disk device 4 and the position sequentially written in the real disk device 12 into the data mapping table 8, the virtual disk device 4 is already stored in the data mapping table 8. When the position of the real disk device 12 is stored in association with the writing position, the position is updated with the current position of the real disk device 12, while when it is not stored, the position of the current real disk device 12 is updated. Is written newly.
[0012]
Further, when the backup program 10 instructs to backup, the data mapping table 8 is copied and stored, and the data at the position of the real disk device 12 registered in the entry on the data mapping table 8 is set to prohibit deletion. ing.
[0013]
Accordingly, data written to an arbitrary position on the virtual disk device 4 is sequentially written on the real disk device 12, and the correspondence between both is stored in the data mapping table 8 to realize high-speed writing, and at the time of backup, the data mapping table It is possible to realize high speed backup by copying 8 and prohibiting deletion of saved and saved data.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Next, embodiments and operations of the present invention will be described in detail sequentially with reference to FIGS.
[0015]
FIG. 1 shows a system configuration diagram of the present invention.
In FIG. 1, a processing device 1 is a computer having a processor and a memory, and executes various processes according to a program. Here, application programs (hereinafter simply referred to as applications) 2 and 3, a virtual disk device 4, And a random light / forward light conversion mechanism 5 and the like.
[0016]
The applications 2 and 3 execute various processes by writing data to random positions of the virtual disk device 4.
The virtual disk device 4 is a virtual disk device 4 to which the applications 2 and 3 write data at random positions. Writes to random positions from the applications 2 and 3 to the virtual disk device 4 are taken in by the random write / forward write conversion mechanism 5, converted to forward write, and written sequentially to the real disk device 12.
[0017]
The random write / sequential write conversion mechanism 5 takes in data written at random positions in the virtual disk device 4 from the applications 2 and 3, converts the data into sequential write, writes it to the real disk device 12, and on the virtual disk device 4. The write position and the position written in the real disk device 12 are associated with each other and stored in the data mapping table 8 for management. Here, the data mapping management program 6, the data bitmap management program 7, a data mapping table 8, a data bitmap table 9, a backup program 10, a real disk I / O management program 11, and the like.
[0018]
The data mapping management program 6 manages the data mapping table 8, and associates the positions of the virtual disk device 4 to which data is written at random and the positions of the real disk device 12 that are sequentially written from a predetermined position. It is written and stored. For example, when data is newly written, the write position of the virtual disk device 4 and the current write position of the real disk device 12 are newly stored in association with each other. The write position of the real disk device 12 stored in association with the write position of the disk device 4 is updated with the current write position of the real disk device 12 (described later with reference to FIGS. 2 to 5).
[0019]
The data bitmap management program 7 is a table for managing empty sectors in the real disk device 12, and manages by setting the distinction between the free area and the used area in association with the disk number and sector number. (See FIG. 3B).
[0020]
The data mapping table 8 stores a random write position of the virtual disk device 4 and a sequential write position of the real disk device 12 in association with each other (see FIG. 3A).
[0021]
The data bitmap table 9 is a table for managing free sectors of the real disk device 12 (see FIG. 3B).
The backup program 10 backs up data on the real disk device 12, and copies and saves the data mapping table 8 and sets a sector deletion prohibition flag in the data mapping table 8 (or prohibits deletion of the sector). And the like (which will be described later with reference to FIG. 6).
[0022]
The real disk I / O management program 11 sequentially writes data to the real disk device 12 (to be described later with reference to FIGS. 2 to 6).
The real disk device 12 is a disk device to which data written at random positions on the virtual disk device 4 is actually written sequentially.
[0023]
FIG. 4 is an explanatory diagram of the present invention.
4, the data mapping table 8 is provided with an entry corresponding to an area (virtual address, for example, sector number) divided from the head of the virtual disk device 4 as shown in FIG. The position (real address, sector number) where the data is sequentially written is stored. In the state shown,
Since the data was written in the random position order of the virtual addresses VA4, VA2 and VA5 on the virtual disk device 4,
In the first time, the data mapping table 8 stores the real address RA1 in which data is sequentially written to the real disk device 12 from the VA4 of the virtual disk device 4 to the position indicated by the arrow.
Second, the data mapping table 8 stores the real address RA2 in which data is sequentially written to the real disk device 12 from the VA2 of the virtual disk device 4 to the position indicated by the arrow.
In the third time, the data mapping table 8 stores the real address RA3 in which data is sequentially written to the real disk device 12 from the VA 5 of the virtual disk device 4 to the position indicated by the arrow.
[0024]
Next, since new data has been written to the data of the virtual address VA2 on the virtual disk device 4,
In the fourth time, the data mapping table 8 is updated with the real address RA4 in which data is sequentially written to the real disk device 12 from the VA2 of the virtual disk device 4 to the position indicated by the arrow (update from RA2 to RA4 in the drawing).
[0025]
As described above, when data is written to the virtual disk device 4, the data is sequentially written to the new real address by the real disk device 12, and the written real address RA is pointed to by the virtual disk device 4. By writing or updating the real address in the corresponding position 8, the random write is converted into the sequential write in the virtual disk device 4 and executed in the virtual disk device 4, and the virtual address and the real address are stored in the data mapping table 8. It can be stored and managed, and high-speed writing by forward writing is possible.
[0026]
FIG. 3 shows an example table of the present invention.
FIG. 3A shows an example of the data mapping table 8. The data mapping table 8 is a management table of corresponding sectors of the virtual disk device 4 and the real disk device 12, and for example, the following information shown in FIG.
[0027]
-Flag (Shared Flag):
・ 1: Sector is shared with the previous version (already backed up)
-0: The sector is not shared with the previous version (not backed up)
・ Disk number:
・ Sector number:
Here, the flag (Shared Flag) indicates whether or not the data in the sector of the real disk device 12 is saved, and in the case of 1, it indicates that the data is shared with the old version data and is set to prohibit deletion. , 0 indicates that the data is not shared with the data of the previous version and that deletion is not prohibited. The disk number and sector number store the correspondence between positions (areas, here, sectors) at which data is written between the virtual disk device 4 and the real disk device 12 (see FIG. 4).
[0028]
As described above, the flag (deletion prohibition flag), the disk number, and the sector number are registered in the data mapping table 8, and the randomly written position (sector) of the virtual disk device 4 and the actual disk device 12 are actually written in order. It is possible to store and manage the positions (sectors) associated with each other. The correspondence management of both sectors is performed by storing the sector of the virtual disk device 4 and the sector of the real disk device 12 as a pair, or the correspondence of the real disk device 12 to the entry corresponding to the sector of the virtual disk device 4. For example, a sector to be stored may be stored.
[0029]
FIG. 3B shows an example of the data bit map table 9. The data bit map table 9 is a management table of empty sectors of the real disk device 12, and stores the following information shown in association with each other.
[0030]
・ Sector number:
-Bit 0 or 1:
0: Empty area 1: In use Here, data is written to the real disk device 12 in order in order by storing bit 0 (empty area) or bit 1 (in-use area) in association with the sector number. Thereafter, empty areas close to the head are searched and written in order, and the areas can be reused at high speed and simply, and efficient sequential writing of the real disk device 12 can be performed.
[0031]
FIG. 2 shows a flowchart for explaining the operation of the present invention.
In FIG. 2, S1 accepts a write request.
In S2, it is determined whether or not writing is the first time. This is because the correspondence between the random write position of the virtual disk device 4 and the sequential write position of the real disk device 12 is not stored in the data mapping table 8 of FIG. Determine. In the case of YES, since it is determined that the writing is the first time, the process proceeds to S5. On the other hand, in the case of NO, since it is found that the data mapping table 8 has already been stored, the process proceeds to S3.
[0032]
In S3, since it has been determined that the writing is not the first time in S2, it is further determined whether or not it is already a backup. This determines whether the data at the write position accepted in S1 has already been backed up and the flag of the corresponding sector in the data mapping table 8 (deletion prohibition flag, Shared flag) is set. In the case of YES, since it is found that the data is backed up and set to prohibit deletion, the process proceeds to S5. On the other hand, in the case of NO, it has been found that the data has not been backed up and that deletion prohibition has been set, so the allocation information of the sector already written in S4 is released (the corresponding entry in the data bitmap table 9 is set to 0). ), And proceed to S5.
[0033]
S5 allocates a new sector.
FIG. 5 shows an operation explanation flowchart (sector allocation) of the present invention.
In FIG. 5, S11 determines whether there is a free area. This is to determine whether the real disk device 12 has a free area to be written in order. In the case of YES, a new sector is allocated in S12 (an empty new sector in the real disk device 12 is allocated), and subsequently data is sequentially written to the allocated sector. On the other hand, in the case of NO in S11, since it has been found that there is no free area to be allocated to the real disk device 12, the process proceeds to S13.
[0034]
In S13, an empty sector near the head of the real disk device 12 is searched. This is done by referring to the data bit map table 9 in FIG. 3B described above to search for an empty sector from the top (search for an empty sector of bit 0 from the top).
[0035]
S14 determines whether there is an empty sector. In the case of YES, since an empty sector has been found, an empty sector is assigned in S15, and subsequently data is sequentially written into the assigned sector. On the other hand, in the case of NO in S14, since it has been found that there is no empty sector, in S16, a write request is made an error due to insufficient empty area.
[0036]
As described above, if there is an empty area (empty sector) in sequence in the real disk device 12, it is assigned and data is written in order. On the other hand, when there is no more empty area, the data bitmap table 9 is referred to from the beginning. The empty area (empty sector) found by searching for the empty area (empty sector) is allocated and the data is sequentially written, so that the empty area of the real disk device 12 can be used effectively.
[0037]
In S6 of FIG. 2, the data mapping table 8 is updated. This is written (updated) and stored by associating the random writing position of the virtual disk device 4 with the writing position to the real disk device 12 in order.
[0038]
S7 writes the data to the real disk device.
With the above procedure, a write request to a random arbitrary position of the virtual disk device 4 is accepted, and the data mapping table 8 is referred to. When the data has already been written and the data has not been backed up, the data is stored. After the area (sector) is released so that it can be reused, a new area (sector) is allocated from a predetermined position of the real disk device 12 in a direction in which the moving direction of the head becomes a constant direction, and data is written in order and data mapping is performed. The table 8 stores the randomly written position of the virtual disk device 4 and the written position in the order of the real disk device 12 in association with each other, so that writing to the random position of the virtual disk device 4 from an application or the like can be performed. Actually, it is converted into a write to the position of the sequential area of the real disk device 12 and executed. It is possible to realize a high-speed light due.
[0039]
FIG. 6 shows a backup explanatory diagram of the present invention. This is to explain the procedure and operation when data is backed up with the configuration shown in FIG.
FIG. 6A shows a backup flowchart.
[0040]
In FIG. 6A, S21 accepts a backup request. In this case, the random write / forward write conversion mechanism 5 in FIG. 1 accepts a backup request from the applications 2 and 3 or the administrator.
[0041]
In S22, the data mapping table 8 is copied. In this case, the data mapping table 8 shown in FIGS. 1 and 3A and FIG. 4 is copied and stored.
In S23, the flags of all entries in the copy source data mapping table 8 are set to "1". This is because the sector flags of all entries in the copy source (current) data mapping table 8, for example, the sector flags (Shared Flag) of all entries in FIG. Set to). As a result, since the sector flags (deletion prohibition flag) of all entries in the current data mapping table 8 are set, the data of all sectors managed in the data mapping table 8 are set to prohibit deletion, This means that all the sector data in the data mapping table 22 copied and saved in S22 has been saved (saved). As described above, the data mapping table 22 can be copied and saved, and the data of all sectors can be backed up simply by setting the flags of the sectors of all entries in the data mapping table 22. Backup is possible.
[0042]
FIG. 6B shows a backup explanatory diagram. The upper (current copy source) data mapping table 8 in the drawing is copied and saved as shown in the lower row, and the sectors of all entries in which the real addresses of the upper current data mapping table 8 are registered are stored. By setting this flag (deletion prohibition flag) to 1 (deletion prohibition), all data can be backed up.
[0043]
Since the data of the virtual address VA4 of the virtual disk device 4 is updated after the backup, the data is written to the sector (real address RA5) in order of the real disk device 12, and the virtual address VA4 of the virtual disk device 4 is written. Update to 5 (RA5) indicated by a right-pointing arrow in the corresponding data mapping table 8 (Note that the flag (deletion prohibition flag) of the old sector RA1 is set to 1 and the backup is not deleted. Have been). As described above, it is possible to update the data using the current data mapping table 8 after the backup, and the deletion prohibition flag of the backed up data (sector) is set so that any data (sector) is stored. Whether the data has been backed up is determined from the data mapping table 8 that has been copied and saved.
[0044]
At the time of backup, the data mapping table 8 is copied to store which data is stored, and the sector itself on the actual disk device 12 in which the stored data is written or a sector management table provided separately is stored. You may make it set the deletion prohibition bit (backed up bit) of the sector.
[0045]
【The invention's effect】
As described above, according to the present invention, a configuration is adopted in which data to be written at random positions in the virtual disk device 4 is sequentially written on the real disk device 12 and the corresponding positions are stored in the data mapping table 8. Therefore, the random write can be executed at high speed, and the data mapping table 8 can be copied at the time of backup, set to prohibit storage and deletion, and can be backed up quickly and easily.
[Brief description of the drawings]
FIG. 1 is a system configuration diagram of the present invention.
FIG. 2 is a flowchart explaining the operation of the present invention.
FIG. 3 is an example table of the present invention.
FIG. 4 is an explanatory diagram of the present invention.
FIG. 5 is a flowchart for explaining the operation of the present invention (sector allocation).
FIG. 6 is an explanatory diagram of backup according to the present invention.
FIG. 7 is an explanatory diagram of a conventional technique.
[Explanation of symbols]
1: Processing device 2, 3: Application (application program)
4: Virtual disk device 5: random write / forward write conversion mechanism 6: data mapping management program 7: data bitmap management program 8: data mapping table 9: data bitmap table 10: backup program 11: real disk I / O management Program 12: Real disk device

Claims (2)

A data mapping table for managing the correspondence between the address information managed by the physical disk and the virtual address information and the deletion prohibition information ;
In response to a write request that specifies virtual address information, if the specified virtual address information is not managed in the data mapping table, the real mapping area is sequentially allocated to update the data mapping table. If it is managed and the deletion is not prohibited, the allocated area on the real disk is released and the real disk area is allocated sequentially to update the data mapping table. Virtual disk control comprising: allocation processing means for sequentially allocating real disk areas and updating the data mapping table without erasing past history; and writing means for writing requested data in the areas allocated by the allocation processing means system. On the computer,
Data mapping table that manages the correspondence between address information and virtual address information managed by a physical disk and deletion prohibition information in response to a write request that specifies virtual address information If not managed, the real mapping area is allocated sequentially and the data mapping table is updated, and if it is managed and deletion is not prohibited, the allocated area is released on the real disk and the real disk An allocation processing means for sequentially allocating areas and updating the data mapping table, and in the case of management and deletion prohibition, allocating real disk areas in sequence and updating the data mapping table without deleting the past history; and Write the requested data to the area allocated by the allocation processing means Writing means and
Virtual disk control program to function as

Images