JP3246146B2 - Data transfer method for external storage device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a magnetic disk drive transfer system which enables a transfer of a magnetic disk drive.

[0002]

2. Description of the Related Art Conventionally, at the time of maintenance of a magnetic disk device or replacement with a new type of magnetic disk device, backup to another magnetic disk device during a time when files stored in the magnetic disk device are not used. From the old magnetic disk drive to the new magnetic disk drive.

[0003] If there is not enough time to shift the business due to 24-hour operation, the business is temporarily stopped,
There was a problem that a migration work had to be performed.

Further, when a new magnetic disk device is migrated, it is necessary to regenerate the configuration of the magnetic disk device and restart the system in order for the system to recognize the magnetic disk device. Had to stop.

A method of replacing a magnetic disk drive during operation of the system is described in, for example, a manual "Program Product VOS3 Disk Double-Writing Control Program WDCP / ES" (issued in June 1992) issued by Hitachi, Ltd. This is described as "replacement of head disk assembly (HDA)".

[0006]

As described above, in the prior art, at the time of transfer of the magnetic disk drive, it was necessary to temporarily degenerate the business or the operating system or stop it for a certain period of time.

An object of the present invention is to provide a method and an apparatus for migrating data in an external storage device, which enable maintenance / migration work of the external storage device without stopping the application program of the business. Also,
The purpose is to enable migration even if the external storage device to be migrated has a different attribute or form.

[0008]

According to the present invention, a computer is provided.
And a plurality of storage devices.
Data to write data from computer to storage device during copying
The storage method, in which the computer
-Manage the status of the destination storage device for each predetermined area,
When receiving an I / O request to write data to the device,
By processing I / O requests based on the status of
Outside without stopping business application programs
It is possible to perform maintenance / transfer work of the external storage device.

[0009]

[0010]

[0011]

[0012]

According to the present invention, a magnetic disk device serving as a file transfer destination is dynamically incorporated into a system, and combined with a magnetic disk device serving as a file transfer source and subjected to dual writing, so that a higher-level application can store a file. File transfer to a new magnetic disk device can be executed even while an input / output request is being executed to the transfer source magnetic disk device.

Further, by locating the dual writing control means at a position for controlling the physical input / output access means, a magnetic disk device having different physical attributes, or multiplex writing by the physical access method or hardware is executed. File transfer is possible even if the attributes / forms of the magnetic disk device such as a magnetic disk device are different.

Further, even if the magnetic disk device is a magnetic disk device set between remote locations, file transfer is possible.

[0015]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a system configuration diagram showing an embodiment of the present invention including a hardware configuration. FIG. 2 is a device management table for dynamically adding / deleting a device while the system is operating.
FIG. 5 is a diagram showing a means for absorbing a difference in physical attributes and realizing a file migration, and FIG. 8 is a diagram showing a migration between a dual writing disk device in the migration of a multiple writing magnetic disk device. FIG. 9 is a diagram showing disk migration to a remote location.

Referring to FIG. 1, a case where a normal input / output request is executed will be described.

The input / output request executed by the user application 100 is transferred to the operating system using the logical input / output access means 102 in the operating system 101.

The logical input / output access means 102 transfers control to the dual writing control means 104. Dual writing control means 104
Then, it is determined whether the magnetic disk device that is the target of the I / O request is migrating the file, or if the file is being migrated, whether the data area of the target I / O request destination is the double write target area. . When the file is not being migrated, the physical input / output access unit 105 executes an input / output request to the magnetic disk device 114 in which the input / output request destination file of the user application 100 is stored.
An input / output request issued from the operating system 101 is issued to a peripheral device via the input / output processor 110 and the channel 111. In the magnetic disk control device 112 which has received the input / output request issued from the channel 111, the magnetic disk drive device 113
, An input / output request is made to the magnetic disk drive 114 in which the target file is stored.

(Embodiment 1) A procedure for adding a new magnetic disk device to the present system and transferring files when an input / output request is executed according to the above-described procedure will be described.

When a new magnetic disk device is added, the addition of a magnetic disk device management table required by software will be described with reference to FIGS.

FIG. 2 shows an existing magnetic disk drive 21.
0 and 211 and a magnetic disk device management table 200 and 201 corresponding to those magnetic disk devices, a magnetic disk device 212 is newly added to the system in which a corresponding magnetic disk device management table is added. 202 shows a state of being generated on the main memory.

The magnetic disk device management table 202 includes:
A device identifier 220 for identifying the magnetic disk device 212 from other magnetic disk devices in the system, a magnetic disk device address 221 used when an actual I / O is issued to the magnetic disk device, and a physical address of the magnetic disk device. Attributes 222 (track capacity, number of cylinders, presence / absence of additional mechanism, etc.).

Such addition / deletion of the magnetic disk device management table is executed by the dynamic configuration control means 103 in FIG.

FIG. 3 is a flowchart showing how the dynamic configuration control means 103 adds a magnetic disk device management table.

Normally, the magnetic disk device management table is created when the system is started up. Therefore, when adding a magnetic disk device, first check whether a table corresponding to the magnetic disk device to be added already exists. (Step 301). If there is already, since it cannot be added, error processing such as outputting an error message (step 306) is performed and the processing ends. If there is no corresponding table, a lock for adding a magnetic disk device management table is secured. This lock is a global lock existing in the system, and holds the state of updating (addition / deletion) of the table chain. This allows
Exclusion with another additional request process is performed (step 30).
2).

Next, a magnetic disk device management table is created on the main memory (step 303), and the device identifier, magnetic disk device address, physical attributes, etc. of the magnetic disk device to be added are set (step 304). After all the setting processes are completed, the lock for adding the magnetic disk device management table is released (step 305), and an input / output request is accepted.

FIG. 4 is a flow chart showing how the dynamic configuration control means 103 deletes the magnetic disk device management table.

In the case of magnetic disk drive deletion, it is first checked whether a magnetic disk drive management table corresponding to the magnetic disk drive to be deleted already exists (step 401). Otherwise, it cannot be deleted, so error processing such as outputting an error message (step 40)
Perform 5) and end the process. If there is a corresponding magnetic disk device management table, a lock for deleting the magnetic disk device management table is secured. Thus, exclusive control with another deletion request process is performed (step 402).

Next, the information set in the magnetic disk device management table is reset (step 403).
As a result, the input / output request is no longer accepted. Thereafter, the lock for deleting the magnetic disk device management table is released (step 405).

FIG. 5 shows how an I / O request is issued to magnetic disk devices having different physical attributes.

In FIG. 5, a new magnetic disk drive 5
Numeral 11 indicates a state added to the system. The magnetic disk device 509 and the magnetic disk device 511 have different physical attributes such as different track capacities, different numbers of mounted cylinders, different additional mechanisms (with disk cache), and use the same physical access method. I / O request cannot be issued.

Therefore, for an input / output request to the magnetic disk devices 509 and 511 temporarily in a double-write state, the access method selecting unit 504 selects an input / output request unit using an access method corresponding to each magnetic disk device. It is.

FIG. 6 is a flowchart showing how to select an access method.

When the access method is selected, first, after the actual input / output device is determined by the input / output distribution means 503, the magnetic disk device management table having the input / output request is referred to (step 601). The physical access method prepared in advance is selected according to the physical attribute of the input / output request (step 602), and the command and data buffer prepared by the input / output request source are converted according to the selected access method (step 603).

Thereafter, control is transferred to the physical access method based on the converted command and data buffer (step 604). That is, for the magnetic disk drive 509, the physical input / output access method 508 depending on the magnetic disk drive 509, and for the magnetic disk drive 511, the physical I / O access means 51 dependent on the magnetic disk drive 511.
By using 0, an input / output request is realized. Also,
The transfer of the file from the old magnetic disk device 509 to the new magnetic disk device 511 is executed by copying the file by the dual writing control unit 502.

Next, a method of executing an input / output request during execution of a file copy will be described. Copy information is managed by bitmap information 505 corresponding to the magnetic disk device 509 and bitmap information 507 corresponding to the magnetic disk device 511.
And conversion table 506 for converting between bitmaps
Is managed by a management table composed of The bitmap information 505 and 507 are stored in an area 51 indicating a non-copy.
2. An area 513 indicating that copying is in progress and an area 514 indicating that copying has been completed. Each area is assigned an address for each copy execution unit (track or cylinder) and manages the copy state. In FIG. 5, addresses 0 and 1 are copied.
, The addresses 2 and 3 are being copied, and the addresses 4 and 5 are not being copied.

When the magnetic disk devices 509 and 511 are set to the double write state, all the bitmap information 507 of the magnetic disk device 511 that is the copy destination is the bitmap information of the magnetic disk device 509 of the non-copy and copy source. It is written that all 505 have been copied. The copy process starts copying from the start address of the bitmap information 505 of the magnetic disk device 509 that is the copy source, and writes the copy source data to the magnetic disk device 511 corresponding to the address converted by the conversion table 506. And the bitmap information 507 is copied. At the end of copying, the bitmap information 507 of the copy destination magnetic disk device 511 is written as bitmap information corresponding to the copied area as copied.
As a result, it is possible to determine whether the data is not copied, copy is being made, or has been copied only by referring to the bitmap information.

FIG. 7 is a flow chart showing the double write control when an I / O request is issued to two magnetic disk devices that are currently being copied.

The dual writing control unit 502, which has received an input / output request from the user application, refers to the bitmap information 505 and 507 and determines whether the area of the input / output request destination has already been copied (step 701). If the copy has been completed, an I / O request is created and executed for both the copy destination and copy source magnetic disk devices (step 702).

At this time, if the input / output request is read, it may be executed only on one of the magnetic disk devices. If the input / output destination is executing a copy (step 704), the request is held in the dual writing control unit until the copy is completed (step 705), thereby guaranteeing data consistency. If the input / output destination has not been copied, the input / output request is executed only to the copy source magnetic disk device (step 706).

A procedure for transferring a file from the old magnetic disk device 114 to the new magnetic disk device 124 in the system configuration of FIG. 1 using the means described above will be described.

It is assumed that the magnetic disk device 124 is a magnetic disk device that was not defined when the system was created.
The addition of the device by software is executed at the timing of a command to input an operator after the work of adding hardware is completed. The magnetic disk device 124 is the input / output processor 12
0, a channel 121, a magnetic disk control device 122, and a magnetic disk drive device 123.

The dynamic configuration control means 103 newly generates a magnetic disk device management table corresponding to the magnetic disk device 124 and prepares for receiving an input / output request.

Next, the magnetic disk device 114 and the magnetic disk device 124 are defined as a magnetic disk device in a double write state by using the double write control means 104. The definition of the dual writing state is written in the system configuration definition file 106 in case the system goes down or restarts during the migration. As described above, input and output to and from the magnetic disk device 114 during this time may be in progress.

Further, the copy of the file is executed by using the double writing control means 104. After the copy is completed, the magnetic disk drive 114 is disconnected from the system and physically removed.

Thus, the user application 100 does not need to stop the input / output request even during the transfer of the file from the magnetic disk device 114 to the magnetic disk device 124.

(Embodiment 2) In the case of maintenance work such as replacing only the medium of the magnetic disk device, a magnetic disk device having the same physical attribute is temporarily set as a spare magnetic disk device for file migration. By transferring files to the spare magnetic disk unit and continuing the work, performing maintenance work on the original magnetic disk unit, and again using the same replacement means to transfer files from the spare magnetic disk unit to the original magnetic disk unit The maintenance of the magnetic disk drive can be performed.

(Embodiment 3) As is clear from FIG. 1, by providing the main control means for file migration on software, peripheral devices other than the CPU, input / output processors, channels, magnetic disk control devices The exchange / transition of the disk drive can be executed by the same procedure.

(Embodiment 4) Next, a case where the present invention is applied to a multiplex writing magnetic disk device will be described.

FIG. 8 shows a procedure for transferring files between dual-write magnetic disks as the most common example of a multiple-write magnetic disk device.

In the procedure, a new magnetic disk device 810 is added to the magnetic disk devices 800 and 801 already in the double writing state, and the file is copied. Therefore, during copy execution and after completion, a triple write state is temporarily established.

Next, in the procedure, the magnetic disk drive 8
01, and the dual writing state is set in the magnetic disk devices 800 and 810.

In the procedure, in the same manner as the procedure, a new magnetic disk device 811 is added to the magnetic disk devices 800 and 810 in the double writing state, and a file copy is performed.
Execute

In the procedure, the magnetic disk device 800 is separated, and the magnetic disk device 810 and the magnetic disk device 811 are separated.
Use to set the double writing status.

The procedure shows that the magnetic disk device 800 is disconnected from the system, and a new double writing state is set by the magnetic disk devices 810 and 811.

In this way, by applying the present invention to a multiplex-write magnetic disk drive, the multiplex-write state is always maintained during the transition of the magnetic disk drive, and
It is possible to improve the system reliability and availability without stopping the execution of the running application.

(Embodiment 5) Next, a case where the present invention is applied when a new magnetic disk device is installed in a remote place will be described.

FIG. 9 shows an embodiment in which a new center is provided at a remote location and a magnetic disk drive is transferred to the new center.

Even when such a system is configured, if the magnetic disk device and the magnetic disk control device which support the input / output access means of the operating system 902 at a remote location are supported, Based on the procedure described above, the operation at the main center is not changed from the conventional one, and the transfer from the magnetic disk device 903 as the file transfer source of the main center 900 to the magnetic disk device 904 as the file transfer destination of the sub center 901 is possible. Becomes

Further, in the case of a transfer to a remote location, there is a possibility that the transfer will be interrupted during the transfer due to a failure in the communication line connecting the main center and the sub center. In such a case, the migration destination file of the sub center 901 is closed and the copy processing is interrupted. If there is a vacant magnetic disk device accessible to the sub-center, the migration process can be continued by redefining the new device and the migration source device as dual writing devices. As described above, according to the present invention, there is an effect that the file transfer of the magnetic disk device can be performed without stopping the system during the execution of the user application. In addition, the attribute /
There is an effect that migration is possible even if the form is different.

[0062]

As described above, according to the present invention,
It is possible to perform maintenance / migration work of the external storage device without stopping the application program of the business. Further, even if the attributes and forms of the external storage device to be migrated are different, migration is possible.

[Brief description of the drawings]

FIG. 1 is a system configuration diagram showing an embodiment of the present invention.

FIG. 2 is a diagram illustrating addition of a device management table accompanying addition of a real device.

FIG. 3 is a flowchart when a device management table is added.

FIG. 4 is a flowchart when deleting a device management table.

FIG. 5 is a configuration diagram that absorbs differences in physical attributes and executes file migration.

FIG. 6 is a flowchart when an access method is selected by the dual writing control unit.

FIG. 7 is a flowchart of a double writing control for a device that is executing a copy.

FIG. 8 is a diagram showing a procedure for explaining an embodiment of file migration between multiple writing disks.

FIG. 9 is a diagram for explaining an embodiment of transferring a file to a remote place.

[Explanation of symbols]

100 user application, 101 operating system, 102 logical input / output access means, 103 dynamic configuration control means, 104 dual writing control means, 105 physical input / output Access means, 106
... System configuration definition file, 110, 120 ... Input / output processor, 111, 121 ... Channel, 112,
122: Disk control device, 113, 123: Disk drive device, 114, 124: Magnetic disk device
200: Device management table for magnetic disk device 210, 201: Device management table for magnetic disk device 211
Table 202, a device management table for the magnetic disk device 212, 210, 211, 212, a magnetic disk device, 220, a device identifier field, 221 a device address field 222, a physical attribute field, and 5
00: user application, 501: logical input / output access means, 502: double writing control means, 503
...... Input / output distribution means, 504 ... access method selection means, 505, 507 ... bitmap information, 506 ...
Conversion table, 508, 510 ... physical input / output access means, 509, 511 ... magnetic disk drive, 512 ...
… Uncopied area, 513 …… Copied medium area, 514…
Copied area, 800 ... Magnetic disk drive A, 80
1 ... Magnetic disk device A ', 810 ... Magnetic disk device B, 811 ... Magnetic disk device B', 900 ...
Main center, 901... Sub-center, 902... Operating system, 903... Magnetic disk device storing a source file, 904... Magnetic disk device storing a destination file.

──────────────────────────────────────────────────の Continuing on the front page (72) Inventor Hiroshi Suzuki 5030 Totsuka-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Inside Software Development Division, Hitachi, Ltd. (72) Inventor Hiroshi Yokota 5030 Totsuka-cho, Totsuka-ku, Yokohama-shi, Kanagawa Stock (56) References JP-A-4-51314 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G06F 3/06 G06F 13/10

Claims (5)

(57) [Claims] 1. A data storage method comprising: a computer and a plurality of storage devices, wherein data is written from the computer to the storage device during data copying between the storage devices, wherein the computer stores The state of the copy source and copy destination storage devices is managed for each predetermined area, and when an I / O request for writing data to one of the storage apparatuses is received, the data of the data is stored based on the state of the area .
A data storage method for processing the input / output request in the storage device that is a copy source and a copy destination . 2. The data storage method according to claim 1, wherein
The status of the storage device indicates at least whether or not the copy has been completed, and the computer is a copy source and a copy destination when the target area of the input / output request has been copied.
Data storage method for processing the output request to the storage device. 3. The data storage method according to claim 1, wherein
I / O for writing data to one of the storage devices
The request is the entry / exit requested by the application program
Data storage method that is a force request. (4)It has a computer and a plurality of storage devices.
A data storage method for writing data to a storage device, The computer is based on an input / output request for one storage device.
Of each storage device registered in the management table.
The management table according to the access method prepared
Process the I / O request to a plurality of storage devices registered in
The data storage method to be performed. 5. The data storage method according to claim 4, wherein
I / O requests to the one storage device are
Which is an input / output request requested from the application program
Storage method.