JPH0527921A - File control system for magnetic disk device - Google Patents

Abstract

PURPOSE:To improve the reliability of a system, to recover a defective part by a parallel operation with a user job and to execute active maintenance. CONSTITUTION:Concerning the file managing system of a form to allocate an empty area (Y file) 11 managed by an operating system(OS) to a temporary area (work area) 10 of a magnetic disk device used by a job while activating the job under the control of the OS, when a defective part (*) is detected at a certain part in the allocated temporary area 10, a surface inspecting/ repairing tool is activated as an abnormal end without releasing the temporary area 10. At such a time, the entire magnetic disk device is not occupied but the file (X file) 10 including the defective part is repaired by the parallel operation with the user job.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of controlling a file containing a defective medium portion in a magnetic disk device.

File storage devices (DASD), such as magnetic disk devices, have been brought along with the recent online computer systems.
In addition, even if a failure occurs, it is required to manage files that do not affect the following jobs.

Further, in order to improve the service to the user, it is required that active maintenance can be performed such that the defective portion is repaired in parallel with the job, and that the repair tool is automatically started without the intervention of the operator. .

[0004]

2. Description of the Related Art FIG. 2 is a diagram for explaining a file control method in a conventional magnetic disk device.

Conventionally, operating systems (hereinafter,
When a job is executed under (OS), the temporary area (work area) (X file) 10 used by the job is managed by the system. Free space group (Y file) 11 of the magnetic disk unit 1 Assign from

At this time, if a defective portion that cannot be written / read is detected in the allocated temporary area 10, recovery processing of hardware and software is performed, and if the recovery processing fails, the record address of the defective portion is recorded in the logging file. Abnormally terminates the job and releases the temporary area.

Normally, when the job is completed, the temporary area
Since the temporary area (work area) 10 is released and placed under the control of the OS so that the (work area) 10 can be used by other jobs, there is a defect in the temporary area (work area) 10. Even when the location is detected, the temporary area 10 is released.

The details of the operation under the operating system (OS) will be described below in the order of processing. (1) A job is started under the operating system (OS).

(2) A file management system under the OS allocates a temporary area (work area: X file, which may be abbreviated below) 10 used in a job from a free area (Y file) 11 of a disk device. (See Fig. 2) (3) Write / read to the allocated temporary area 10 to execute the job.

(4) If a read error (data check) occurs at a location in the allocated temporary area 10,
Recovery processing is performed by the OS or the control device, and if it is an uncorrectable read error, the defective cylinder, track, and record position are recorded as error logging information in a logging file (not shown). (See B in Fig. 2) (5) The OS records the error logging information in the logging file and then releases the temporary area 10 with the defective part (*),
It should be an area that can be reallocated as a free area in the system. Further, the OS abnormally terminates the job being executed when it detects a hard error (uncorrectable read error). (See B and C in Figure 2)

[0011]

From the above, the following problems occur. When the job is executed after the process of (5) above, the defective part described above is again allocated as the temporary area 10, so the following jobs may end abnormally (that is, many errors occur). There was a problem.

Further, based on the information logged in the process of (4) above, the uncorrectable defective portion (cylinder, track, record position) is recognized, and then the surface inspection / repair tool of the magnetic disk device is used. The entire device is occupied, and the defective portion is skipped or the alternate track processing is performed to repair the defective portion.

That is, in the conventional method, since the file including the defective portion is opened, it is not possible to recognize in which file the defective portion exists. Therefore, the system occupies the entire disk device. However, it is necessary to repair the defective portion using the surface inspection / repair tool of the magnetic disk device.

Further, in the conventional method, the file including the defective portion is opened and the execution of the subsequent job is started.
Moreover, since the file including the defective portion cannot be recognized and the system needs to manage the entire disk device as described above, for example, the operator may find that abnormal terminations frequently occur. At the time of detection, after temporarily stopping the execution of the job, the correction parameters are manually extracted from the above logging information, input to the system, and the above-mentioned surface inspection / repair tool of the magnetic disk device is started. There was a need. That is, since the repair tool cannot be automatically started, problems such as erroneous input and time-consuming input that occur when the correction parameters are manually input are caused.

In view of the above-mentioned conventional drawbacks, the present invention prevents frequent occurrence of abnormal termination when a succeeding job is executed when it is detected that there is a defective portion on the medium during execution of the job in the magnetic disk device. In addition, it is an object of the present invention to provide a file control method for repairing a medium, which can improve the reliability of the system and improve the service to the user.

[0016]

FIG. 1 is a diagram showing an embodiment of the present invention. The above problems can be solved by the file control method in the magnetic disk device configured as follows.

(1) A job is started under the operating system (OS), and a free space managed by the operating system (OS) is stored in a temporary area (work area) 10 of the magnetic disk device 1 used by the job. In a file management method of allocating an area, when a defective portion (*) is detected at a certain place in the allocated temporary area (work area) 10, the temporary area (work area) 10 is released. Instead, it is configured to terminate abnormally.

(2) A job is started under the operating system (OS), and a temporary area (work area) 10 of the magnetic disk device 1 used by the job has a free space managed by the operating system (OS). In the file management method of allocating an area, when a defective portion (*) is detected at a certain position in the allocated temporary area (work area) 10, the temporary area (work area) 10 is released. By holding the magnetic disk device 1 without using the magnetic disk device 1, only the temporary area (work area) 10 is restored without occupying the entire magnetic disk device 1.

(3) A job is started under the operating system (OS), and a free space managed by the operating system (OS) is stored in a temporary area (work area) 10 of the magnetic disk device 1 used by the job. In the file management method of allocating an area, when a defective part (*) is detected at a location of the allocated temporary area (work area) 10, the temporary area (work area) 10 is not released. Retain and configure to launch the repair tool.

[0020]

In other words, in the present invention, the operating system
(OS) has the following processing. In the temporary area (work area: X file), for example,
When an uncorrectable read error (data check) is detected, the job terminates abnormally, but the status is occupied without releasing this temporary area (control information data control block (DCB), data space control block). (DSCB) is also in the exclusive state}.

As a result, since the file including the defective portion is not assigned to the succeeding job, it is possible to prevent the abnormal termination that frequently occurs in the conventional method. The OS performs the above-mentioned process (4), that is, when a read error (data check) occurs at a portion of the allocated temporary area 10, the OS or the control device performs a recovery process to make an uncorrectable read error. If so, the process records the defective cylinder, track, and record position as error logging information in a logging file (not shown) .At the same time, save the error logging information in the system storage area as well. deep.

Further, the system occupies the temporary area including the defective portion, and the surface of the medium is inspected for the defective portion based on the error logging information saved as described above.
Run the maintenance tool to be repaired during system operation.

As a result, the operating system (OS) can recognize the file including the defective portion even for the disk device having the defective portion. While the file is started up and the file is repaired, the parallel operation of other user jobs, specifically, the time-sharing process of the repair process and the succeeding job can be performed, which enables active maintenance.

Therefore, in the file control method according to the present invention, when a non-recoverable defective portion is detected in the temporary area when the job is executed under the operating system (OS), the temporary area is not released, and the subsequent processing is performed. Subsequent jobs can be executed without adversely affecting the jobs, and the system reliability can be improved. Further, since the system does not release the temporary area including this defective portion, the repair tool can be started while executing the subsequent job, and the repair tool can use the error logging information saved in advance as a basis. Thus, the repair parameter can be extracted, the repair process can be executed, and the effect of active maintenance can be obtained.

[0025]

Embodiments of the present invention will be described in detail below with reference to the drawings. FIG. 1 described above is a diagram showing an embodiment of the present invention, (a) schematically shows an operation flow, (b) shows a format example of file management information (DSCB control information), C) shows a format example of device management (UCB control information).

In the present invention, a job is started under the control of the operating system (OS), and the temporary area (work area) 10 of the magnetic disk device 1 used for the job is
In the file management method of allocating a free area managed by the operating system (OS), when a defective portion is detected at a portion of the allocated temporary area 10, the temporary area 10 is not released. , Start the surface inspection / repair tool as abnormal termination. Further, at this time, the means for controlling the file including the defective portion so as to be repaired by the parallel operation with the user job without occupying the entire magnetic disk device 1 is a means necessary for implementing the present invention. Is. The same reference numerals indicate the same objects throughout the drawings.

The file control method in the magnetic disk device according to the present invention will be described below with reference to FIG. a. Jobs are executed under the operating system (OS), and free space (Y file) 11 to space (X file)
10 will be assigned. At this time, the operating system (OS) sends the control information (D
SCB) {Refer to Fig. 1 (b)} and create this X file
10 is the UC that tells which device and in what position
It is managed by B control information (see Fig. 1 (c)).

B. During execution of the above-mentioned job, for example, an uncorrectable read error (data check) occurs {there is a defective portion. The part indicated by * in the X file 10 of FIG. 1}. The sense information (cylinder value and track value) when this defective portion is detected is saved in the system storage area.

C. The file management under the operating system (OS) does not release the temporary area (X file) 10 and abnormally terminates the job in the occupied state. DCB (data control block) and DSCB (data space control block) of the job that ended abnormally
The control information such as is monopolized by the system.

D. Since the file (X file) 10 including the defective portion is exclusively used by the system, the X file 10 including the defective portion described above must be excluded in the subsequent user job even if this magnetic disk device 1 is used. Since it is allocated from the free space (Y file) 11, abend can be prevented.

E. X file exclusively owned by c. Above
Cylinder value and track value (track position) of sense information saved for 10 defective parts and control information (DSCB, UCB, etc.) of operating system (OS)
Obtain the volume serial number, device address, and device name from the media, and use these as parameters to inspect and repair the medium surface (skip control processing, alternate track allocation)
Have the system call the tool.

In the case of the present invention, as described above, since the system holds the file (X file) 10 including the defective portion, the normal job managed by the system can be performed without affecting the subsequent jobs. Since the file (Y file) 11 is allocated and the job execution instruction can be given, the repair tool can be automatically started and executed by the execution of the job and the parallel operation, specifically, the time-sharing operation.

In this way, the track can be used normally by removing the defective portion of the X file 10 without affecting the succeeding job. Further, as described above, active maintenance is realized by allowing a user job to operate in parallel with a device having a defective portion during surface inspection. As an example, jobs using the Y file 11 indicated by d can operate simultaneously.

F. After the surface inspection is completed and the defective portion is removed, the temporary area (X file) 10 occupied by the operating system (OS) is opened and made available as a free area.

As described above, according to the present invention, the job is started under the operating system (OS), and the temporary area (work area) 10 of the magnetic disk device 1 used for the job is used.
In the file management method of allocating a free area managed by the operating system (OS), the temporary area 10 is released when a defect is detected at a location of the allocated temporary area 10. Without that, the surface inspection / repair tool is started as an abnormal termination. Further, at this time, the magnetic disk device 1 is not entirely occupied, but a file including the defective portion is repaired by a parallel operation with a user job.

[0036]

As described above in detail, the file control method in the magnetic disk device of the present invention does not open the temporary area when a non-recoverable defective portion is detected in the temporary area during execution of a job. And the system holds
Since the repair tool is started and the temporary area including the defective portion is repaired in parallel with the execution of the subsequent job, it does not adversely affect the subsequent job. The nature is enhanced. Also, since the system occupies only the partial area including the defective portion, the defective portion can be repaired without occupying the entire disk device medium, and active maintenance that repairs the defective portion in parallel with the user job is possible. Becomes In addition, in the conventional method, after stopping the system once, based on the logged information,
Although a surface inspection / repair tool was used after a person investigated an uncorrectable defect part, the present invention holds a file including the defect part and executes a subsequent job when the defect part is detected. It is also possible to activate the surface inspection / repair tool (automatic activation).

[Brief description of drawings]

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

FIG. 2 is a diagram illustrating a file control method in a conventional magnetic disk device.

[Explanation of symbols]

1 Magnetic disk unit (DASD) 10 Temporary area (work area) or X file 11 Free space group or Y file

Claims (3)

[Claims] 1. A magnetic disk drive used by a job when a job is started under the control of an operating system (OS).
In the file management method of allocating the free space managed by the operating system (OS) to the temporary area (work area) (10) of (1), the allocated temporary area (work area) (10) A file control method in a magnetic disk device, wherein when a defective portion (*) is detected at a certain location, the temporary area (work area) (10) is not terminated and abnormal termination is performed. 2. A magnetic disk drive used by a job when a job is started under the operating system (OS).
In the file management method of allocating the free space managed by the operating system (OS) to the temporary area (work area) (10) of (1), the allocated temporary area (work area) (10) , When a defective portion (*) is detected at a certain location, the temporary area (work area) (10) is held without being opened, so that the entire magnetic disk device (1) is not occupied, A file control method in a magnetic disk device, wherein only the temporary area (work area) (10) is restored. 3. A magnetic disk device used by the job when a job is started under the operating system (OS).
In the file management method of allocating the free space managed by the operating system (OS) to the temporary area (work area) (10) of (1), the allocated temporary area (work area) (10) When a defective portion (*) is detected at a certain location, the temporary area (work area) (10) is held without being opened, and the repair tool is started up. method.