JPH04213123A - Preventive maintenance system for fault of electronic computer - Google Patents

Abstract

PURPOSE:To allow a computer system to have a self-diagnostic function, history managing function, and alarm function in order to operate a preventive maintenance by early discovering the fault of an outside storage device by an on-line operation. CONSTITUTION:The self-diagnosis of the outside storage device is automatically operated in a fixed cycle in parallel with the normal operation of an electronic computer in order to cope with the preventive maintenance of the fault of the outside storage device of the electronic computer. The diagnosed result is held and updated, an error history is generated, and the number of times of an error is monitored. An alarm is given to an operator in the case of more than the fixed number of times of the error, and the preventive maintenance is operated in order to prevent the serious fault.

Description

[Detailed description of the invention]

0001

[Field of Industrial Application] This invention relates to computer systems, and in particular, computer failure prevention that performs periodic online self-diagnosis of external storage devices before a failure occurs in the external storage device and recovery becomes impossible. It concerns the maintenance method.

0002

2. Description of the Related Art If a storage device in an electronic computer malfunctions due to its usage, it often leads to immediate system failure, and simply having a low failure rate does not provide peace of mind. Most failures occur when a program previously stored in a storage device is loaded when necessary, or the computer cannot read the program correctly or the contents loaded are incorrect, and in this case, the computer does not operate normally. Therefore, preventing these failures is a factor in increasing the reliability of the system, but conventional methods cannot be said to be appropriate.

The first method, which is already common, is a method in which maintenance personnel run a diagnostic program for the device during periodic inspections, check for errors, and perform maintenance.

[0004] The second method, which is less common than the above-mentioned method, stores the history of the error when an error occurs in the area when the computer system operates and accesses the storage device. In this method, maintenance personnel check the history and carry out maintenance during periodic inspections.

[0005] The error referred to here refers to an error that can be retried in both software and hardware.The number of retries is determined for each, and if the error is recovered within that time, generally no error message is output.

[0006] In the second method, even if a retry is successful, the occurrence is regarded as an error and the occurrence is generally counted, and the number of retries is also counted.

[0007]Furthermore, the progress of storage device errors is generally manifested in an increase in the number of cases in which recovery is possible through retries and an increase in the number of retries.

[0008]

[Problems to be Solved by the Invention] Conventional failure preventive maintenance methods for electronic computers involve checking the external storage device by periodically running a diagnostic program offline by maintenance personnel, and checking for errors stored in the computer storage device. Error analysis must be performed by collecting information, which requires manpower and time for diagnosis, and if a failure occurs before the next periodic inspection, failure detection may be delayed and the severity of the failure may progress. There was also a consideration that it would end up happening.

[0009] This invention was made in order to solve the above-mentioned problems, and it operates the self-diagnosis function of the computer in parallel with the normal operation of the computer and during the period when the operation is stopped. The purpose is to obtain a fault preventive maintenance method for electronic computers that allows online fault monitoring and detection.

0010

[Means for Solving the Problems] A failure preventive maintenance method for a computer according to the present invention periodically performs self-diagnosis of an external storage device in parallel with access operations of the external storage device that the computer periodically performs. The self-diagnosis result is held and updated to generate an error history, and the system also has a function of monitoring the number of errors and outputting an alarm when an error occurs more than a certain number of times.

[0011]

[Operation] According to the present invention, a self-diagnosis for failures in the external storage device of a computer is automatically performed in parallel with the normal operation of the computer, an error history is generated based on the self-diagnosis results, and the error history is generated a certain number of times. This system integrates external storage device diagnosis, storage of diagnostic results, and error alarm management, such as notifying the operator of an alarm in response to the occurrence of an error, to perform preventive maintenance to prevent the computer from causing serious failure.

0012

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. In FIG. 1, 1 indicates the H/W configuration of the computer system. 11 is a main storage device, 12 is a central processing unit, 13
The input/output control device 14 is the target external storage device.

Further, 2 represents a S/W image within an external storage device within the computer system. 21
22 is an OS (operating system), and 22 is a utility program that implements a preventive maintenance method, and has three functions. Further, 23 is application software. The three functions of the utility program 22 are as follows.

(a) Online inspection function: A function that diagnoses all areas of the storage device at arbitrarily specified intervals without affecting the operation of the computer system, and when an error occurs, it is sequentially stored in the error history function. be done. (b) Error history function: A function that stores the error occurrence status in the function a, converts it to a specified format, updates it sequentially, and holds it as a history. Error information is stored in H/
This applies to all items restored by W retry and S/W retry. (c) Inspection management function: A function that refers to error information stored in the error history function at specified intervals without affecting the computer system, and generates an alarm according to the preventive maintenance standards and the computer system scale, for example. On a computer system, an alarm message is output when a certain number of errors due to H/W retry is exceeded, which is not considered an error.

Next, an actual preventive maintenance method will be explained. To explain the sequence of FIG. 2 as an example, first, sequence 1 is access of application software of an external storage device in a normal computer system. When accessing the external storage device, if the access ends normally, the computer system continues the operation in the next sequence. However, according to sequence 2, if an error occurs, the error history function starts collecting error information. During operation of the application software, the online inspection functions of sequences 3 to 5 (dotted lines) are sequentially accessing the external storage device at a cycle that does not affect the computer system. When the process ends normally (sequences 3 and 4), the operation continues, and when an error occurs (sequence 5), error information is stacked by the error history function in the same way as application software. At this time, the determination processing operation is executed at the same time as the operation of the inspection management function is collected by the error history function.

FIG. 3 shows an example of the determination processing of the inspection management function (taking a disk device as an example). When a collectable data check error that can be saved by retrying occurs, it is determined whether it has occurred 20 times, and if the number of times is within the standard, the determination process is completed as shown in the figure. Next, if the reference value shown in the figure is exceeded, an alarm program will be activated to alert the user, and other actions will be taken according to the agreement with the computer system user.

This preventive maintenance method is particularly effective for fixed head disk devices, but can also be applied to external storage devices such as optical (magnetic) disks, floppy disks, and magnetic tape devices.

[0018]

[Effects of the Invention] As described above, according to the present invention, a computer is provided with a function to perform self-diagnosis, error history, and failure alarm of an external storage device in parallel with normal access to the external storage device. This allows for online self-diagnosis, which saves human labor, and has the effect of allowing preventive maintenance to be carried out quickly and reliably.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a hardware configuration diagram of a computer system and a software image of an external storage device in a failure preventive maintenance method for an electronic computer according to an embodiment of the present invention.

FIG. 2 is a sequence explaining the operation of the method of this embodiment.

FIG. 3 is a flowchart showing an example of the determination processing of the inspection management function in the method of this embodiment.

[Explanation of symbols]

1 Computer system 2 Software image of external storage device 14
External storage device 21 Operating system 22 Utility program

Claims (1)

[Claims] Claim 1: A computer periodically performs self-diagnosis of the external storage device in parallel with the access operation of the external storage device that is periodically performed, maintains and updates the self-diagnosis results, generates an error history, and detects errors. A fault preventive maintenance method for electronic computers characterized by having a function of monitoring the number of errors and outputting an alarm when an error occurs more than a certain number of times.