JPS6162941A - Inspection system of information processor - Google Patents

Abstract

PURPOSE:To prevent a data change and to take over the state at a stop by actuating a patrol execution task of each processor at constant intervals of time, and saving information required for the next start-up when an error is detected and stopping the execution task. CONSTITUTION:Patrol execution tasks 32 and 33 for testing processors 41 and 42 and an operating system 31 which controls the tasks 32 and 33 are stored in a main storage device 30, and the processors 41 and 42 are connected to the device 30. Then the system 31 actuates the tasks 32 and 33 and an instruction to be tested, etc., are transferred to a message area and executed; when the execution result coincides with an expected value, the same process is repeated a specific time later. When not, the expected value is transferred to a message output area and an error message is outputted to an output device to write information required for the restarting of the system on a disk and also stop the system.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an inspection method for an information processing device, and more particularly to a Luz check method for inspecting an information processing device under control of an operating system.

[Prior art]

Conventionally, failures in this type of information processing equipment have been detected by check circuits within the equipment, and when failures that cannot be checked by the check circuits occur, they are detected in the form of logical contradictions in software such as the operating system. Ta. For this reason, the latter faults that are expected to be detected in the form of logical contradictions in software may not be detected even in software because software is not created for the purpose of detecting hardware faults.

In this case, the data would be garbled, and since the failure was detected a long time after the failure occurred, there was a fatal problem in that data recovery would be difficult.

[Purpose of the invention]

The purpose of the present invention is to prevent data from becoming garbled in the event of a failure that cannot be detected by software, and at the same time to enable the operating system to continue operating in the same state as when it was stopped. There are two ways to provide an inspection method.

[Structure of the invention]

The information processing device inspection method of the present invention provides patrol execution tasks for each processing device for testing the processing devices connected to the main storage device, and an operating system that controls these patrol execution tasks in the main storage device. The operating system activates all of these patrol execution tasks at regular intervals, and the activated patrol execution task performs inspection of the corresponding processing device. When the patrol execution task detects an error, the operating system The operating system immediately executes an emergency stop process to save information necessary for the next startup.
It is characterized by stopping the execution of the operating system itself.

〔Example〕

Embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an information processing apparatus to which the information processing apparatus inspection method of the present invention is applied. On the main storage device 60, an operating system (hereinafter referred to as O8) 61
and a processing device (hereinafter referred to as CP) connected to the main storage device 6o.
A patrol execution task 32.33 for testing 41.42 (referred to as U) is stored.

Next, the operation of patrol execution in the information processing apparatus of this embodiment will be explained.

Since the logic of the patrol execution tasks 32 and 36 is the same, the patrol execution task 62 will be described below.

FIG. 2 is a flowchart showing the operation of the patrol execution task 32, which is an inspection program for the CPU 41. First, the patrol execution task 32 is activated from process 11 (:). Control is transferred to the next process 12, and information about the CPU number to be tested and the instruction to be tested is transferred to the message area.

In process 16, the instruction to be tested is executed. Processing 14
Now, compare the above execution results with the expected values. If the results match in judgment 15, the process moves to judgment 20. Judgment = 3- In decision 20, if it is not the last test, the next test is repeatedly executed from process 12. If it is the last test, control of the patrol execution task 62 moves from judgment 20 to process 21, and 1
Set the timer.

After one hour has elapsed, control returns from decision 22 to process 11, and the test is performed again.

If the execution result and the expected value do not match in judgment 15, control of the program moves to process 16;
The execution results in step 14 and the expected values used in step 14 are transferred to the message sender. Thereafter, in process 17, the error message is edited and output to a message output device (not shown). When control of the program is transferred to process 18, a stop request is sent to 0861, and 0831 sends information necessary to restart the system as an emergency stop process, such as the fact that a CPU has failed, and the failed CPU number (not shown). Write to disk. Thereafter, control moves to process 19, and the OS 31 immediately performs system stop processing.

This prevents the occurrence of data corruption due to malfunction of 0PU41.42, and also makes it possible to operate 0831 by taking over the information at the time of failure.

Although the above explanation was about the patrol execution task 62 which is the inspection program for the CPU 41, it can be easily inferred that the present invention is also applicable to the patrol actual information processing device which is the inspection program for the CPU 42.

〔Effect of the invention〕

As explained above, the present invention executes a program for checking the normality of processing devices at regular intervals equal to the number of processing devices, and when an error is detected, requests the operating system to stop execution. The operating system immediately executes an emergency stop process to save the information needed for the next startup, and then stops the operating system itself. At the same time, it has the effect of making it possible to take over the information when a failure occurs and run the operating system.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of an information processing apparatus to which the information processing apparatus inspection method of the present invention is applied, and FIG. 2 is a flowchart showing an outline of patrol execution for carrying out inspection in the information processing apparatus of FIG. 30...Main storage device, 31...Operating system, 62.66...
-Patrol execution task, 41.42...processing device. Patent applicant: NEC Corporation Representative: Susumu Uchihara, patent attorney
': Figure 1

Claims (1)

[Claims] Patrol execution tasks for each processing device for testing the processing devices connected to the main storage device and an operating system that controls these patrol execution tasks are stored in the main storage device, and the operating system executes these tasks at regular intervals. Start all patrol execution tasks,
The activated patrol execution task checks the corresponding processing device, and when the patrol execution task detects an error, it issues a request to the operating system to stop execution, which causes the operating system to immediately restart the next startup. An inspection method for an information processing device, characterized in that the operating system itself stops running after executing an emergency stop process to save information.