JPH02141834A - Trouble recovery system for processor - Google Patents

Abstract

PURPOSE:To quickly perform the trouble recovery without intervention of a maintenance man by automatically performing IMPL again before reporting the occurrence of trouble from a service processor to a display device, and recovering the trouble without affecting in the operation of the other processor. CONSTITUTION:A service processor 1 is provided with a mechanism 2 which analyzes trouble of disk control processors 10 and 11 and disk devices 12 and 13. When trouble occurs in the service processor 1, a watchdog timer 4 expires because a clear signal 5 cannot be sent to this timer, and a time-out signal 6 is sent to a trouble recovery mechanism 3. This mechanism 3 automatically loads the program of the service processor 1 again to raise the system (re- IMPL). Then, trouble is automatically recovered without affecting in the other processor. Thus, the trouble recovery is performed without intervention of the maintenance man.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a failure recovery method for a service processor in an electronic computer having a multiprocessor configuration.

[Conventional technology]

As described in Japanese Patent Application Laid-Open No. 63-59637, conventional devices monitor failures in one processor using a watchdog timer, and if a processor failure occurs and the watchdog timer is not reset for a certain period of time, the timer counter is reset. If an overflow occurred, the number of resets up to that point would be displayed on the display, and the processor configuration was also a single processor system.

[Problem to be solved by the invention]

The above-mentioned conventional technology does not take into account recovery from a fault, and after a fault is detected, the operator has to perform fault recovery depending on the cause of the fault, and there is a problem in that an operation for recovery is required.

SUMMARY OF THE INVENTION An object of the present invention is to provide a function to automatically perform fault recovery after a fault is detected, and to perform fault recovery without affecting other processors.

[Means to solve the problem]

In order to achieve the above purpose, the service processor that detects a fault automatically performs a re-IMPL (system startup by reloading the program) before reporting the fault to the display device, thereby affecting the operation of other processors. The system is designed to recover from a service processor failure without any trouble.

[Effect]

When a failure occurs in the service processor, the service processor's failure recovery mechanism automatically re-IMPLs the service processor (starts up the system by reloading the program).As a result, failure recovery is performed only by the service processor, and other processors are without affecting operation.

No malfunctions.

〔Example〕

An embodiment of the present invention will be described below with reference to FIG.

The service processor 1 is a disk control processor 10
．． The disk control processor 10.11, which has a mechanism 2 for analyzing failures in the disk device 12.11 and the disk device 12.13, receives a control signal 19.20 from the central processing unit 14 and performs read/write operations on the disk device 12.13. signal 17,1
8 and to the central processing unit 14 via control signals 18 and 19. The service processor 1 sends a control signal 15.22 to the disk control processor 10.11 to provide disk failure information 16.21.
The disk failure analysis mechanism 2 receives the data and analyzes the cause of the failure. The display device 8 sends a control signal 18 to the service processor 1.
is sent, the fault information 17 is obtained, and this is displayed. The watchdog timer 4 is a timer that monitors failures in the service processor 1, and generates a timeout signal 6 if it does not receive a clear signal 5 for a certain period of time. While the service processor 1 is operating normally, the failure recovery mechanism 3 sends a clear signal to the watchdog timer 4 at regular intervals, so the timeout signal 6 is not generated.

When a failure occurs in the service processor 1, the clear signal 5 cannot be sent to the watchdog timer 4, so the watchdog timer times out and sends a timeout signal 6 to the failure recovery mechanism 3.

The failure recovery mechanism 3 automatically re-IMPLs the service processor 1. IMPL (initial microprogram load) loads the program of the service processor 1 from the external storage device 24 into the memory 7. Program sends reset signal 9 to service processor 1
Turning on will clear the fault.

If the cause of the failure is not resolved, IMPL is applied multiple times to recover from the failure. Even while the service processor 1 is recovering from a failure using IMPL, the central processing unit 14 can exchange control signals 19.20 with the disk control processor 10.11, and the disk control processor 10.11 can read/write the disk devices 13 and 14. Light signal 17.18
can be exchanged. During failure recovery, the disk I/O operations of the disk control processor 10.11 are not affected.

According to this embodiment, since a fault in the service processor is automatically recovered, there is an advantage that the fault can be recovered quickly without the intervention of maintenance personnel.

〔Effect of the invention〕

According to the present invention, when a processor failure occurs, it can be automatically recovered without affecting other processors.
Failure recovery can be performed without the intervention of maintenance personnel.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of one embodiment of the present invention. 1... Service processor. 2...Disk failure analysis mechanism. 3...Failure recovery mechanism. 4...Watchdog timer. 8...Display device.

Claims (1)

[Claims] 1. In an electronic computer consisting of multiple processors, a service processor, and a display device, after detecting a fault in the service processor, re-IMPL (system startup by reloading the program) only the service processor, without affecting the operation of other processors. A failure recovery method for a processor, characterized by providing a mechanism for recovering from a failure without causing any damage to the processor.