JPH07311697A - Display fault system in computer system - Google Patents

Abstract

PURPOSE:To easily judge a fault part at a main function module level by setting a specific value in the nonvolatile register corresponding to a function module in which the fault has occurred and turning ON a light emission device corresponding to the function module. CONSTITUTION:Light emission devices 112 to 121 are provided nearby the respective function modules 102 to 111 of respective hardware modules (panel) 101 and are made to illuminate by setting a specific value in nonvolatile fault history registers. Then the specific value is set in the nonvolatile fault history register corresponding to the function module, whose fault is detected by a detecting means, among the function modules 102 to 111 to turn ON the light emission device corresponding to the function module which gets out of order among the light emission devices 112 to 121. Consequently, the fault part can easily be judged on the main function module level.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer system for displaying the occurrence of a failure.

[0002]

2. Description of the Related Art Conventionally, in order to notify that a failure has occurred in a computer system, an abnormality warning light is provided on an operation panel of the computer, or an abnormality warning light is provided in each of a plurality of hardware modules constituting the computer. Then, it has been performed to turn on the light when an abnormality is detected. By turning on these warning lights, the user can know the abnormality of the computer system and can confirm the hardware module in which the abnormality has occurred.

Here, the hardware module indicates various boards, buses, etc. connected to the system bus, and each of the boards is provided with a plurality of smaller function modules such as a processor and various controllers. It is assumed that

The above-mentioned warning light is generally turned on by a software operation or a hardware mechanism, and is turned on when an abnormality is detected. That is, once the power is turned off, the warning light is extinguished, and even if the power is turned on again, the warning light does not turn on until the abnormality is detected again.

Therefore, when investigating the cause again after turning off the system power, it is not possible to rely on the warning light, and the record of the operating state (system log) left inside the computer or when an abnormality is detected is detected. It is necessary to investigate based on the information in the hard copy of the warning message output on the console screen of or the result of executing the diagnostic program. It takes time and time to identify the failure point with the smallest replaceable unit (function module). It took a lot of work.

That is, in the conventional computer system, the failure display by the warning light when a failure occurs is only in the unit of the hardware module or in the unit of the computer system,
Moreover, there was no guarantee that the warning light would turn on again when the power was turned off.

In recent years, computers having excellent fault tolerance have been developed. Such a computer is called a fault tolerant computer. The fault-tolerant computer has a hardware / software configuration in which when a hardware failure occurs, the failure is detected, the cause of the failure is separated from the system, and the remaining hardware can continue the processing. Some fault tolerant computers have a fault detection function for each of the main function modules that make up the main hardware module.

A fault-tolerant computer generally has hardware redundancy, and when a failure occurs in a part of the hardware modules, the redundant hardware module takes over the processing of the failed module.

Therefore, the redundancy of the module may be lost after the failure occurs, and it is necessary to replace the failed module with a normal one as soon as possible after the failure.

Therefore, it is necessary to identify the components to be replaced accurately and in the smallest possible replacement unit. further,
It should be possible to easily verify that all failed modules have been removed from the system. In addition, it is important that the malfunctioning function module can be easily identified when the malfunctioning hardware module is brought back for investigation.

However, in the conventional computer system,
There is only the system log and hard copy information of the output message on the console screen when an error is detected, and there is an error warning light for each hardware module that lights up only when an error is detected. Since the warning light does not turn on, the system log and the output message on the console screen were used to check for abnormalities and determine which component to replace.

Therefore, it takes time and effort to find a defective module, and if there are a plurality of defective modules, it may be forgotten to replace them. In addition, if it is difficult to reproduce the failure in a single inspection after taking out the failed hardware module, it may be difficult to find the cause of the failure at the function module level, which is the smallest unit that can replace the failed part. It was

[0013]

As described above, in the conventional computer system, although the computer system or the hardware module is provided with the abnormality warning light, even when the failure is indicated by the warning light when a failure occurs, There is no guarantee that the warning light will be turned on again when the power is turned off once, and it has been difficult to find the cause at the function module level, which is the smallest unit that can replace the failure point.

The present invention has been made in consideration of the above circumstances, and it is an object of the present invention to provide a computer system capable of easily determining a failure location at a main function module level.

[0015]

To achieve the above object, the present invention has a plurality of hardware modules provided with a plurality of function modules, and at the function module level when a system failure occurs. In a computer system having a detection means for detecting a failure occurrence point, a nonvolatile register corresponding to each of the function modules of each of the hardware modules, and provided in the vicinity of each of the function modules of each of the hardware modules A first light emitting device which is turned on by setting a predetermined value in the non-volatile register, and a predetermined value is set in the non-volatile register corresponding to the function module in which the failure detected by the detection means has occurred, Funk with a failure Characterized by comprising a control means for lighting the first light emitting device corresponding to ® emission module.

A second light emitting device corresponding to each of the hardware modules is provided for each hardware module, and even one of the first light emitting devices corresponding to the function module constituting the hardware module is turned on. If so, the second light emitting device corresponding to the hardware module is turned on.

Further, if a third light emitting device corresponding to the entire system is provided and at least one of the second light emitting devices corresponding to the hardware module is turned on,
The third light emitting device is turned on.

[0018]

According to this structure, when a failure occurs, the system performs failure diagnosis, sets the nonvolatile register corresponding to the device causing the failure, and turns on the corresponding light emitting device.

As a result, the location of the failure can be easily determined at the level of the main function module at a glance, and even if the power is turned off once and then the power is turned on again, the light emitting device indicating the failure lights up. Even if the failure does not reoccur immediately after the power is turned on again, the hardware module / function module that caused the failure can be easily confirmed.

Further, even when the hardware module in which the failure has occurred is removed from the computer system and the cause of the failure is specified, such as a take-out test, the light emitting device is turned on even when the power is turned on for each hardware module. It is possible to judge, and it is easy to judge the defective chip module (function module).

The light emitting device corresponding to the hardware module is turned on as long as the function module causing the failure exists in the hardware module (for example, even if the failure does not reoccur), and the light emitting device is turned on in the computer system. As long as there is even one hardware module running, the light emitting device corresponding to the entire system lights up.

Therefore, the light emitting device corresponding to the system will not be turned off unless all the defective hardware modules are replaced. That is, it can be confirmed that all the defective hardware modules have been replaced by the light emitting devices corresponding to the entire computer system.

[0023]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows one of the boards (hardware modules) 101 constituting the computer system according to this embodiment. The computer system is a board 10 shown in FIG.
1 is stored in the computer main body in a plurality (see FIG. 2 described later).

In the computer system, when a failure occurs, the hardware module is replaced on a board basis as shown in FIG. In addition, the computer system is
It shall have means for finding a failure point and shall be able to find the failure point in units of the main function module of each board. Here, description will be made assuming that the means for detecting a failure is configured by hardware and the means for discovering a failure location is configured by software.

As shown in FIG. 1, a board 101 is mounted with a plurality of function modules 102 to 111. Function modules 102-111
When a failure occurs, it is possible to determine in which function module the problem has occurred by means of failure detection / failure location finding.

Further, each function module 102
LED111 as a light emitting device near
2 to 121 are provided. Although not shown in FIG. 1, nonvolatile failure history registers corresponding to the LEDs 112 to 121 are provided on the board 101 by the number of function modules 102 to 111 (= L).
EDs 112 to 121 are prepared). In addition,
Here, the “non-volatile register” means that the data in the register is saved even after the power supply is cut off (even when the power is turned off).

A board failure history warning lamp 122 is mounted on the front surface of the board 101. The board failure history warning light 122 includes the function modules 102 to 1
The LEDs 112 to 121 provided in 11 are turned on when any one of them is turned on.

On the rear surface of the board 101, the board 1
Bus connectors 123 to 125 for connecting 01 to the system bus of the computer main body are provided. Next, FIG. 2 shows the addresses of the failure history registers as seen from the software (means for finding failure points). The size of the register corresponding to each function module is 4 bytes. By setting the most significant bit (MSB) to "1", the LED (of the function module) corresponding to the failure history register lights up. It has become.

Here, it is assumed that the address of the failure history register of the board 101 starts from the hexadecimal address 1000 (address of the entire system). Figure 2
As shown in, the function module 102 (LE
It corresponds to each failure history register arranged from the address “1000” in order from D112). Each fault history register can be accessed by the processor at its respective address.

FIG. 3 shows an external view (view from the front) of the computer system main body 201. The computer main body 201 shown in FIG. 3 is composed of six boards 202 to 207. Each of the boards 202 to 207 has the same configuration as the board 101 shown in FIG.

The front of each board 202-207 is shown in FIG.
As described above, the board failure history warning lights 208 to 21
3 is provided. A system failure history warning light 215 is provided on the control panel 214 of the computer main body 201.

The system failure history warning light 215 is lit when any one of the boards 202 to 207 board failure history warning lights 208 to 213 is lit.

Next, the operation of this embodiment will be described with reference to the flow chart shown in FIG. FIG. 4 shows a procedure of processing for displaying a failure when a failure occurs in the computer system. In the computer system of this embodiment, it is assumed that the hardware detects the failure and the software finds the failure location.

First, when the processor of the computer system is notified of the failure occurrence by the hardware (failure detection means) (step S301), it executes the diagnostic program (failure location finding means) to diagnose the entire system. Start (step S302).

As a result, it is possible to determine the failure location (there may be multiple locations) at the function module level (step S303). The processor sets the most significant bit (MSB) of the failure history register corresponding to the function module determined to have a failure to "1" (step S304).

After that, the board including the failed function module is separated from the system by software (step S305), the normal processing is restarted on the remaining boards (step S306), and the processing when the failure occurs is ended (step S305). S307).

Thus, in the computer system of this embodiment,
When a failure actually occurs, the most significant bit of the failure history register corresponding to the function module determined as the failure is set.

As a result, the LED near the failed function module is turned on, the board failure history warning lamp of the board including the failed function module is turned on, and the system failure history warning lamp 215 is also turned on.

As in the conventional case, the system failure history warning lamp 2
By referring to 15, the user knows that a failure has occurred,
The offending board in which the failure occurred can be identified.
In addition, the present invention can identify a failed function module on a failed board.

When replacing a board in which a failure has occurred,
The system failure history warning light 215 goes out only after all the problem boards have been replaced, so that it can be confirmed that all the failed boards have been replaced.

Further, even if it is a temporary failure, a value indicating that a failure has occurred remains in the failure history register. Therefore, if the power is turned on again, the failure will not occur immediately even if the failure does not occur again. It is possible to determine the board and function module that caused the error.

For this reason, it is possible to determine the failed function module even with the failed board alone, and the efficiency of maintenance work is improved. In the present embodiment, steps S302 to S304 in FIG. 4 have been described as being performed by software, but there is no problem if these processes are performed by hardware.

In this way, each board 202-207 is
A non-volatile register corresponding to each function module is provided, and when a failure occurs, a value is set in the non-volatile register corresponding to the failed function module so that the light-emitting device provided near each function module can be installed. Can be turned on.

This makes it possible to easily determine the location of a failure in function module units at a glance.
Also, since the value is set in the non-volatile register,
Even if the power is turned off and then turned on again, the light-emitting device that indicates a failure lights up, and even if the failure does not recur immediately after the power is turned on again, it is easy to check the hardware module / function module that caused the failure. Become.

Further, even when the power is turned on in units of hardware modules, such as during a take-out test, the light emitting device is also turned on, so that the failure location of the module can be determined even in the test of the module alone, and the determination of the failed function module is easy. It will be.

[0046]

As described above in detail, according to the present invention, it becomes possible to easily determine the failure location at the main function module level.

[Brief description of drawings]

FIG. 1 is a diagram showing an external configuration of a board (hardware module) that constitutes a computer system according to an embodiment of the present invention.

FIG. 2 is a diagram showing an address map of a failure history register in this embodiment.

FIG. 3 is a diagram showing an external configuration (front) of a computer main body according to the present invention.

FIG. 4 is a flowchart showing a processing procedure when a failure occurs in this embodiment.

[Explanation of symbols]

101, 202-207 ... Board, 102-111 ... Function module, 112-121 ... LED, 2
01 ... Computer main body, 122, 208 to 213 ... Board failure history warning light, 214 ... Control panel, 215 ...
System failure history warning light.

Claims (3)

[Claims] 1. A computer system having a plurality of hardware modules provided with a plurality of function modules, and having a detection means for detecting a failure occurrence location at the function module level when a failure occurs in the system, Non-volatile registers respectively corresponding to the respective function modules of the respective hardware modules, and lighted by setting a predetermined value in the non-volatile registers provided in the vicinity of the respective function modules of the respective hardware modules. First to do
And a non-volatile register corresponding to the malfunctioning function module detected by the detecting means are set to a predetermined value, and the first light emitting device corresponding to the malfunctioning function module is turned on. A failure display method for a computer system, comprising: a control means. 2. A second light emitting device corresponding to each of the hardware modules is provided for each hardware module, and even one of the first light emitting devices corresponding to a function module constituting the hardware module is turned on. If it is, the failure display method of the computer system according to claim 1, wherein the second light emitting device corresponding to the hardware module is turned on. 3. A third light emitting device corresponding to the entire system is provided, and if any one of the second light emitting devices corresponding to the hardware module is turned on, the third light emitting device is turned on. The fault display method for a computer system according to claim 2, wherein