JPH0395634A - Restart control system for computer system - Google Patents

Abstract

PURPOSE:To omit the formation of a duplex system to quickly restart of an operation of a computer system at occurrence of a fault with no intervention of an operator and at the same time to analyze the fault later by providing a fault detection means and an action selecting/setting means to a system controller. CONSTITUTION:An action selecting/setting means 24 is provided to a system controller 2 to set the information which shows whether the state of a processor 6 should be previously recorded or not and whether the processor 6 should be restarted or not. When a fault detection means 21 of the controller 2 detects a fault of the processor 6, the contents of the means 24 are read to decide whether the state of the processor 6 should be recorded or not and whether the processor 6 should be restarted or not. When an instruction is received for recording the state of the processor 6, the information on the state of the processor 6 is recorded. Then the processor 6 is restarted if the restart of this processor is instructed. Thus it is possible to properly cope with the occurrence of a fault based on the contents of the means 24 with no duplex system required. Furthermore the fault can be analyzed later and therefore a system operating form is easily selected.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a computer system restart control method for automatically restarting a computer system when a failure occurs and its operation is stopped.

(Prior Art) When configuring a computer system, measures are taken to deal with failures such as hardware failures and program runaways. For example, if you want to prevent operations from stopping and processing being delayed, you can set up the same computer system as a backup computer system in addition to the computer system that normally operates, and immediately switch the operation to the backup computer system in the event of a failure. be. Furthermore, if the suspension of operation is not a serious problem, a single computer system without a backup computer system is used.

FIG. 2 shows a block diagram of a conventional single calculation system.

In the figure, a computer system 1 includes a system control device 2 that controls and manages the entire system, an operator conversation device 3 and a processing device 5 connected to the system control device 2.
It is composed of.

The system control device 2 is composed of a processor and the like that monitors whether the processing device 5 is activated and its operations are performed normally. The operator conversation device 3 is
This terminal device consists of a keyboard, a display, etc. operated by an operator, and is used to input various drawing commands to the system control device 2. The processing device 5 actually performs data calculations, and is composed of a processor 6, a main storage device 7, and a disk device 8. Note that each part is connected by a common bus 4. The processor 6 is
It controls and manages each part constituting the processing device 5. The main storage device 7 is composed of a RAM and the like that stores programs and data necessary for the operation of the processor 6. The disk device 8 stores data using a recording medium such as a hard disk, and is provided with a magnetic disk 9.

In the computer system I having the above structure, the system control device 2 stores in advance a boot program for starting the processor 6. Further, the magnetic disk 9 of the disk device 8 stores an initial program for loading a basic program that controls the operation of the system.

Now, when a boot command is manually input from the operator conversation device 3 to the system control device 2, the system control device 2 transfers the boot program to the main storage device 7 via the common bus 4. Furthermore, a startup command is issued to the processor 6. When the processor 6 receives this startup command, it executes the boot program stored in the main storage device 7, and then reads the initial program from the disk device 8 and stores it in the main storage device 7. Furthermore, the initial program is executed to finish starting the basic program that controls system operation and operation. The above procedure completes the startup of the computer system 1.

Now, in order to notify the system control device 2 that the operation is normal, the processor 6 issues a reset instruction for the monitoring timer provided in the system control device 2 at regular intervals. If the timer reset command is not issued and the monitoring timer performs a predetermined count, the system control device 2 determines that the program in the processor 6 has gone out of control, and forcibly stops the operation of the processor 6. Also,
When the processor 6 recognizes that a hardware failure has occurred in each part constituting the processing device 5, it stops its operation by itself. In this case, since the system control device 2 is not interrupted, the system control device 2 can recognize that the processor 6 has stopped.

Now, when the system control device 2 recognizes that the operation of the processor 6 has stopped, it makes the operator recognize the occurrence of the failure through the operator conversation device 3 or the like.

When the operator recognizes that a failure has occurred in the computer system 1 and the operation has stopped, the operator investigates the cause of the failure and then restarts it, that is, inputs a startup command using the operator conversation device 3. Become.

Now, in the case where the operation stoppage time is limited and the cause of the failure can be canceled by resetting the processor 6, etc., the operator inputs a reset command for the processor 6 from the conversation device 3. do. At this time, in order to analyze the cause of the failure later, in response to a reset command from the processor 6, the processor 6 itself primarily records the contents of internal registers (status information) indicating what state the processor 6 is currently in. The process of copying to the storage device 7 is performed. The expert will manually issue a command to execute a program to store the status information data copied to the main storage device 7 in the disk device 8 from the operator conversation device 3, and will then manually issue a command to restart the system. . Based on the information copied to this disk device 8, maintenance and
A service person will analyze the failure of the computer system l.

In addition, in the case of the computer system 1 that is operated unmanned without an operator or the like, a method for automatically restarting the computer system 1 is that the processor 6 detects a failure, and when the processor 6 is stopped, the main memory of the processor 6 itself is Referring to the fixed address of device 7,
Some systems automatically restart unconditionally, that is, perform operations based on an initial program or the like already stored in the main storage device 7.

(Problem to be solved by the invention) However, computer systems operated unmanned1
In this case, the system restarts unconditionally after its operation is stopped, resulting in the problem that status information at the time of failure cannot be saved. Further, in the computer system 1 which saves status information indicating the state of the processor 6 when the system is stopped and then starts up again, there is a problem in that an operator must always be present. Furthermore, there is a method of duplicating the computer system and continuing operation with another system while the faulty system is being investigated.
A problem has arisen in that it costs about twice as much to prepare two identical systems.

The present invention has been made with attention to the above points, and it is necessary to construct a redundant system.In the event of a failure, it is possible to quickly resume operation (restart) without operator intervention. The purpose of the present invention is to provide a computer system restart control method that can perform a failure analysis later.

(Means for Solving the Problems) A computer system restart control method of the present invention includes a processor and a system control device that controls failure detection and failure recovery of the processor, and the system control device The present invention is a combination of a failure detection means for detecting a failure and an operation selection setting means for setting an operation selection as to whether or not to record the status of the processor and restart the processor when the failure occurs.

(Operation) In the above method, information on whether to record the status of the processor and whether to restart it is set in advance in the operation selection setting means. When the failure detection means of the system control device detects a failure of the processor, it reads the contents of the operation selection setting means and determines whether to record the status of the processor and whether to restart the processor. . If an instruction to record the status of the processor has been given, status information indicating the status of the processor is recorded. Further, if a restart instruction has been given, the processor restart processing is executed. In this way, depending on the contents of the operation selection setting means, appropriate countermeasures can be taken when a failure occurs.

(Embodiment) FIG. 1 shows a block diagram of a computer system according to the present invention.

In the figure, a computer system 1 has a system control device 2 that controls and manages the entire system, an operator conversation device 3 connected to this system control device 2, and a common It is composed of a processing device 5 connected via a bus 4. Note that the same parts as in FIG. 2 are denoted by the same reference numerals, and redundant explanation will be omitted.

The system control device 2 is provided with a failure detection means 21, a status information copying means 22, a restart execution means 23, an operation selection setting means 24, and a ROM 25 storing a boot program. The fault detection means 21 detects a fault in the processing device 5, such as a program runaway in the processor 6 or a stoppage of the processor 6, and is composed of a monitoring timer (for example, a watchdog timer). The status information copying means 22 issues an instruction to record (copy to the disk device 8) the contents of the internal register 6a provided in the processor 6 when the fault detection means 21 detects a fault. Note that the internal register 6a is a normal register that the processor 6 has as a function. Restart execution means 2
3 outputs a restart instruction to the processor 6 when the failure detection means 21 detects a failure. The operation selection control means 24 is composed of registers and the like for setting whether or not the situation information copying means 22 executes a copy of the situation information and whether or not the restart execution means 23 executes a reboot. be.

FIG. 3 shows a configuration diagram of the operation selection setting means.

The figure shows the configuration of a 32-bit register corresponding to the data length (32 bits) of the system control device 2.

In the figure, the 30th bit is a restart instruction flag 24a for setting whether to execute a restart, and the 31st bit is a copy instruction flag 24b for setting whether to execute recording of status information. be.

For both flags, if the content is ``1'', the process is executed, and if the content is ``0°'', the process is not executed.

Now, returning to FIG. 1 again, the present invention will be explained.

In the computer system with the above configuration, it is assumed that the contents of both the restart instruction flag 24a and the copy instruction flag 24b of the operation selection setting means 24 are set to "1".Therefore, the status information copying means 22 A process of copying the contents of the internal register 6a of the processor 6 to the magnetic disk 9 of the disk device 8 via the main storage device 7 is executed.

Similarly, the restart execution means 23 restarts the processor 6.

Now, when a boot command is input from the operator conversation device 3 to the system control device 2, the system control device 2 transfers the boot program to the main storage device 7 via the common bus 4. Furthermore, a startup command is issued to the processor 6. When the processor 6 receives this startup command, it refers to the boot program stored in the main storage device 7, reads the initial program from the disk device 8, and stores it in the main storage device 7. Furthermore, the initial program is executed, and the startup of the basic program that controls system operation and operation is completed. The above procedure completes the startup of the computer system 1.

By the way, in order to notify the system control device 2 that the operation is normal, the processor 6 issues an instruction to reset the monitoring timer provided in the failure detection means 21 at regular intervals. If the timer reset command is not issued and, for example, the processor 6 goes out of control and no interrupt occurs,
The monitoring timer continues counting. If this count value exceeds a predetermined value, the system control device 2 determines that a failure has occurred in the processor 6, and forcibly stops the operation of the processor 6. Further, the processor 6 stops its operation by itself when an abnormality in the main storage device 7 or an access error in the disk device 8 occurs. In this case as well, the system control device 2 is not interrupted, and the fault detection means 21 detects the fault.

When detecting a failure in the processing device 5, the failure detection means 21 activates the status information copying means 22 and the restart execution means 23.

First, the situation information copying means 22 is configured to copy the operation selection setting means 24.
The content of the third l bit is read and it is determined whether the content of the third l bit is ``1'' or not.Here, since the content is previously set to ``1'', the status information copying means 22 copies the status information. will be executed. Specifically, first, the system control device 2 issues an instruction to the processor 6 to copy the contents (status information) of the internal register 6a. Upon receiving this instruction, the processor 6 temporarily transfers the contents of the internal register 6a to an empty area of the main storage device 7. After that, the system controller 2
At the same time as the processor 6 is reset, the program for copying status information stored in the main storage device 7 is started to operate by referring to it. When processor 6 receives this activation instruction, it accesses disk device 8 and copies the status information stored in main storage device 7 to magnetic disk 9 .

Next, the restart executing means 23 of the system control device 2 reads the contents of the operation selection setting means 24 and determines whether the contents of the 30th bit is "1°". Since it is set to 1°, reboot execution method 2
3 instructs the processor 6 to restart.

When processor 6 receives this reboot instruction, it refers to the boot program stored in main storage device 7 and starts its operation.

If the 30th bit of the operation selection setting means 24 is set to "O", the restart execution means 23 of the system control device 2 displays the occurrence of a failure via the operator conversation device 3. In addition, when only restarting is to be executed, the 31st bit of the operation selection setting means 24 is set to ゛゜O.
You can set it to ゛゜.

As described above, recording of status information and restart are executed based on the contents of the operation selection setting means.

The present invention is not limited to the above embodiments.

In the embodiment, the procedure for copying the status information indicating the status of the processor 6 (the contents of the internal register 6a) to the main storage device 7 and then copying it again to the disk device 8 was explained.
For example, the data may be directly copied from the processor 6 to the disk device 8.

(Effects of the Invention) According to the computer system restart control method of the present invention configured as described above, when a failure occurs in a processing device, it can be automatically restarted without any personnel present, and status information can be recorded. I can do it. Therefore, even in systems that are operated unmanned, failure analysis work can be performed later. In addition, since it is possible to select whether or not to automatically restart each system and whether or not to record status information, it is possible to easily change the operation mode depending on the environment in which the system is installed. You can choose.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a computer system according to the present invention, FIG. 2 is a block diagram of a conventional single computer system, and FIG. 3 is a block diagram of an operation selection setting means. 2... System control device, 3... Operator conversation device, 5... Processing device, 6.
... Processor, 6a... Internal register, 7... Main storage device, 8... Disk device, 21... Fault detection means, 22... Status information copying means, 23... Restart execution Means, 24...Operation selection setting means.

Claims (1)

[Scope of Claims] A processor; a system control device configured to detect a failure in the processor and control failure recovery; 1. A computer system restart control method, comprising an operation selection setting means for recording the status of the processor and setting an operation selection as to whether or not to restart the processor.