JPS6215893B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention detects when a failure occurs in the active computer system in a multiple computer system having an active computer system and a standby computer system, and This invention relates to an automatic switching control method that automatically causes a standby computer system to continue processing of an active computer system by using an automatic restart function upon recovery from an abnormality.

Conventionally, system switching for this type of multiplexed computer system has been carried out in two ways: an operator manually switches the system by determining alarm signals etc. issued by a monitoring device, and another method involves diagnosing the active computer system from the standby computer system and changing the system to the active computer system. BACKGROUND ART Systems have been adopted in which, when an abnormality is detected in a computer system, the standby computer system issues a switching command to a system switching device to automatically switch the system.

However, in the former method, in which the operator manually switches the system, the process required to restart the system is relied on manually, so there is a recovery time, that is, the time from when the system goes down until the next process is executed. It had the disadvantage of requiring a considerable amount of time. In addition, in the latter method of diagnosing the active computer system from the standby computer system, a diagnostic function is added to both the active computer system and the standby computer system, data is exchanged between the systems, and the diagnostic program is periodically executed. Since the system needs to be executed manually, it has the drawback that the devices and programs that make up the system are complicated and expensive.

The present invention provides a relatively simple device that utilizes the interrupt function at the time of a power failure and the automatic restart function at the time of power failure recovery, which the conventional computer system had, and which the computer system has conventionally had. By using this program, an inexpensive automatic switching control method is provided which eliminates the above-mentioned drawbacks in a relatively small-scale multi-computer system.

In the automatic switching control method of the present invention, both the active computer system and the standby computer system each have an interrupt function in the event of a power failure and an automatic restart function in the event of recovery from a power failure. In a multiple computer system equipped with means for monitoring the system status and transmitting a pseudo power failure signal to the standby computer system, when a fault in the active computer system is detected, the monitoring means sends the standby computer to the standby computer system. For the system,
By sending a signal with a different voltage level, current level, or timing compared to the original power supply error signal, that is, a pseudo power supply error signal, the standby computer system is initialized and restarted to become the active system. It is something that makes you

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

Computer system A is composed of a main storage device 1-A, a central processing unit 2-A, a channel 3-A, an input/output control device 4-A, a typewriter 9-A, etc., and a computer system B is composed of a main storage device 1-A. B, central processing unit 2-B, channel 3-B, input/output control device 4-
B, typewriter 9-B, etc.
The monitoring device 8 monitors the operating status of the computer system A and the computer system B, changes the system configuration, etc. Magnetic disk pack control device 5
Each of the communication control devices 6 has a changeover switch therein, and can be connected to either the computer system A or the computer system B in response to a changeover command from the monitoring device 8. In the figure, dotted lines indicate the pseudo power supply abnormality signal and the pseudo power supply abnormality recovery signal 10.

Assume that computer system A is currently in use and is executing an online job, and computer system B is executing an offline job or is in a stopped state. At this time, the magnetic disk pack control device 5 and the communication control device 6 are connected to the computer system A. Note that an online program is stored in a part of the magnetic disk pack device 7, and the online program is read out from the beginning of the main storage device 1-B, loaded into the main storage device 1-B, and executed. The program for this is stored.

System switching is performed as follows. Assume that the monitoring device 8 detects a failure in the active computer system A. First, the monitoring device 8 sends a pseudo power supply abnormality signal 10 to the central processing unit 2-B, the magnetic disk pack control device 5, and the communication control device 6. By receiving the pseudo power supply abnormality signal 10, the standby computer system B enters the same state as at the time of the power supply abnormality, and if an offline job is being executed, the processing is interrupted and initialized. Upon receiving the pseudo power supply abnormality signal 10, the magnetic disk pack control device 5 and the communication control device 6 disconnect from the active computer system A and perform a switching operation to connect to the standby computer system B. conduct.

Typically, the above switching operation is completed within a few seconds.
Thereafter, the monitoring device 8 cancels the pseudo power supply abnormality signal 10 and transmits the pseudo power supply abnormality recovery signal 10.
As a result, the already initialized central processing unit 2-B executes the program stored at the beginning of the main storage device 1-B, reads the online program from the magnetic disk pack device 7, and stores it in the main storage device 1-B. 1. Load into B. By creating the online monitor program so that it starts executing after the load is completed, the standby computer system can process online jobs as the active computer system.

Above, we have explained an example in which the pseudo power supply abnormality signal serves both to initialize the standby computer system and to issue a switching command to the magnetic disk pack control device, communication control device, etc.; Alternatively, the above switching command may be performed using another switching signal.

As explained above, the present invention uses a pseudo power supply abnormality signal to initialize and restart a standby computer system and to perform switching operations of a magnetic disk storage control device, a communication control device, etc. Computer systems are equipped with an interrupt function in the event of a power failure and a restart function in the event of recovery from a power failure, so automatic switching can be easily achieved by connecting a pseudo power failure signal from a monitoring device to the standby instrumentation system. can be made to perform an action. Further, there is no need for means for coupling between computer systems or a complicated diagnostic program, which were required in the conventional method of diagnosing an active computer system from a standby computer system. Therefore, according to the present invention, an automatic switching control system can be realized at low cost by adding a simple function to an existing multiple computer system.

[Brief explanation of the drawing]

The figure is a block diagram of one embodiment of the present invention. 1-A, 1-B...Main storage device, 2-A, 2-B
...Central processing unit, 3-A, 3-B...Channel, 4
-A, 4-B...I/O control device, 5...Magnetic disk pack device, 6...Communication control device, 7...Magnetic disk pack device, 8...Monitoring device, 9-A, 9-B
...Typewriter, 10...Pseudo power supply abnormality signal and pseudo power supply abnormality recovery signal.

Claims (1)

[Claims] 1. An active computer system and a standby computer system that generate an interrupt when a power supply abnormality occurs and are automatically restarted when the power supply abnormality is recovered, and a fault monitoring means that detects a fault in these systems, the fault monitoring means detects a failure in the active computer system, transmits a pseudo power supply abnormality signal to the standby computer system, initializes the standby computer system, and then causes the standby computer system to become active. An automatic switching control method characterized by switching control.