JPS6172496A - System resuming system of duplex information processor - Google Patents

Abstract

PURPOSE:To limit a service interruption to minimum by writing the content of a self system memory to an external memory by a preliminary processor and reading the content of a file in the self system memory when a difficulty is detected during the operation of both systems and one system is operated. CONSTITUTION:An external memory FMO is used as a system file and an external memory FM1 is used as a failure information collecting fie. During a normal operation, a central processing unit CC#1 of a present system and an internal memory MM#0 of the present system are in a condition of operation (ATC) and a central processing unit CC#1 of a preliminary system and an internal memory MM#1 of the preliminary system are in the condition of a standby (SBY). When both the systems are operated and an abnormal condition is detected and one system is operated, the unit CC#1 writes automatically the contents of the memory MM#1 to the memory FM1 and reads the contents of the memory FMO to the memory MM#1 and backs up the memory.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to an information processing device that has a redundant equipment configuration, and in particular detects an abnormality during the operation of the device and automatically corrects the failure when a system down occurs. The present invention relates to a system restart method for a redundant information processing device that allows information to be collected and the system to be restarted quickly.

In online systems such as electronic exchange systems, main devices such as processors and memories are often duplicated to ensure continuous operation of the system. If a failure occurs during operation of such a redundant system, it is desired to restart the system with as little disruption to system operation as possible.

[Conventional technology]

FIG. 4 is a diagram showing an example of a conventional duplex information processing device.

In the figure, CC#O is the active system central processing unit, CC#1 is the backup system central processing unit, MM#0 is the active system internal memory, and MM#
1 is the backup system internal memory, EMA#O is the active system emergency circuit, EMA#1 is the backup system emergency circuit, B
US#O is the active bus, BUS#l is the backup bus, IO
C is input/output control device, FMO is external memory device, FMI
is an external memory device. The same symbols represent the same objects throughout all the figures below.

In FIG. 4, the external memory device FMO is used as a system file, and the external memory device FM1 is used as a failure information collection file.

During normal operation, the active system central processing unit CC#0 and the active system internal memo + JMM#0 are in the operating state (hereinafter referred to as ACT).
The standby central processing unit CC#1 and the standby internal memory MM#1 are in standby state B (hereinafter referred to as SBY).

In both systems, the information necessary for operation is always stored in the internal memory M of the active system.
The data is input to M#0 and the backup system internal memory MM#1 and updated, but the backup system does not operate.

FIG. 5 is a diagram illustrating an example of a conventional system restart method, in which fa) is the operating state of the active system, (bl is the operating state of the standby system, (C) is the operating type, and (d) is the operating state of the active system. Indicates the service status respectively.

In an online system with a redundant configuration like this, the fifth
If an abnormality occurs in the active system as shown in Figure (a) (indicated by an x mark), the active system emergency circuit'
EMA #0 operates and the standby system enters the ACT state.

Therefore, as shown in FIG. 5 (C1), the two-system operation changes to one-system operation. During this time, the service state is temporarily stopped as shown in FIG. 5(d), but the time is in seconds.

Spare central processing unit CC#L and backup internal memory MM
When #1 is in the ACT state, if an abnormality occurs in the standby system (indicated by *), the standby system emergency circuit EMA#
1 operates to bring the active system central processing unit CC#O1 and the active system internal memory MM#0 into the ACT state.
), the contents of the system file are read from the external memory device FMO during the period indicated by the dotted line, and the contents of the system file are read from the external memory device FMO, and
Must be stored at 0.

The time required for the above operation is usually several minutes, and the disadvantage is that the system user cannot use his/her terminal (including telephone) during this time.

Particularly in the case of an online system such as an electronic switching system that must be operated constantly, the long system interruption time is a major social problem.

Another drawback is that it is necessary to write out the information remaining in the spare memory at the time of abnormality detection to the external memory device FMI.

[Problem that the invention seeks to solve]

The purpose of the present invention is to eliminate the above-mentioned drawbacks of the conventional system, and to provide serviceability and maintenance that minimizes the service interruption time for online system users even if the system goes down by detecting system abnormalities during both system operation. An object of the present invention is to provide a system restart method for a duplex information processing device with excellent performance.

[Means for solving problems]

The solution to the problem is that in an information processing device that has a dual processor and memory configuration, if an abnormality is detected during both system operation and one system operation is started, the standby processor automatically This is achieved by writing the contents of the own system liquid memory to external memory and reading the contents of the system file into the own system liquid memory.

[Effect]

According to the present invention, when a system abnormality is detected during dual-system operation and the system becomes single-system operation, the backup system (formerly active system) processor writes the pre-01u system memory to an external memory device and saves system files. Since the data is read and memory backed up, and the system can be restarted quickly when the next abnormality is detected, the time required to restart the system is greatly shortened, resulting in a significant improvement in the service status.

〔Example〕

FIG. 1 is a diagram for explaining the operation of the redundant information processing apparatus according to the present invention.
(b) is the operating state of the standby system, (C) is the operating type, (d+
indicates the service status.

FIG. 2 is a flowchart illustrating the operation of the duplex information processing apparatus according to the present invention.

FIG. 3 is a diagram illustrating the contents of system control information.

The present invention will be described in detail below with reference to the drawings. During normal both system operation, emergency times iEMA#O1 and EMA#1
Both counters have initial values, and the 5BY-BUSY flag is off.

When the active system emergency circuit EMA#0 detects an abnormality in the active system central processing unit CC#0 and an EMA interrupt is generated, processing 1 starts.

In process 1, it is checked whether the 5BY-BUSY flag is in the off state B (backup central processing unit CC#1 is not processing), and if it is not being processed, in process 2 the active central processing unit CC #1 is checked.
C#O and the standby system central processing unit CC#1 are switched, and in process 3, the EMA counter of the standby system emergency circuit EMA#1 is incremented by 1.

If the 5BY-BUSY flag is on in process 1 (standby central processing unit CC#1 is in use), processes 2 and 3 are skipped.

In process 4, the processor mode is identified, and if it is the active system, the own memory is restarted in process 5, and in process 6, the operation type is determined based on information such as the EMA counter and interrupt cause, and both systems are operated. A decision is made as to whether it will be single-system operation or single-system operation.

If the standby central processing unit CC#1 is identified in process 4, the SOY-BUSY flag is checked and checked in process 7. If the SBY-BUSY flag is off, process 8; if the EMA counter is at a constant value (n), process 9.
The 5BY-BUSY flag is set to on, and in process 10 the contents of the spare internal memory MM#l are written to the external memory FMI which is the failure information collection file, and in process 11 the contents of the external memory device FMO which is the system file are written. Write to backup internal memo MF, 4 #1, process 1
In step 2, the SBY-BUSY flag is set to off, making it capable of interrupts.

If the EMA counter is not at the constant value (n) in process 8, processes 9, 10, 11, and 12 are skipped, and the process waits for an interrupt.

Furthermore, if the SBY-BUSY flag is on in process 7, a rest-foot process is performed to continue the process before the interruption.

The above operations are summarized in the flowchart shown in FIG.

In Fig. 1(a), the thin solid line indicates the external memory FM1.
The dotted line represents the time required to read the system file.

As shown in the system information in Figure 3, the EMA counter, system mode (0/1 = dual system operation/single system operation), and 5
The BY-BUSY7 lag (0/1 = off/on) is the same value for both systems, but the processor mode (0/1 = active system/standby system), processor machine number (0/I - active system/standby system) The values for the system are different for the active system and standby system.

In the above explanation, an example has been described in which one processor is used in each of the active system and the standby system, but the present invention is equally applicable to the case where a plurality of processors are used. Furthermore, in the above explanation, two external memories, a system file and a failure information collection file, are used, but the same effect can be achieved even if the system file and failure information collection file coexist in one external memory.

〔Effect of the invention〕

As explained in detail above, according to the present invention, even if an abnormality is detected in the system during operation of both systems and the system goes down, serviceability and maintenance that minimize service interruption time for online system users can be achieved. This has the great effect of realizing a system restart method for duplexed information processing equipment with excellent performance.

[Brief explanation of the drawing]

FIG. 1 is a diagram for explaining the operation of the duplex information processing apparatus according to the present invention. FIG. 2 is a flowchart illustrating the operation of the duplex information processing apparatus according to the present invention. FIG. 3 is a diagram illustrating the contents of system control information. FIG. 4 is a diagram showing an example of a conventional duplex information processing device. FIG. 5 is a diagram illustrating an example of a conventional system restart method. In the figure, CC#O is the active system central processing unit, CC#1 is the backup system central processing unit, MM#O is the active system internal memory, and MM#
1 is a backup internal memory, EMA#0 is a working emergency circuit, EMA#1 is a backup emergency circuit,
BUS#O is the active bus, BUS#1 is the backup bus, I
OC is input/output control device, FMO is external memory device, FM
I is an external memory device. Koto (Kuchi Hayabusa 2 figure)

Claims (1)

[Claims] In an information processing device with a duplex configuration of processors and memory, if an abnormality is detected during operation of both systems and one system is operated, the standby processor automatically transfers the contents of the memory of its own system to the external memory. 1. A system restart method for a redundant information processing apparatus, characterized in that the contents of a system file are written to the memory of the system itself, and the contents of the system file are read into the memory of the system itself.