JPS626302A - N:1 back-up system - Google Patents

Abstract

PURPOSE:To improve the reliability of an n:1 back-up system by providing a storage output device to this back-up system to sample and store the output of an arithmetic controller. CONSTITUTION:An arithmetic and control unit 1 performs various operations and delivers these results and (n) pieces of such unit 1 are provided here. While a single back-up device 2 is set to these units 1. The device 2 is connected to the units 1 via a data collecting means 5 like a data transmission line. When the unit 1 is working normally, the output of this unit 1 is selected by an output switch device 3 and applied to an external device. Here a storage output device 4 is provided to sample and store the output of the unit 1 in a prescribed period. Thus the output of the device 4 is applied to the external device via the device 3 in case the device 2 has a trouble.

Description

[Detailed Description of the Invention] [Industrial Application Field] This invention is constructed by providing one backup device for a plurality of arithmetic and control devices that perform predetermined arithmetic operations and output the results. 1. Regarding improvements to the backup system.

[Conventional technology]

Generally, the method of backing up n arithmetic and control units is to duplicate or triple all of the individual arithmetic and control units.
A 1:n backup method in which one arithmetic and control device is backed up by one backup device is known. Among these, the former 1:n backup method has the advantage that the reliability of the system improves as the number of nOs increases, but has the problem of significantly increasing costs. Therefore, in order to ensure reliability while suppressing cost increases, it is currently necessary to rely on the latter n:1 backup method.

[Problem that the invention seeks to solve]

However, with this nil backup method, if the backup device fails or if at least one arithmetic and control unit fails during backup operation, backup becomes impossible, and the output corresponding to the failed arithmetic and control unit becomes undefined. There is a risk of adverse effects on the controlled object (external equipment).

Therefore, in the n:1 backup method, the present invention prevents the corresponding calculation output from becoming undefined even if at least one arithmetic and control unit fails when the backup device fails or during the backup operation. The purpose is to improve reliability.

[Means for solving problems]

A memory output means is provided for sampling and storing the output of the arithmetic control means at a predetermined period and outputting the previous value based on a predetermined instruction, and the output of each arithmetic control means is sampled and stored by the memory output means. By keeping
To improve the reliability of the system by preventing unstable output from being given to external equipment even in the event of a failure as described above.

[Effect]

FIG. 1 is a main diagram that best represents the features of this invention. In the figure, 1 (ia to In) is an arithmetic and control unit, 2
3 (3a to 3n) is a backup device, 3 (3a to 3n) is an output switching device 54c4a to 4n) is a storage output device, and 5 is a data collection means.

That is, a plurality (n) of arithmetic and control devices 1 that perform various arithmetic operations and output the results are installed, and one
In such a backup system in which a number of backup devices 2 are provided, it is necessary to constantly collect information on the arithmetic and control devices 13 to 1n on the backup device 2 side. It is connected to the device 1 via a data collection means 5 such as a data transmission line. Therefore, as long as the arithmetic and control unit 1 is normal, its output is selected by the output switching device 3 and given to the external device, but if any one of the arithmetic and control units 1 fails, its function is transferred to the backup device 2.
The output of the backup device 2 is selected by the output switching device 3 and given to the external device.

The above is a general n:1 backup system, but
In this case, the output of the arithmetic and control unit 1 is sampled at a predetermined period by the storage/output device 4 and stored. And when the backup device fails,
Alternatively, if at least one of the arithmetic and control devices 1 fails during backup operation, the output from the storage output device 4 can be selected by the output switching device 3 and given to the external device, so that the output to the external device can be made unstable. This will further improve reliability.

〔Example〕

FIG. 2 is a block diagram showing an embodiment of the present invention.

Here, one arithmetic and control device 1a, one output switching device 3a, and one storage output device 4a are shown as representatives of each.

That is, the arithmetic and control unit 1a has a central processing unit (CPU) 1
1. Read only memory (ROM) 12. Random access memory (18, AM) 13. Digital/analog converter (D/A converter) 14 and operational amplifier 1
5, etc., and such a configuration has a backup device [
The same applies to 2. That is, 21tiCPU, 2
2 is an OM, 23 is an AM, 24 is a D/A converter, and 25 is an operational amplifier.

The storage output device 4a includes a comparator 41, an up/down counter 42. It is composed of a D/A converter 43, an operational amplifier 44, shift registers 45, 46, and the like. The comparator 41 constantly tracks the output of the arithmetic and control device 1a by comparing the output from the arithmetic and control device 1a with the output value (P point O value) of the memory output means 4a, and the up/down counter 42 Refresh the content in real time. Shift registers 45 and 46 store the contents of the up/down counter 42 using sampling clocks C and L. Therefore, shift register 4
5 stores the latest value, and register 46 stores the previous value. When a failure as described above occurs, the contents of the second stage shift register 46 are preset into the up/down counter 42, which is converted into an analog signal by the D/A converter 43 and then sent via the operational amplifier 44. The signal is outputted and introduced into the analog switch 3a.

Here, we will explain the reason why the data output from the storage output device 4 in the event of a failure as described above is assumed to be the data that was sampled last time.0 Figure 3 shows the timing of sampling and the occurrence of a failure.

FIG. 3 is a reference diagram for explaining the relationship with detection timing.

Figure (A) shows the relationship between sampling timing and data, and shows the relationship between sampling timing and data. , 1. , 1.・・・・・・
・Data DOs D at predetermined intervals Ts such as tn
l s D2...Dn is sampled.

Here, if the time point at which a failure occurs is tk, then in a case like the one shown in the same figure (b) where a failure can be determined at the same time as this failure occurs, sampling data can be prevented from being taken in by this, but in the case shown in the same figure (b) As in c), when there is a delay time τ between the timing tk of failure occurrence and the timing of determining failure, there is a high possibility that the latest data Dntim sampled during this time has already arrived. Therefore,
The storage output device 4 outputs the previous value instead of the latest data, thereby avoiding the above-mentioned inconvenience. Since we have installed a storage output device that samples and stores the output of the control device, even if the backup device fails, or if another arithmetic and control device fails during backup operation, the output to external equipment will remain undefined. This provides an advantage in that the normal calculation value before the arithmetic and control unit malfunctions can be maintained. In addition, in systems that require time after the arithmetic and control unit fails until backup can be performed using a backup device, the storage output device temporarily outputs data during that time, and when backup becomes possible, By outputting from the backup device, backup can be performed without punching or shock.

[Brief explanation of drawings]

Figure 1 is the main diagram that best represents the features of this invention;
The figure is a book diagram showing an embodiment of this invention, and FIG. 3 shows sampling timing and failure occurrence. This is a reference diagram for explaining the relationship with detection timing. ◇ Symbol explanation 1 (la~In)... Arithmetic control unit, 2...
...Backup device, 3 (3a to 3n)...
・Output switching device, 4 (4a to 4n)...Storage output device, 5...Data collection means, 11.21.
...Central processing unit (CPU), 12.22...
...ROM, 13.23...RAM. 14.24,43...D/A Humpata, 15
゜25.44...Operational amplifier, 41...
- Comparator, 42...up/down counter, 45, 46...e...shift register.

Claims (1)

[Claims] In an n:1 backup system comprising one backup means for a plurality of arithmetic and control means, the output from the arithmetic and control means is sampled and stored at a predetermined period, and the previous value is read upon receiving a predetermined instruction. A memory output means for outputting the output, and an output selection means for selecting a predetermined one of the outputs of the memory output means, the backup means, and the arithmetic control means and outputting the selected one to an external device are provided respectively corresponding to the arithmetic control means, and When the backup means malfunctions or when at least one arithmetic control means malfunctions during the backup operation, the corresponding storage output means is selected by the output selection means, and the output is output from the storage output means to an external device. n:1 backup system.