Fault monitoring system in loosely coupled multi-processor system

[Detailed Description of the Invention]

[0001]

[Industrial Application]

The present invention relates to the failure surveillance method in a loose coupling multiprocessor system, Between the processor for surveillance and the processors for surveillance is especially related with the failure surveillance method in the loose coupling multiprocessor system which performs the failure surveillance in the multiprocessor system combined by the channel of a communication line or a data bus.

[0002]

[Description of the Prior Art]

Although the processor for surveillance will reset at an interval for a definite period of time conventionally using the circuit which alarms to the failure surveillance method in this kind of loose coupling multiprocessor system if not reset beyond the (a) definite period of time, If abnormalities occur, the interval will be extended, an alarm will be emitted, and the supervisory timer method and the processor for (b) surveillance which a surveillance processor recognizes perform self-diagnosis at the time of necessity, The result is passed to a surveillance processor, arbitrary data is periodically sent to the processor for surveillance from a self-diagnostic system and (c) surveillance processor, data is made to return, as a result of passing the processing for which it opted beforehand, and methods, such as a communication response method which carries out a failure judging with data as a result, are adopted.

[0003]

[Problem(s) to be Solved by the Invention]

In order to carry out processing which the conventional failure surveillance method mentioned above took beforehand by the processor for surveillance, the contents of surveillance are fixed and change is not easy, There is a fault that it is difficult to change the target part of diagnosing the failure portion discovered by the surveillance result still in detail or diagnosis.

[0004]

It is in providing the failure surveillance method in the fault joint multiprocessor system which the object of the present invention removed the fault mentioned above, and changed easily by having given flexibility to the contents of surveillance, and made easy a discovered change of the target part of detailed diagnosis of a failure portion, and diagnosis.

[0005]

[Means for Solving the Problem]

A failure surveillance method in a loose coupling multiprocessor system of the present invention, A diagnostic program is sent to the processor for surveillance in the form of processing units, such as a job or a task, between a processor for surveillance, and a processor for surveillance from a communication line or the processor for surveillance of a multiprocessor system with a channel of a data bus, It has the composition which inspects a processing result which it is at the completion time of processing of the diagnostic program, and was returned from the processor for surveillance by the processor for surveillance, and judges a state.

[0006]

[Example]

Next, the present invention is explained with reference to drawings.

[0007]

Drawing 1 is a block diagram showing the composition of one example of the present invention.

[0008]

The example shown in Drawing 1 makes an example the case where one processor for surveillance and processor for surveillance are combined via a data bus, respectively, and is with processor 1 for surveillance, having processor 2 for surveillance, and data bus 3 -- Drawing 1 -- in addition -- three diagnostic programs 4a, 4b, and 4c of the form of processing units, such as a job or a task, The diagnostic program which was sent out to processor 2 for surveillance via data bus 3 from processor 1 for surveillance, and became processing unit 5, and processing result 6 sent out to processor 1 for surveillance via data bus 3 from processor 2 for surveillance are written together and shown.

[0009]

Next, operation of this example is explained.

[0010]

Processor 1 for surveillance and processor 2 for surveillance which were connected by data bus 3 communicate mutually, and have a function.

[0011]

A plurality of diagnostic programs 4a, 4b, and 4c of the form of processing units, such as a job or a task, are saved at processor 2 for surveillance, and it is periodically sent out via data bus 3 to processor 2 for surveillance. In the example of Drawing 1, diagnostic program 4c is sent to processor 2 for surveillance via data bus 3, it performs as processing unit 5 within processor 2 for surveillance, and processing result 6 is returned via data bus 3 after the completion of processing to processor 1 for surveillance.

[0012]

In processor 1 for surveillance, the contents of the post-processing result 8 are inspected, and the existence of failure of Preprocessor 2 for surveillance is judged by the right or wrong.

[0013]

Thus, the thing for which the program for diagnosis is sent to the processor for surveillance in the form of processing units, such as a job or a task, and failure of the processor for surveillance is diagnosed based on the answered diagnostic processing result from processor 1 for surveillance, The contents of diagnosis can be chosen if needed, moreover corresponding to the load of a system, a setup of a diagnostic item becomes possible including an addition, and the fault diagnosis including management of an early fault point can process them efficiently.

[0014]

[Effect of the Invention]

As explained above, in the present invention, a diagnostic program is sent out to a processor for surveillance per processing from a processor for surveillance. Therefore, it is possible to choose diagnostic contents if needed, and diagnosis of the processor for surveillance and the device connected to it becomes possible including the number of diagnostic items according to the load of the system, An addition of diagnostic contents also has the effect that it becomes possible and dealing with a new fault point becomes easy.

[Brief Description of the Drawings]

[Drawing 1]

It is a block diagram showing the composition of one example of the present invention.

[Description of Notations]

1 Processor for Surveillance 2 Processor for Surveillance 3 Data Bus 4a~4c Diagnostic program 5 Processing Unit 6 Processing Result