JPS63291132A - Fault deciding device for composite computer system - Google Patents

Abstract

PURPOSE:To correctly judge a fault location, to attain a suitable fault processing and to improve the reliability of a composite computer system by supervising mutually other computer with plural computers and exchanging the supervisory information. CONSTITUTION:A computer having a transmission right of a health signal transmits successively the health signal to other computer and stores the fault supervising information collected already by an opponent computer. When the information collection is completed, the fault of the computer is decided, and the stored fault information is informed to other computer. A fault deciding processing is executed on the computer having the transmission right of the health signal at a constant time internal and a health signal transmission right is cyclicly transferred between computers. Thus, plural computers mutually supervise other computer and the supervisory information is exchanged. Thus, the fault location is correctly judged, a suitable fault processing is executed and the reliability of the composite computer system can be improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an information processing system. In particular, the present invention relates to a failure determination device for a compound computer system.

〔overview〕

In a multicomputer system fault determination device, multiple computers mutually monitor other computers and exchange monitoring information to accurately determine the location of the fault, perform appropriate fault handling, and improve the performance of the multicomputer system. This is designed to improve reliability.

[Conventional technology]

Conventionally, failure determination of computers constituting a compound computer system or communication paths connecting them has been made by a single computer.

[Problem that the invention seeks to solve]

In the conventional method described above, a normal computer is treated as a failure due to a mistake made by one of the computers that performs failure determination, or a failure is determined to be a computer failure even though there is a communication path failure, resulting in failure of the system to handle the failure. , it is necessary to disconnect the faulty computer, continue running the job, or switch the communication path, which causes confusion between the computers, and the entire system falls into an unstable state, making normal system operation impossible. It has the problem of causing

SUMMARY OF THE INVENTION The present invention solves these conventional problems and aims to provide a device that accurately determines faults.

[Means for solving problems]

The present invention is characterized in that in a composite computer system in which a plurality of computers are interconnected by communication paths, each computer is provided with a fault monitoring means for monitoring faults in other computers.

The fault monitoring means includes means for sending a fault monitoring signal to other computers, means for storing fault monitoring information responded from other computers in response to this signal, and detecting the faulty computer from the information stored by the means. ``a means for determining, and the plurality is three or more.

[Effect]

A computer that has the right to transmit health signals sequentially transmits health signals to other computers, and stores fault monitoring information that the other computers have already collected.

Once this information collection is complete, we will determine the computer's failure.
The stored fault information is reported to other computers.

Failure determination processing is performed on a computer that has the right to transmit health signals at regular time intervals, and the right to transmit health signals is cyclically delegated between computers.

In this way, multiple computers mutually monitor other computers and exchange monitoring information, thereby accurately determining the location of a failure, performing appropriate failure handling, and improving the reliability of the complex computer system. be able to.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention using three computers, and FIG. 2 is a block diagram showing the fault monitoring function and overall configuration of the embodiment of the present invention. The embodiment of the present invention is composed of a computer 1.2.3, a bus 4 as a main communication path, and a bus 5 as an auxiliary communication path.

The computer 1 is equipped with a processor 6, a main memory 7, and bus control units 8 and 9, and the main memory 7 includes a fault monitoring signal sending means 3 for monitoring other computers and communication paths.
1, a fault monitoring information storage means 32 that stores the information obtained by the fault monitoring signal sending means 31 in the fault monitoring information table 10, and a fault monitoring information storage means 32 that notifies other computers of the fault information stored by the fault monitoring information storage means 32. A fault monitoring function 11 and a fault monitoring information table 10 include a fault monitoring information notifying means 33 and a faulty computer determining means 34 for determining a faulty computer based on the fault monitoring information notified by the fault monitoring information notifying means 33. It is equipped.

Computers 2 and 3 also have processors 12 and 1, respectively.
8, main storage devices 13 and 19, and bus control unit 14.
15.20 and 21 are included, and like the computer 1, the main storage devices 13 and 19 have fault monitoring functions 17 and 2.
3 and fault monitoring information tables 16 and 22, respectively.

In the embodiment of the present invention configured as described above, a computer having the right to transmit a health signal sequentially transmits a health signal to other computers using the fault monitoring signal sending means 31, and the fault monitoring information that the other computer has already collected is transmitted. Fault monitoring information storage means 3
Collect by 2. When this information collection is completed, the failed computer determining means 34 determines the failure of the computer.

This fault determination processing is performed at regular time intervals on a computer that has the right to transmit health signals, and the right to transmit health signals is also cyclically delegated between computers.

FIG. 3 shows an example of the contents of the failure monitoring information table 10 of the computer 1. Here, N84 (i≠j11≦
i, j≦3, N, , = O11) indicates the state of computer j detected by computer i; if Ni, = 0, it is normal; Ni, =
If it is 1, it indicates an abnormality. Computer i calculates these N1j (1≦
i, j≦3) is sent back to the calculation stage by the failure monitoring information notification means 33 along with a response signal when receiving a health signal from another computer.

Then, this information from all computers (N(1, t#j, l≦
When all the calculations (i, j≦3) have been completed, the condition Σ N, J≧2 (A) is checked, and the faulty computer determining means 34 determines that the calculation aj that satisfies this is faulty.

One example will be explained here. As shown in FIG. 4, the computer 3 receives a time T. Suppose that a failure occurs and it is no longer able to respond to health signals from other computers. At time T, computer 1 obtains the right to transmit a health signal and sequentially sends health signals to other computers 2 and 3. The computer 2 also sends N2 and N23 in the fault monitoring information table 16 in the computer 2 together with the response signal.

After receiving this information, the computer 1 updates Nt3 in the failure monitoring information table 10 in the computer 1 with the information, and then determines that the computer 2 is normal and sets Nlt to 0.

On the other hand, since a failure has occurred in computer 3, a response signal cannot be sent back to the health signal from computer 1. For this reason, computer 1 cannot update the failure monitoring information N31%N,2 of computer 3 in the failure monitoring information table 10 and remains in the same state, but eventually computer 1 will detect an abnormality in computer 3. Let N13 be 1. As a result, the contents of the failure monitoring information table 10 become as shown in FIG.

In this failure monitoring information table 10, N1. is "l", which indicates that computer 1 cannot confirm the status of computer 3. Since communication path failure may be the cause of this, the computer 1 temporarily sends the health signal to the computer 3 again using the auxiliary bus 5 and performs the same failure determination.

As a result, if there is no response from computer 3 as usual, computer 3 is determined to be abnormal and stored as N13 on the failure monitoring information table IO on computer 1, and if there is a response from computer 3, it is determined to be a route failure. Then, each computer is notified that the bus 5 will be used for communication with all computers thereafter, and N+3= on the failure monitoring information table 10 of computer l.
It is changed to O.

Furthermore, at time Tz (-T1 +T, T is the inter-computer delegation interval for failure processing), computer 2 obtains the right to transmit a health signal and collects failure monitoring information from each computer, and computer 1
The fault monitoring information table 16 is created on the computer 2 in the same manner as the previous example. If the fault has not been recovered in computer 3 at this time, the fault monitoring information table 16 will be as shown in Figure 6, and condition (A) will be applied to computer 3.
is satisfied, and the two-hand calculator of calculators 1 and 2 determines that computer 3 is at fault.

〔Effect of the invention〕

As described in detail above, according to the present invention, a plurality of computers mutually monitor other computers and exchange monitoring information, thereby making it possible to accurately determine the location of a failure. This has the effect of being able to handle failures and improving the reliability of the compound computer system.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention. FIG. 2 is a block diagram showing the fault monitoring function and overall configuration of the embodiment of the present invention. FIG. 3 is a diagram showing an example of the contents of a failure monitoring information table according to an embodiment of the present invention. FIG. 4 is a time chart assuming that a failure occurs in the computer 3 in the embodiment of the present invention. FIG. 5 is a diagram showing the contents of a failure monitoring information table for computer 1 obtained from the results of failure processing for computer 1 starting at time T in the embodiment of the present invention. FIG. 6 is a diagram showing the contents of a failure monitoring information table for computer 2 obtained as a result of failure processing for computer 2 starting at time T2 in the embodiment of the present invention. 1.2.3...Calculator, 4.5...Bus, 6.12
．． 18...Processor, 7.13.19...Main storage device, 8.9.14.15.20.21...Bus control unit, 10.16.22...Fault monitoring information table, 1
1.17.23... Fault monitoring function, 31... Fault monitoring signal sending means, 32... Fault monitoring information storage means, 3
3... Fault monitoring information notification means; 34... Fault pointer calculator determination means.

Claims (3)

[Claims] (1) A failure determination device for a compound computer system in which a plurality of computers are interconnected by communication paths, wherein each computer is provided with a failure monitoring means for monitoring failures of other computers. (2) The fault monitoring means includes a means for sending a fault monitoring signal to other computers, a means for storing fault monitoring information responded from other computers in response to this signal, and a means for storing fault monitoring information from the information stored by this means. The apparatus according to claim 1, further comprising means for determining a faulty computer. (3) The device according to claim (1), wherein the plurality is three or more.