JPH02278435A - Duplexed computer system - Google Patents

Abstract

PURPOSE:To execute the fault diagnosis of a communication interface part in a computer system itself without affecting another computer system by deciding the presence and absence of the fault in the communication interface part according to whether the contents of a baton reception counter are updated for each prescribed time or not. CONSTITUTION:In a first computer system FC 1, the contents of a baton reception counter 25 to be sent are monitored by a decision part 16 and when the contents are updated for each prescribed time, it is judged that the operation of a communication interface part 2 is normal. On the other hand, when the contents of the baton reception counter 25 are not updated for a time longer than the prescribed time, it is judged that a baton frame can not be received or transmitted since there is the fault in the communication interface part 2. Then, the switch of duplexed control is instructed to make a second computer system FC 2 side being in a standby state up to the moment real work. Thus, the fault of the communication interface part 2 can be detected without affecting communication operation.

Description

[Detailed Description of the Invention] <Industrial Application Field> The present invention relates to a redundant computer system connected to a communication line, and more specifically, each computer system has an internal communication interface section connected to the communication line. The present invention relates to a redundant computer system having a communication interface unit that can be easily diagnosed for failure.

<Prior Art> As one method for increasing the reliability of computer systems, the double-1-hi method has been adopted in the past.

In this case, each computer has a function to diagnose whether its own operation is normal or not, and if a failure is detected in the operating system, it will go into standby mode. Ikekata's computer system, which had been in use since then, was switched over to use for actual work.

Here, the self-diagnosis functions mainly include memory access check (detection of parity errors and no responses), arithmetic check, and Ilo response check.

FIG. 5 is a conceptual diagram showing the internal configuration of one computer system. In the figure, 1 is a processor unit, 2 is connected to the processor unit 1 via an internal bus BSI, and
This is a communication interface unit connected to the communication line L1.

The processor section 1 includes a microprocessor 11 and a memory 1.
2 and an interface 13 to the internal bus BSI. The communication interface unit 2 also includes a line interface 21, a buffer memory 22, and an internal bus interface 23.

In this way, when the computer system includes the communication interface section 2, it is necessary to diagnose this section as well.

Diagnosis of the communication interface section 2 is performed using the buffer memory 2.
2 and the internal bus interface 23 can be easily diagnosed by writing specific data and reading and verifying the data, similar to the diagnosis of the memory of the processor section.

On the other hand, the diagnosis of the line interface 21 connected to the communication line L1, for example,
A loopback method is used, such as reading back the data output.

<Problems to be Solved by the Invention> However, in a system where the communication line L1 has a bus structure and multiple computer systems are connected, it is possible to interfere with the communication of other computer systems while the communication line is in use. Therefore, there is a problem that a loopback test cannot be performed.

The present invention was made in view of such problems, and its purpose is to avoid affecting the communication of Ike's computer system.
The object of the present invention is to realize a redundant computer system that can perform failure diagnosis of its own communication interface and improve reliability.

Means for Solving the Problems> The present invention, which solves the above-mentioned problems, is a redundant computer system connected to a baton bus type communication line that performs a communication operation when the communication right (baton frame) is handed over to itself. , each of the above-mentioned total y:tm systems includes a baton detection section that detects a baton detection section addressed to itself, a baton reception counter that updates the contents when this baton detection section detects a baton frame, and a baton reception counter that updates the contents of this reception counter. and a determination unit that instructs switching of redundant control so that the computer system that has been in a standby state performs actual work if the contents of the reception counter are not updated even after a predetermined reading time has elapsed. be done.

The communication right (baton frame) indicating that communication can be performed using the communication line is sequentially circulated through each system connected to the communication line.If the communication interface section is normal, each system connected to the communication line A baton frame is detected by the baton detection unit at predetermined time intervals corresponding to the number of baton frames.

The contents of the baton reception counter are updated upon detection of a baton frame by the baton detection section.

If the communication interface section is operating normally.

The contents of the baton reception counter are updated every time the baton frame is circulated. The determination means determines whether or not there is a failure in the communication interface unit based on whether the contents of the baton reception counter are updated at predetermined time intervals.

<Example> Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention. In the figure, PCI and Fe2 are connected to the redundant first .
Both of the second computer systems are connected to the communication line L1, and their internal configurations are almost the same.

1st. In the second computer systems FC1 and FC2,
Reference numeral 1 denotes a processor unit, and 2 a communication interface unit for a communication line L1, which is connected to the processor unit 1 via an internal bus BS1.

Communication line L1 sequentially passes communication rights that allow access to this communication line to each computer system connected to communication line L1 and systems other than these computers, and when communication rights reach one's own location, A baton bus (also called token bus) method is used to communicate with each other.

In the processor section 1, 14 is a ready signal generation means for outputting a ready signal based on the content of self-diagnosis; 15 is a duplex control section that receives the ready signal and controls whether it is in an operating state or in a standby state; 16 is the judgment section,
A signal from the readout means 26 of the communication interface section, which will be described later, is input, and the ready signal generation means 14 is generated based on the value.
It controls the operation of the

In the communication interface unit 2, 24 is a baton detection unit that monitors the signal on the communication line Ll and detects that a baton frame indicating the communication right of the communication line L1 has been circulated to itself, and 25 is a baton detection unit 24. A baton reception counter 26 increments the contents by "1" when a baton frame is detected at , and reading means 26 reads the contents of the reception counter 25 and sends the data to the determination section 16 via the internal bus BSI.

FIG. 2 shows the first part in FIG. FIG. 2 is a diagram showing the connection relationship between duplex control units 15 in the second computer systems PCI and FC2.

Each computer system receives a ready signal RDY1 . NAND circuits NG1 to which RDY2 are respectively input. NG2, and the outputs of each NAND circuit are connected to each other so as to be applied to the input terminal of the NAND circuit of the other computer system, so that the two NAND circuits constitute a flip-flop circuit.

This allows the side that activated the ready signal earlier to
The control permission signal 10CE is activated, the control right is acquired, and the device enters the operating state, while the other side performs a duplex switching operation in which it enters the standby state.

The operation of the system configured in this way will be explained next. Here, the first system ';:ta system PCI is in operation state,
Assume that the second computer system FC2 is in a standby state.

In the first computer system, when a communication request occurs, it becomes possible to perform a communication operation when a baton frame is passed to the first computer system, that is, when the baton detection unit 24 receives the baton frame, and When the communication operation is completed (transmission is completed), the baton frame is transmitted to the next computer system.

If a baton frame addressed to itself is received when there is no data communication request, it immediately transmits the baton frame to the next computer system.

Here, the computer system that sent the baton frame is
After the baton frame is sent, there is a failure in the communication interface section 2 of the next computer system, and the transmitted baton frame cannot be received, or the communication frame cannot be sent in accordance with the communication request, or the baton is sent to the next computer system. Errors that prevent frame transmission are detected by monitoring signals on the communication line 1.

If such an error is detected, the computer system changes the destination of the baton frame according to predetermined rules, and retransmits the baton frame to a computer system that can correctly receive the baton frame.

In this way, if a failure occurs in each computer system or other system connected to the communication line L1, the computer xtll & system in which the failure has occurred is removed from the baton frame circulation route and continues to operate normally. The baton frame will be circulated between the computer system and the system in the pond.

In a normally operating computer system, the baton detection section 24 detects a baton frame every time it comes around, and updates the contents of the baton reception counter 25.

The reading means 26 reads the contents of the baton reception counter 25 and sends the contents to the determination section 16 via the internal bus BSI.

The determining unit 16 monitors the contents of the incoming baton reception counter 25, and if the contents are updated every predetermined time (the time required for one baton frame to go around), the determination unit 16 monitors the contents of the baton reception counter 25, and if the contents are updated every predetermined time (the time required for one baton frame to go around), the determination unit 16 updates the contents of the communication interface unit 24. Determine that the operation is normal. On the other hand, if the contents of the baton reception counter 25 are not updated for a period longer than the predetermined time, it is determined that the communication interface section 2 is malfunctioning and the baton frame cannot be received or transmitted.

When it is determined that a failure has occurred in the communication interface section 2, the determination section 16 sends a message to the ready signal generation means 14,
Instructs the duplication control unit 15 to make the ready signal RDYI inactive.

In the duplication control unit 15, when the ready signal RDY1 becomes inactive, the output of the NAND circuit NGI is inverted, and the flip-flop 71 circuit composed of the NAND circuits NGI and NG2 is inverted. As a result, the control permission signal 10CE to the second computer system FC2 becomes active,
This side will be switched to operational status and communication functionality will be restored.

Note that in the second X-machine system FC2 that was in a standby state, the above-described operation in the determination unit 16 is not performed. Diagnosis of the communication interface unit 2 on the standby side is performed by transmitting a diagnostic frame via the communication line L1 and receiving it again under the condition that the communication interface unit 2 of the computer system in the operating state is operating normally. This is done by doing the following.

FIG. 3 is a conceptual diagram of the main parts of an embodiment of the pond of the present invention.

In the embodiment shown in FIG. 1, redundant control sections for controlling switching of the redundant computer systems are provided in the first and second computer systems, respectively, or in this embodiment, a dedicated control section is provided between the two. A redundant control device DXC is provided.

In this case, the redundant control device DXC is the first. Ready signal RDYI directly from each second computer system PCI, FC2
, RDY2, and depending on the states of these signals, it is determined which side is to be put into operation, and a control permission signal is output.

The determination unit 16 detects a failure in the communication interface unit 2 based on whether the contents of the baton reception counter 25 are updated at predetermined intervals, and if a failure is detected, sends a ready signal to the duplication control unit DXC. make inactive. This operation is similar to the embodiment shown in FIG.

FIG. 4 is a configuration block diagram showing yet another embodiment of the present invention.

In this embodiment, a communication interface section 2 of each computer system is connected to a communication line L1 via a coupler section 3, and a prohibition switch 4 is provided to the coupler section 3 to stop the transmission/reception function during maintenance or the like. It has been established.

The status signals indicating the connection state of the prohibition switch 4 are the first . Second computer system PC1,F
It is applied to each determination unit 16 in C2.

The coupler section 3 modulates or demodulates the transmitted signal or the received signal, and the prohibition switch 4 is turned on (prohibited state) when performing maintenance, for example, and prohibits transmitting and receiving operations. The corresponding total of X machine systems are removed from the baton frame's patrol route in the same way as when the failure occurred.

In this embodiment, when the status signal from the prohibition switch 4 does not indicate a prohibited state, the determination unit 16 performs an operation of determining whether the contents of the baton reception counter 25 are updated at predetermined intervals, If the status signal from the prohibition switch 4 indicates a prohibited state, it is determined that the operation of the baton detection section 24 is normal even if the contents of the reception counter 25 are not updated at predetermined time intervals.

According to this embodiment, even if a prohibition switch is provided to stop the transmission/reception function during maintenance or the like, it is possible to accurately detect a failure in the communication interface section.

In addition, in the above explanation, it is not assumed that there is only one type of baton frame circulating on the communication line L1, but in addition to the normal baton indicating communication rights, diagnosis for automatic recovery of a failed system is performed. Prepare a diagnostic baton for
The present invention can also be applied to a system in which the diagnostic baton is circulated at regular intervals. That is, in the case of such a system, the baton detection section may leave the contents of the reception counter as is when a diagnostic baton is detected, and update the contents of the reception counter only when a normal baton is detected.

<Effects of the Invention> As explained in detail above, according to the present invention, failures in the communication interface section can be detected without affecting communication operations, and even in response to failures in the communication interface section,
It is possible to provide a system that can quickly switch between duplexes.

Furthermore, if a prohibition switch is provided to prohibit the transmission/reception function for maintenance purposes, a failure in the communication interface section can be detected without affecting the state of the switch.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing an embodiment of the present invention; Fig. 2 is a block diagram showing the structure of a duplex control section; Figs. 3 and 4
The figure is a block diagram showing an embodiment of the pond according to the present invention, and FIG. 5 is a conceptual diagram showing the internal structure of one computer system in a conventional device. PCI...first control section, Fe2...second control section, Ll...communication line, 1...processor section, 2...communication interface section, 14...ready signal generation means , 15... Duplex control unit, 16... Judgment unit, 24...
・Baton detection unit, 25...Baton reception counter, 26...Reading means, Fig.

Claims (3)

[Claims] (1) A redundant computer system connected to a communication line using a baton pass method that performs a communication operation when the communication right (baton frame) is handed over to oneself, and each computer system has a baton detection system that detects a baton frame addressed to itself. a baton reception counter whose contents are updated when the baton detection section detects a baton frame; and a baton reception counter whose contents are updated when the baton detection section detects a baton frame; A redundant computer system comprising: a determination unit that instructs switching of redundant control so that a computer system in a standby state performs actual work. (2) Each computer system is coupled to a communication line via a coupler section, and the coupler section is provided with a prohibition switch that instructs the prohibition of the communication function, and the determination section determines whether the prohibition switch instructs the prohibition of the communication function. If the contents of the reception counter are not updated even after a predetermined period of time has elapsed under conditions where the computer system is not in a standby state, instruct the redundant control to be switched so that the computer system that had been in standby mode performs the actual work. A redundant computer system according to claim 1, characterized in that: (3) Each computer system has a ready signal generation means that inputs an instruction signal from the determination unit and outputs a ready signal based on the content of the self-diagnosis, and a standby state to see if it will become operational upon receiving the ready signal. and a duplexing control section for controlling whether or not the signal becomes active, and the ready signal generating means makes the ready signal inactive in accordance with an instruction signal from the determining section. computer system.