KR100394736B1 - Method And Apparatus For Fail Recovery And Board Selecting In Dual System - Google Patents

Abstract

In the redundant system, the status of the board maintaining active and standby is monitored through watchdog monitoring to perform redundancy switchover, and to output the reset signal to the error-prone board to automatically recover the software error.

The present invention detects whether an error occurs in the board by accessing the board in charge at regular intervals, and error detection means for outputting a control signal for switching the operation mode of the board when an error is detected; Arithmetic means for logically calculating the operation mode selection signal applied from the detection means and the counterpart board, and outputting the operation selection signal to the own board side; and if the error occurrence signal of the own board is detected from the error detection means, the board side of the board Error recovery means for outputting a reset signal for error recovery.

Therefore, since the automatic recovery is made for the errors caused by the software, the convenience of operating the system is provided.

Description

Board Selection and Error Recovery Device and Method in Redundant System {Method And Apparatus For Fail Recovery And Board Selecting In Dual System}

The present invention relates to a redundancy system, and more particularly, to monitor the state of a board maintaining active and standby by watchdog monitoring to perform redundancy switching, and to generate an error. The present invention relates to a board selection and error recovery device in a redundant system that outputs a reset signal to the board so that software errors can be automatically recovered.

In general, the exchange system is equipped with two identical boards in mirror form for minimizing subscriber complaints by maintaining continuous call setup, and one of them is kept active. The other board is trying to stay on standby.

For example, if the A board maintains the active state among the two boards in the mirror form as described above, the B board maintains the standby state. In this state, the B board maintains the standby state. If the access signal is not detected by the watchdog circuit, the watchdog circuit determines that an error has occurred in the A board maintaining the active state, and transfers the B board maintaining the standby state to the active state and generates an error in the active state. After A board is switched to standby, it waits for the recovery of the A board in error.

As described above, an error occurs in the A board that is in the active state and the switchover to the B board that is in the standby state is performed. After recovering the error, there is a troublesome problem of remounting, and even if a simple error occurs, the board cannot be recovered by itself and must be removed from the slot.

The present invention has been made in view of the above-mentioned general problems, and its object is to detect the active and standby states of each board in a redundancy system through watchdog monitoring to perform redundancy switchover, and to switch to the standby state due to an error occurrence. A reset signal is applied to one board to enable automatic recovery in case of an error caused by software, and to output an alarm signal to the operator for maintenance when an error that cannot be recovered automatically occurs on both redundant boards. To provide convenience and reliability to the operation of the system.

1 is a block diagram illustrating a board selection and error recovery apparatus in a redundant system according to the present invention;

2 is a flowchart of a reset generating means in a board selection and error recovery apparatus of a redundancy system according to the present invention.

<Explanation of symbols for the main parts of the drawings>

10A, 10B: watch dog detector 20A, 20B: end gate

30A, 30B: Reset occurrence and timer circuit

The present invention for achieving the object as described above, in the board selection and error recovery apparatus in a redundant system, by accessing its own board at regular intervals to detect whether or not the error of the corresponding board, the error occurrence is detected Error detecting means for outputting a control signal for switching the operation mode to the board in charge thereof; Calculating means for performing a logic operation on the control signal applied from the error detecting means and the operation mode selection signal applied from the counterpart board, and outputting an operation mode selection signal to the board in charge of the controller; And error recovery means for outputting a reset signal for error automatic recovery to the corresponding board side when an error occurrence of the own board is detected by the error detection means.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

As can be seen in Figure 1, the board selection and error recovery apparatus in the redundant system according to the present invention, the A board and the B board is the same configuration, that is, watch dog detection unit (10A, 10B) and end gate (20A, 20B) And a reset generation and timer circuits 30A and 30B to maintain a mirror shape, wherein the watchdog detection units 10A and 10B are responsible for their own maintenance at regular intervals when the boards they are maintaining are active. The board accesses the board to detect whether the board has an error, and if it is determined that the board in charge of the board cannot execute a normal function due to an error, it outputs a control signal for switching to another board in the standby state.

The AND gates 20A and 20B are logically multiplied by a signal output from the watch dog detectors 10A and 10B and applied to one terminal and a signal applied to the other terminal after being inverted after being output from the AND gate of the counter board. Outputs an active or standby signal to the board responsible for it.

When the reset generation and timer circuits 30A and 30B detect an error signal for their own board output from the watchdog detectors 10A and 10B, they output a Reset * signal for error recovery to their board. At the same time, the counter starts and outputs the reset * signal. If the error recovery is not detected even after a certain time has elapsed, it is determined that the board has a hardware error and outputs a system error signal.

In addition, when a non-recoverable hardware error occurs in both boards configured as a separate board and implemented as the redundancy, a system by logically multiplying the reset occurrence and the system error signal applied from the timer circuits 30A and 30B And gate 100 is provided that is a logical operation means for outputting a signal for a system error to the operator of.

In the present invention having the technical configuration as described above, the operation of performing system monitoring and redundancy switching and recovery of the error occurrence board are as follows.

For example, if the A board is active and the B board is in a standby state, and each board is in a normal state, the signal output from the watchdog detection units 10A and 10B is in a "high" state. Outputs the signal of.

Therefore, the signal output from the AND gate 20A of the A board maintains the "high" state, and the signal output from the AND gate 20B of the B board maintains the "low" state.

As described above, while the A board is active and the B board maintains standby, the watchdog detection unit 10A on the A board side accesses the state of the A board, which is in charge of itself, at set time intervals, to which A is responsible. Determine if an error is detected on the board.

If the state of the A board detected by the watchdog detection unit 10A is determined to be an error, the own board maintains the standby state and outputs a control signal for switching the B board holding the standby state to the active state. do.

Accordingly, the AND gate 20A of the A board is inverted from the signal input to one terminal of the watch dog detector 10A from the output terminal of the AND gate 20B of the B board which is in the standby state. Logically multiply the signals to switch the state of the A board in charge from the active state to the standby state, and the AND gate 20B of the B board maintaining the standby state is applied from its watchdog detection unit 10B. The state of the B board is transferred from the standby state to the active state by performing a logical multiplication with the signal to be output from the AND gate 20A of the A board and the inverted state.

In the case where an error occurs in the A board maintaining the active state, the signal output from the watchdog detection unit 10A outputs a signal of the "low" state for requesting redundancy switching, and maintains the standby state. Since the AND gate 20B of the board outputs a signal of the "low" state, the AND gate 20A of the A board is inverted from the "low" signal input to one terminal and the "low" signal to the other terminal. Logically multiply the "high" signal to cause the A board to transition to standby as a "low" state.

At this time, the signal of the "low" state output from the A board end gate 20A is input in the "high" state as an inverted input to the other terminal of the B board end gate 20B, and the watch of the B board in this state. Since the signal output from the dog detector 10B also outputs a "high" signal due to the normality of the B board, the AND gate 20B of the B board outputs a "high" signal to activate the B board in the standby state. Switch to.

As described above, the A board in the active state is switched to the standby state due to the occurrence of an error, and the B board in the standby state is transferred to the active state. The reset generation and timer circuit unit 30A outputs a reset signal Reset * for recovery of the generated error to the board side thereof, and simultaneously operates a counter circuit to execute counters of a set time (S101 and S102).

When the elapsed time set as the counter is detected in the state in which the reset signal Reset * for error recovery is output as described above (S103), it is determined whether an error of the A board in charge of the self has been recovered (S104).

If it is determined that the error has been recovered, the standby state is maintained until an error occurs in the B board and is transferred. If the error recovery is not detected within the time set by the counter, operation and maintenance of a hardware error occurred in the A board. A signal for notifying a repairer is output to induce recovery from an error generated by dismounting from the slot (S105).

And gates mounted on the external board by signals output from both boards when an error occurs in both boards configured as redundant in the above, and thus is not restored even in the occurrence of reset and control signals by the timer circuits 30A and 30B. By the logical product operation of (100), a signal of a system error is output to the operation and maintenance personnel.

As described above, according to the present invention, when an error occurs in a board in charge of the watchdog in a redundant system, another board that is in standby state is transferred to an active state, and a reset signal is applied to the board. It is possible to recover errors by software by outputting and provide convenience for error recovery, and to eliminate unnecessary operation of removing the board from slot for error recovery.

Claims (5)

In the board selection and error recovery device in a redundant system, Error detection means for accessing its own board at regular time intervals to detect whether an error occurs in the corresponding board, and outputting a control signal for switching the operation mode to the owning board when an error is detected; Calculating means for performing a logic operation on the control signal applied from the error detecting means and the operation mode selection signal applied from the counterpart board, and outputting an operation mode selection signal to the board in charge of the controller; And error recovery means for outputting a reset signal for automatic error recovery to the corresponding board side when an error occurrence of the own board is detected by the error detection means. . The method of claim 1, wherein the calculation means comprises an AND gate, and the control signal of the error detection means is input to one terminal, and the signal inverting the operation mode selection signal of the counter board is input to the other terminal. Board selection and error recovery in a redundant system, characterized in that the logical multiplication operation. The method of claim 1, wherein the error recovery means outputs a reset signal for automatic error recovery, starts a counter, determines whether an error recovery of the corresponding board has been performed within a set time, and then outputs a corresponding signal. Board selection and error recovery device in a redundant system, characterized in that. The system error notification signal is output by logically calculating a signal output from the error recovery means of both boards according to whether the board recovers the error from the error recovery means of both boards. Board selection and error recovery apparatus in a redundant system, characterized in that it comprises a calculation means to a separate board. In the board selection and error recovery method of a redundant system, Determining whether an error is detected in a board in charge of an activated state accessed at a set time interval; When an error occurs in the active board in the active state, a redundant switchover is performed with other boards that remain in the inactive state, and at the same time, a reset signal for automatic error recovery is output to the error generating board, and then the error recovery is verified. Counting the set time to perform; Determining whether an error of a board which attempted automatic error recovery has been recovered through the output of the reset signal when the set time has been detected, and notifying an operator of an error of a system according to the determination result. Board selection and error recovery in redundant systems.