JPH07111504A - Communicating function maintenance diagnostic system - Google Patents

Abstract

PURPOSE:To provide a communicating function maintenance diagnostic system which can evade the misdiagnoses due to the faults of the abnormality processing side and also can improve the maintenance diagnostic reliability. CONSTITUTION:A maintenance interface 9 selects an abnormality pattern out of an abnormality pattern library group 91 in response to the maintenance diagnostic request of an operator. An abnormality pattern generating part 10 generates a abnormal cell to the selected abnormality pattern and sends the cell to a diagnostic point node through a cell transmitting part 3, a circuit interface 5 and a circuit 6. At the diagnostic point node, the abnormal cell received through the interface 5 and a cell receiving part 4 is detected at an abnormality detecting part 7. Then an abnormality analyzing/informing part 8 analyzes the abnormality by comparing it with a fault data base 81 and checks whether the abnormality detecting side correctly operates to a pseudo fault.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a communication function maintenance / diagnosis system for carrying out maintenance / diagnosis of communication functions of a system for performing communication using fixed length packets.

[0002]

2. Description of the Related Art For example, as a typical example of a conventional communication function maintenance / diagnosis method for detecting an abnormality of a communication device and performing maintenance thereof in a system for performing communication using fixed-length packets, a communication function failure is shown. When a failure occurs, a communication mechanism is detected by a detection mechanism prepared in advance, and the result is notified to the CPU that manages and operates the communication function, and the CPU periodically reads the status for failure detection, It is known to detect normality / abnormality and notify the operator of the abnormality.

Among them, the abnormality detecting mechanism according to the former method can be considered as a structure as shown in FIG. 4, for example. That is, FIG. 4 is a method of detecting an abnormality by interrupt control, and when an abnormality detection unit 101 realized by hardware detects an internal abnormality of a hardware mechanism or an external abnormality such as a network failure, the interrupt control unit 102 is activated. Then, an interrupt is generated in the CPU 110, and the abnormality is notified by software that monitors a failure such as the occurrence of an abnormality in the CPU 110. After the abnormality is determined, it is notified to a higher CPU or an operator. .

The abnormality detecting mechanism according to the latter method can be considered as a structure as shown in FIG. That is, FIG. 5 is for performing abnormality detection by polling, and similarly to the mechanism shown in FIG. 4, an internal abnormality or an external abnormality is detected by the hardware abnormality detection unit 101, and the abnormality is displayed as status information for the CPU 110. After displaying on the display unit 103, the CPU 110 periodically performs a polling process on the display unit 103 to discriminate the abnormal information, and when the abnormal occurrence is recognized, the upper CP
It is designed to notify U or the operator.

As can be seen from the above description, the conventional communication function maintenance diagnosis method employs a passive method of coping with the abnormality when it occurs, and the abnormality is detected and notified. No consideration was given to malfunctions due to abnormalities in the processing system itself. Therefore, for example, when a failure including hardware failure that cannot be predicted when it occurs occurs, if the failure of the abnormality processing system itself for grasping the abnormality has already occurred, the hardware at this time It became impossible to correctly notify the operator of the abnormal condition.

In order to eliminate such inconvenience, a test device for inducing a bit error or the like in a line is provided outside the communication device, and this test device creates an abnormal condition in a pseudo manner to detect a failure in an error processing system. A method of performing a motion test is also possible. However, this method cannot always simulate the occurrence of a failure unique to the target communication device, and lacks simplicity.

[0007]

As described above, in the above-mentioned conventional communication function maintenance diagnosis method, no consideration is given to erroneous diagnosis due to a failure of the abnormality processing system itself, and a hardware failure or the like whose occurrence time is unknown. If a failure that includes the failure processing system itself has already occurred when a failure that includes a failure occurs, it is not possible to accurately notify the higher-level processing unit such as the operator of the failure, and the reliability of maintenance diagnosis is significantly increased. There was a problem of lowering it.

The present invention eliminates the above problems, prevents the abnormality of the abnormality processing system itself, and always notifies the abnormality of the diagnosed side accurately, thereby improving the reliability of the communication function maintenance diagnosis. It is an object of the present invention to provide a communication function maintenance diagnosis method that can be performed.

[0009]

SUMMARY OF THE INVENTION A communication function maintenance diagnosis method of the present invention is an abnormality detecting means for detecting an abnormality of each communication device in a node of a network which communicates using a cell composed of fixed length packets, and the abnormality detecting means. Abnormality analysis / notification means for analyzing the detection result of the detection means and notifying the abnormality occurrence state, pseudo failure generation means for generating a pseudo failure state in each communication device of the own node or another node, and the pseudo failure occurrence The failure detecting means determines whether or not the failure detecting operation is performed as expected by the abnormality detecting means, and the pseudo failure occurs in the communication device of the own node or another node periodically or at the time of designation. Is generated, and the failure status of the abnormal processing system is grasped while monitoring the response operation of the abnormal processing system with respect to the pseudo failure.

[0010]

According to the present invention, a pseudo failure condition is set to each communication device in the own node or another node by using a cell including an abnormal condition generated based on an abnormal occurrence pattern prepared in advance, or at a designated time. On the other hand, while inducing, the abnormal processing system in the own node and the other node detects the pseudo failure and performs processing for confirming whether or not the expected operation is normally performed in accordance with the abnormal occurrence pattern at the pseudo failure.

According to this processing, when the expected operation is not normally performed, it can be recognized that the abnormal processing system has malfunctioned. In this case, it is possible to always maintain the abnormal processing system in a normal state by finding and repairing a defective part in advance, and under the stable operation of the abnormal processing system, it is impossible to predict when the detected part will occur. Even in the case of the above abnormality, the abnormality can be accurately detected and accurately notified to the operator or the upper CPU.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 is a block diagram showing a main configuration of a communication system to which a communication function maintenance diagnosis method of the present invention is applied. In FIG. 1, 1-1 to 1-m are terminals, 2 is a terminal interface, 3 is a cell sending unit, 4 is a cell receiving unit, 5 is a line interface, 6-1 to 6-n is a line, and 7 is abnormality detection. Section, 8 is an abnormality analysis / notification section, 9 is a maintenance interface,
Reference numeral 10 is an abnormal pattern generator.

In this system, as software,
At least an abnormality occurrence pattern library group 91 and a failure database 81 are prepared. The abnormality occurrence pattern library group 91 is registered with control information for pseudo occurrence of a desired failure as an abnormality occurrence pattern corresponding to the hardware in the system. As the specific control information, for example, information in which cell header information such as a sequence number, information length, or segment type is abnormal can be considered. In addition, the failure database 81 registers failure data used as a comparison target of the abnormality occurrence pattern when determining whether or not the pseudo failure generated based on the abnormality occurrence pattern is as expected. Is.

Normal data cell communication in this communication system is performed as follows. For example, a data cell transmitted from the terminal 1-1m is converted into a fixed-length packet by the cell transmission unit 3 via the terminal interface 2 and then transmitted from the line 6-1 to the destination node via the line interface 5. It Also, for example, the cell addressed to the terminal 1-m in the own node is the line 6-n to the line interface 5
Is received by the cell receiving unit 4 through the terminal interface 2 and is further delivered as received data to the terminal 1-m through the terminal interface 2.

In the communication system of the present invention having such a communication function, when an abnormality occurs in a certain communication device in the system, the abnormality detecting section 7 detects the abnormality,
After the abnormality analysis / notification unit 8 analyzes the abnormality, the analysis result is notified to the maintenance interface 9. The operator is
A normal communication function can be maintained by recognizing the abnormality of the communication device by the notification and dealing with the abnormality recovery.

In this maintenance diagnostic method, it is needless to say that when the abnormality of the communication device occurs, if the abnormality analysis / notification unit 8 itself has a failure, the failure of the communication device cannot be correctly notified. . As a countermeasure against this point, in the present invention, an instruction for maintenance diagnosis is given periodically or at a predetermined time, a pseudo fault is caused according to this instruction, and the response operation of the abnormality analysis / notification unit 8 to this pseudo fault is monitored. Meanwhile, it is diagnosed whether the abnormality analysis / notification unit 8 itself operates properly.

Here, as a specific operation for generating the pseudo fault, for example, when the operator gives an instruction for maintenance diagnosis of the communication function, the maintenance interface 9
Reads out the abnormality occurrence pattern corresponding to the above-mentioned instruction from the abnormality occurrence pattern library group 91, and then, according to this abnormality occurrence pattern, the abnormality pattern generation unit 10 determines that the cell header information such as the sequence number, the information length or the segment type is abnormal. An abnormal cell is generated, and the target communication device is controlled so as to be in a pseudo failure state using this abnormal cell.

Further, as a concrete example of the corresponding operation of the abnormality processing system thereafter, the abnormality detecting section 7 detects the pseudo fault of the communication device generated as described above, and then the abnormality analysis /
The notification unit 8 compares the detection result with a failure database registered in advance, and determines whether or not the abnormality processing is correctly performed according to the comparison result. By this control, the operator can always notify the maintenance interface 9 of the failure of the communication device accurately by maintaining the abnormality analysis / notification unit 8 in the normal operation state based on the inspection result.

In the present invention, a pseudo fault is generated periodically or at a desired time, and a method of diagnosing whether or not the pseudo fault matches an expected one is, for example, between a diagnosis instruction node and a diagnosis point node. Two methods are conceivable: a method of performing the diagnosis and a method of performing the diagnosis by itself.

According to the former method, the abnormal pattern generated from the abnormal pattern library group 91 is notified to the abnormal pattern generating unit 10 through the maintenance interface 9 in accordance with the maintenance diagnosis instruction from the operator or the upper processing unit, After the abnormal cell corresponding to the abnormal pattern (the cell header information such as the sequence number, the information length or the segment type is abnormal) is inserted into the transmission information by the cell transmission unit 3, the node at the diagnostic point is connected through the line 6. Send out.

On the other hand, in the latter method, after the abnormal pattern generation unit 10 generates an abnormal cell based on the above-mentioned abnormality occurrence pattern selected from the abnormality occurrence pattern library group 91, the abnormal cell is combined with loopback or the like. Then, the pseudo failure is simulated by causing the error to flow back in the own node, and it is confirmed whether the abnormality handling mechanism (abnormality analysis / notification unit 8 or the like) of the own node device operates properly.

As described above, according to the present invention, a desired abnormality occurrence pattern (making the cell header information such as the sequence number, the information length or the segment type abnormal) can be selected according to the instruction of the operator or the upper CPU, An abnormal cell generated by this abnormality occurrence pattern is sent to a specific path periodically or at a designated time, so that the hardware part that controls the communication function of transferring cells to a logical path or a physical line Anomalies such as failure information that is notified when a malfunction or network failure occurs are generated randomly and artificially.On the other hand, the error handling system side predicts the behavior and actual behavior predicted from the failure occurrence pattern selected for this pseudo failure occurrence. The failure status of the abnormal processing system itself is diagnosed by comparing the failure occurrence states of the above.

More detailed processing of the maintenance diagnosis system of the present invention will be described with reference to FIGS. First, FIG.
(A) shows an example of maintenance diagnosis across nodes.
In the case of this example, a specific communication path or a physical line itself is designated, and the diagnostic instruction node (# 0) issues a maintenance diagnostic instruction of the communication function to the diagnostic point node (# 2) in advance. After that, the cell including the abnormality is sent from the diagnosis instruction node (# 0) to the diagnosis point node (# 2) by the method described above. At the diagnostic point node (# 2), the abnormal cell is received by the cell receiving unit 4 through the line interface 5.

At the time of this reception, when the abnormality detecting unit 7 detects an abnormal cell, the abnormality analyzing / notifying unit 8 then responds to the abnormality occurrence pattern notified from the diagnosis instruction node (# 0) in advance based on the detection result. Determine if the anomaly occurred as expected. It should be noted that this determination is performed by a fault database 81 that is a database of faults that should occur in correspondence with the fault occurrence patterns in the fault occurrence pattern library group 91, and the above-mentioned fault occurrence that is notified in advance from the diagnostic instruction mode (# 0). It can be determined by comparing the pattern and whether or not they match.

FIG. 2B shows an example of so-called maintenance diagnosis of the self node in which the diagnosis instruction node and the diagnosis point node are the same. In this case, the abnormality pattern generation unit 10 simulates an abnormality in its own node according to the abnormality generation pattern selected by the operator or the instruction from the higher-level processing unit and causes a pseudo fault to occur in each communication device. After that, the abnormality detection unit 7 detects an abnormality from each communication device in its own node by the same determination processing as that of the diagnosis point node (# 2). Then, based on the detection result, the abnormality analysis / notification unit 8 determines whether or not the abnormality expected from the abnormality generation pattern is correctly activated, while comparing the abnormality generation pattern with the failure database 81.

Next, the diagnosis sequence when the diagnosis point node shown in FIG. 2A is not the self node will be described in detail with reference to FIG. First, in the diagnosis across nodes, the diagnosis instruction node (# 0) selects an abnormality occurrence pattern corresponding to an instruction from the higher-level processing unit or the like from the abnormality occurrence pattern library group 91 (pattern types are A, B, ... ... X), and notifies the diagnosis point node (# 2) of the maintenance diagnosis instruction request primitive (S1) having the pattern type as additional information.

On the other hand, the designated diagnostic point node (#
In 2), if the self node is operable, the maintenance diagnosis instruction response primitive (S2) is sent to the diagnosis instruction node (# 0). At this point, the diagnosis instruction node (# 0) follows the diagnosis point node (# 0) according to the selected abnormality occurrence pattern.
Send the abnormal cell (S3) to 2). On the other hand, the diagnosis point node (# 2) determines whether or not a failure expected in advance occurs according to the received abnormal cell in the failure database 81.
(The database type is indicated by a, b ... x)
Judge by comparing with.

If an abnormality detection result that matches the prediction is obtained, the diagnosis is passed and the diagnostic result response primitive (S4) with the notification content "normal" is sent to the diagnostic instruction node (# 0). If the error occurs even once or if the error exceeds a certain threshold value, it is considered that a failure of the abnormal processing system has occurred, and the content of the diagnostic result response primitive (S4) is set as "abnormal". , And notifies the maintenance interface 9 of its own node or the diagnostic point node (# 2), etc., and entrusts the higher-level processing.

[0029]

As described above, according to the present invention,
Periodically or at the time of designation, a cell containing an abnormal state is used to artificially generate a failure in the communication device of its own node or another node, and it is checked whether the abnormal processing system operates normally against this pseudo failure. Therefore, when the abnormal processing system malfunctions, it is possible to repair the defective part in advance and keep the abnormal processing system in a stable operating state. It is possible to perform highly reliable maintenance diagnosis while avoiding the situation where the abnormality cannot be notified due to the failure of the system itself.

[Brief description of drawings]

FIG. 1 is a block diagram showing a main configuration of a communication system to which a communication function maintenance diagnosis method of the present invention is applied.

2 is a conceptual diagram showing two application examples of the communication function maintenance diagnosis method of the present invention in the communication system shown in FIG.

FIG. 3 is a sequence control diagram according to a maintenance diagnosis method that spans nodes among the methods shown in FIG.

FIG. 4 is a conceptual diagram showing an example of an abnormality detection notification mechanism according to a conventional maintenance diagnosis method using interrupt control.

FIG. 5 is a conceptual diagram showing an example of an abnormality detection notification mechanism according to a conventional maintenance diagnosis method using polling.

[Explanation of symbols]

 1-1 to 1-m terminal 2 terminal interface 3 cell transmitting unit 4 cell receiving unit 5 line interface 6-1 to 6-n line 7 abnormality detection unit 8 abnormality analysis / notification unit 81 fault database 9 maintenance interface 91 abnormality occurrence pattern Library group 10 Abnormal pattern generation unit 101 Abnormality detection unit 102 Interrupt control unit 103 Display unit 110 CPU

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical indication location H04L 29/14 9371-5K H04L 13/00 313

Claims (5)

[Claims] 1. An abnormality detecting means for detecting an abnormality of each communication device in a node of a network which communicates using a cell composed of fixed-length packets, and analyzes a detection result of the abnormality detecting means to notify an abnormality occurrence state. Abnormality analysis / notification means, pseudo failure generation means for generating a pseudo failure state in each communication device of its own node or another node, and the abnormality detection means predicts the pseudo failure by the pseudo failure generation means. It is equipped with a judgment means for judging whether or not the failure detection operation has been carried out,
Communication function maintenance characterized by generating a pseudo fault in the communication device of its own node or another node, and grasping the fault status of the abnormal processing system while monitoring the response operation of the abnormal processing system to the pseudo fault Diagnostic method. 2. An abnormality occurrence pattern library group, which is a library of abnormality occurrence patterns, is registered in advance, and the pseudo-fault occurrence means causes an abnormality based on a designated abnormality pattern in the abnormality occurrence pattern library group. 2. The communication function maintenance diagnosis method according to claim 1, wherein a cell including a state is generated, and the pseudo fault state is generated by circulating the abnormal cell in its own node or transmitting it to another node. 3. The abnormality occurrence pattern has cell header information such as a sequence number, an information length or a segment type of an abnormal cell to be generated by the abnormality occurrence pattern as at least its information element. Communication function maintenance diagnosis method. 4. A failure database, which is a database of failure situations predicted for each abnormality occurrence pattern, is registered in advance, and the judgment means is responsive to a comparison result between a pseudo failure actually detected and the failure database. 3. The communication function maintenance diagnosis method according to claim 1, wherein it is determined whether or not the failure detection operation of the abnormality detection means is as predicted. 5. When causing a pseudo fault to another node, the fault occurrence pattern used for inducing the pseudo fault prior to the abnormal cell is transmitted from the pseudo fault generating means of the own node to the other node. The communication function maintenance diagnosis method according to claim 4, wherein