JP2001251300A - Network maintenance system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem where realization of simplification or efficiency of procedures or actions, until a problem is solved due to the analysis of failure according to a situation at the time of the generation of the failure or a situation after the occurrence of the failure by a maintenance person, when the failure is generated due to any factor whose analysis is necessary in network maintenance. SOLUTION: This system is provided with an individual network maintenance part 100, arranged for each network equipment constituting a network and a network maintenance part 200 connected via a network 920 to plural sub- networks including the individual network maintenance part 100. Then, a maintenance even is collated in advance with the past data, and whether the similar failure is generated in a related network in the past is confirmed. Also, information enough to be recognized based on the operation of an agent is always stored in a maintenance record data base memory 207 in the network maintenance part 200, and the contents are compared by an event information part 206.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a network maintenance system, and more particularly, to a network maintenance system for monitoring the status of network devices, transitioning the status of the network devices, and grasping fault information that needs to be analyzed manually.

[0002]

2. Description of the Related Art The prior art of such a network maintenance system or related technology is described in, for example, "Failure Judgment Method Using Pattern Bitmap" in JP-A-3-153140 and "Communication Network Configuration" in JP-A-6-237250. A failure countermeasure determining device "," A power system monitoring and controlling device and a recording medium storing the program "in JP-A-11-122815, and" Remote maintenance device "in JP-A-10-301810.

FIG. 1 shows an example of a conventional network maintenance system.
It is shown in FIG. In a network group (SUB01) composed of a lower layer network and an upper layer network,
Consider the communication between A and B. The trunk line of the upper layer network mainly used for communication between A and B is NW1-N
When W2 is set, it is assumed that an instantaneous interruption (failure) has occurred. Transmission / reception data is congested on the trunk lines NWa, NW1, NW2, and NWb, and returns to the original state when recovered. In the case of this example, it is assumed that the line card of NW2 for NW1 has failed, and congestion → recovery → congestion → recovery → congestion → recovery → line disconnection.
It recovers by detouring using W4. afterwards,
A maintenance person obtains information contents from MATa, analyzes the state of NW1 and NW2 using his experience and knowledge, and finally replaces the line card of the problematic NW2.

FIG. 16 shows an example of remote maintenance according to the prior art. According to this conventional technique, a plurality of network groups SUB01 to SUB4 are connected to the MAT to collect all maintenance information, and a MAT supervising this maintenance has a maintainer. The failure information that has occurred in SUB01 is MAT
It is assumed that they will gather at the MAT via a. In this example, it is assumed that a similar failure has occurred in the network of SUB03 in the past. It is the maintenance person who determines that the causes of the network are very similar and that the recovery measures are the same, and if there is a time difference, it is necessary to organize the records. Therefore, even if the same failure content occurs in the maintenance range, the same planned action is performed from the determination to the recovery action. Although this action is quickly recognized by the database and the experience of the maintainer, the procedure from the judgment to the recovery action is not simplified and efficient.

[0005]

The above-mentioned prior art has the following problems to be solved. First, regarding a maintenance method, when a failure occurs due to a factor that needs to be analyzed, a state at the time of occurrence or a state after the occurrence is regarded as a failure, and a maintenance person subsequently deals with the failure. Therefore, the actual network state at the time of occurrence is determined and derived from the results and contents after occurrence, and it is difficult to quickly resolve the situation.

Next, regarding utilization of maintenance information, a fault that has occurred in another sub-network in the past in the same maintenance range in a network will be treated as a different item even if the same line card fails due to the same fault. Only the knowledge, experience or past reports of the maintainer will recognize that identity. Therefore, there is a problem that the procedure and actions until the problem is solved are not simplified.

[0007]

SUMMARY OF THE INVENTION Accordingly, one of the objects of the present invention is to provide a network maintenance method for quickly solving a network failure when a failure occurs due to a factor requiring analysis. Another object of the present invention is to provide a network maintenance method that simplifies a procedure and an action to solve a problem in a network having the same maintenance range.

[0008]

A network maintenance method according to the present invention is for maintaining a network to which a plurality of network devices are connected. When a failure occurs due to a factor requiring analysis, the failure occurs. It is also characterized in that an individual network maintenance unit is provided for each network device, which also grasps the state after the occurrence of the failure or the state before the occurrence of the failure and the state before the occurrence of the failure, and automatically determines the cause and the remedy before the maintenance person responds.

According to a preferred embodiment of the present invention, there is provided a network maintenance unit which is connected to a plurality of networks including the above-mentioned individual network maintenance unit and transmits / receives data to / from the individual network maintenance unit.
The individual network maintenance unit is characterized by having a timer for measuring a time for capturing data captured from the network device and a maintenance information auxiliary control unit for holding the captured data for a predetermined time and performing an analysis process. Further, the network maintenance unit includes a program control unit connected by a common bus, a maintenance setting database storing a maintenance program from the program control unit, and an event inference unit transmitting and receiving a result of matching and inference of maintenance events to the maintenance program control unit. And a maintenance record database memory for recording and storing maintenance events. The network maintenance unit includes a data transmission / reception buffer, and is capable of transmitting a maintenance program stored in the maintenance setting database to the individual network maintenance unit via the network.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The configuration and operation of a preferred embodiment of a network maintenance system according to the present invention will be described below in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the configuration of a preferred embodiment of a network maintenance system according to the present invention. The individual network maintenance unit 100 includes an input / output checker 101, a timer A102, a device setting checker 103, a flow information memory 104, a maintenance setting storage unit 105, an agent unit 106, a timer Q107, a timer B108, and a network 920 connected to the line side. An input / output checker 109 and a maintenance information auxiliary control unit 110 are provided. An existing network function 900 is connected to the input / output checkers 101 and 109 and the device setting checker 103.

The individual network maintenance unit 100 shown in FIG.
Are the network devices NW1 shown in FIG.
NW2 and NWa are connected. The input / output checkers 101 and 109 of the individual network maintenance unit 100 receive data input / output, respectively. The timer A102 and the timer B108 measure the supplementary time of the above-described data. The device setting checker 103 exchanges network device parameters with the existing network function 900. The maintenance information auxiliary control unit 110 controls each component in the network maintenance unit 100. In the maintenance information auxiliary control unit 110, a timer Q107 used for timing measurement to send the state to the network maintenance unit 200 described later with reference to FIG. 2 and a maintenance setting in which a survey item from the network maintenance unit 200 is set. Storage unit 105, agent unit 106 in which an autonomous investigation program (hereinafter, referred to as an agent) from network maintenance unit 200 is stored.
And a flow information memory 104, which is a memory area where an agent executes data storage and analysis of the input / output checkers 101 and 109 for a certain time.

It is the network maintenance unit 200 shown in FIG. 2 which constitutes a part of the network maintenance unit according to the present invention and establishes the final maintenance information based on the information of the individual network maintenance unit 100. It is a part mounted on 16 MATs. This network maintenance unit 200
Is the maintenance setting receiving unit 201, the maintenance display information buffer 20
2. Report display database 203, report control unit 204, maintenance setting database 205, maintenance record database memory 207, event inference unit 206, maintenance program control unit 208, event detection unit 209, information DM
A (direct memory access) control unit 210, an agent control unit 211, a data transmission / reception buffer 212, and a network flow information memory 213. The maintenance setting database 205, event inference unit 206, maintenance record database memory 207, and maintenance program control unit 208 are connected by a common bus 214. The maintenance setting receiving unit 201 and the maintenance display information buffer 202 store the existing PC (personal computer) function 9
The input / output device 930 and the display device 940 are connected to the existing PC function 910. The data transmission / reception buffer 212 is connected to the network 920.

Here, the individual network maintenance unit 100 shown in FIG. 1 and the network maintenance unit 200 shown in FIG. 2 are connected via a network 920, and mutually exchange information with a data transmission / reception buffer 212. The maintenance setting receiving unit 201 receives a request for a maintenance person from the network maintenance unit 200. The maintenance display information buffer 202 passes the maintenance information. Data from the data transmission / reception buffer 212 is transmitted / received by the information DMA control unit 210, and data to be stored is accumulated in the network flow information memory 213.
If the content of the data is a content related to a maintenance event, the data is transmitted to and received from the event detection unit 209. Information on the agent in the maintenance event is transmitted / received to / from the agent control unit 211 and passed to the event detection unit 209 as special event information. Information DMA controller 210
Is connected to a maintenance program control unit 208 that sends and receives event information control and agent information control. The maintenance program control unit 208 has a program for controlling the maintenance itself, is connected to the maintenance setting reception unit 201 so that a maintenance person can perform the basic setting, and performs a report control for displaying a report based on the setting. Part 204
It is connected to the. The maintenance program uses the common bus 214
Through the maintenance setting database 205.

Similarly, as described above, the common bus 214 has an event inference unit 206 for transmitting and receiving maintenance event matching and inference results to and from the maintenance program control unit 208, and a maintenance record database memory for storing maintenance events as records. 207 are connected. The above-described report control unit 204 is connected to a report display database 203 which records a display method for displaying a report to a maintenance person, and a maintenance display information buffer 202 which is an interface for actually displaying the report.

Next, the operation of the network maintenance system according to the present invention will be described with reference to FIGS. First, the individual network maintenance unit 100 shown in FIG. 1 is mounted on each of network devices such as network devices NW1, NW2 and NWa as shown in FIGS. 15 and 16, and the network maintenance unit 200 shown in FIG. It is mounted on the MAT as shown. Before explaining each function,
Data conversion of maintenance events in the network will be described.
FIG. 3 is an example of abstracting the network of SUB01. If the network shown on the left is divided into constituent elements and described in an abstract manner, the network can be expressed as shown on the right. Furthermore, by converting the abstract description into a pattern as shown in FIG. 4, the data is stored in the form of data. FIG. 5 is a pattern diagram when a failure occurs. In the communication between A and B, 10 between the trunk line systems 1 (circled numerals) and 10 (circled numerals)
It is assumed that the trunk line card on the (circled number) side fails and instantaneous interruption occurs. If an instantaneous interruption occurs, data cannot be transmitted / received in the portion (1), and congestion occurs.

In the instantaneous interruption, when the normal state returns immediately, the phenomenon that the line returns to the normal state in spite of the line failure, and the congestion state is repeated even after a period of time, is repeated. FIG. 6 shows a state in which the instantaneous interruption continues, the operation becomes unstable, and finally, a trunk line failure appears on the upper layer network. Further, FIG. 7 shows a state pattern when a detour route is calculated between each upper-layer network and detour is performed in the existing technology. FIG. 8 shows the state of the temporary recovery as maintenance information together with the information on the occurrence of the failure. In this movement, it is possible to give information to the maintenance person, but the contents indicate that the temporary restoration has been performed and that the cause is a failure of the trunk line card.

Therefore, in the network maintenance system according to the present invention, the maintenance setting is performed from the network maintenance unit 200 to the individual network maintenance unit 100 according to the intention of the maintenance person. This will be described specifically. First, assuming that the maintenance person instructs to perform monitoring and state transition from the input / output device 930, the setting information is input to the maintenance setting receiving unit 201 via the existing PC function 910. The input setting information is transmitted to the maintenance program control unit 208 and stored in the maintenance setting database 205. While referring to the maintenance program in the maintenance setting database 205, the maintenance program control unit 208 instructs the event detection unit 209 to arrange an agent that performs an autonomous investigation on each network.

The main purpose of the autonomous survey at this time is to notify or require items set by the maintenance person or past history. The event detection unit 209 gives an instruction from the maintenance program control unit 208 to the agent control unit 211. The agent control unit 211 has the information DM
The data is transmitted to the network 920 through the existing PC function 910 via the data transmission / reception buffer 212 through the A control unit 210. The transmitted agent enters the determined individual network maintenance unit 100. If sent to a plurality of networks, the above operation will be repeated. Also, when an agent is transmitted, the maintenance program control unit 208 sends an information transmission request to the information DMA control unit 210 to any of the individual network maintenance units 100.
Or send it to Further, the information DMA control unit 210 transmits how much data is input, and accordingly, the information DMA control unit 210
The area of the network flow information memory 213 is secured and the destination of the agent is determined. Individual Network Maintenance Department 1
In 00, all of the transmission / reception data of the existing network function 900 and the maintenance data are stored in the timer A102 and the timer B108.
Is set by the maintenance information auxiliary control unit 110, and the input / output checker 101 performs a check. Checked data is
The maintenance setting storage unit 10 stored in the flow information memory 104
5 periodically deletes the information according to the command of the setting information of the agent staying in the agent unit 106.

Also, a timer Q107 has a regular watchdog timer so as to prevent the occurrence of a failure by itself and affecting higher-level maintenance. When the agent arrives at the input / output checker 101 from the network 920, the information is transmitted to the maintenance information auxiliary control unit 110, and does not pass through the existing network function 900 side.
It is taken into 10. Therefore, maintenance methods and procedures,
The information to be transmitted as the maintenance event is fetched and stored in the agent unit 106.

At this time, the maintenance method and procedure and the information to be transmitted as the maintenance event are all stored in the maintenance setting storage unit 105, and the maintenance information auxiliary control unit 110 stores the data passing through the input / output checker 101 thereafter. Is loaded into the flow information memory 104, and event information that matches the conditions is filtered. When the content that matches the maintenance event is fetched, the content is transmitted to the agent unit 106, and the agent transmits information to the network maintenance unit 200 via the network 920 via the reverse route. When the agent arrives at the data transmission / reception buffer 212 from the network 920 through the existing PC function 910, the information D
The operation enters the MA control section 210. The brought data itself is stored in the network flow information memory 213, and the agent transmits the result to the event detection unit 209 by the agent control unit 211.

The event detection unit 209 transmits the return of the agent to the maintenance program control unit 208. The maintenance program control unit 208 records a set of data from the network flow information memory 213 via the information DMA control unit 210 in the maintenance record database memory 207 via the common bus 214. If the purpose of this agent ends here, information D from the maintenance program control unit 208
Network flow information memory 213 in MA control unit 210
To release the associated memory area. If there is still a schedule for reinvestigation, the agent is sent again to the corresponding individual network maintenance unit 100 by the above-described method.

Here, the agent will be described in detail.
As described above, there are two types: a type for returning a report every time, and a type for sending one's own signature information and data after resident.
In the latter case, no report is sent to the event detection unit 209 and the maintenance program control unit 208 via the agent control unit 211. Only data is recorded in the network flow information memory 213 via the information DMA control unit 210. The time when this recorded information is used is when the agent returns. The information brought back by the agent will be described. In this embodiment, it is assumed that the content of the command of the agent examines the network state of NW1 and NW2 in chronological order. In this case,
An agent is arranged in each individual network maintenance unit 100 of each of the network devices NW1, NW2, NWa and NWb in FIG. 9 and FIGS. 10 to 1
2 are collected and recorded in the network flow information memory 213.

FIG. 13 and FIG. 14 show these in a time-series and simplified manner. When the agent returns, these contents are recorded in the maintenance record database memory 207 in the same manner as described above. Assume that a similar failure has occurred in the past. In this case, even if the network positions are different, the operation of the precursor and the state operation in the time series have the same pattern. The maintenance program control unit 208
Event inference unit 2 always connected to common bus 214
06, the event inference unit 206 determines that the maintenance event has just occurred and the maintenance record database memory 207.
Comparing the maintenance events stored in.

For example, in the case of the specific embodiment of the present invention, as shown in FIG. 13, PT1, PT2, PT3,
When PT4 and the maintenance event change, it can be assumed that a problem has occurred in the line of the upper layer network of the related connection destination. Further, in this series, PT5, PT6, PT5 in FIG.
7. Considering that PT8 is present, it is inferred that PT1 of NW2 is the root cause. In addition, PT
If it is assumed that 1, PT5, PTa, and PTb are repeatedly generated, another subnetwork (for example, S
If a similar failure has occurred in UB03) in the past, the event inference unit 206 determines that if PT1 and PT5 are interconnected networks, there is a possibility that a line disconnection failure will occur due to PT5 and thereafter. Infer. Then, the maintenance program control unit 208 receives this content, and if the content to be dealt with is already in the maintenance setting database 205,
The contents (for example, detour route switching, etc.) are implemented.
If the content to be dealt with is unknown or the result of the deal is reported to the maintenance person as a report, the report control unit 204
Pass information to Then, the report display database 20
3 and also accumulate the report. When the accumulated report content is completed, the report control unit 204 instructs the maintenance display information buffer 202 to display the report, and the existing PC function 9
The information is displayed on the display device 940 or the like via the display 10.

As an application of the network maintenance method according to the present invention, cooperation of agents can be considered. The agent operates the maintenance program control unit 2 as described above.
It is an autonomous survey program issued from 08. Depending on the location, more than one may be sent at the same time. In this case, the maintenance events to be collected given to each agent are expected to include duplication and the same contents, and the resources and efficiency of the network maintenance unit 200 and the individual network maintenance unit 100 may be deteriorated. There is. Therefore, by providing a mechanism for considering a data format in consideration of handling of common data in the maintenance information auxiliary control unit 110 of the individual network maintenance unit 100 and the event detection unit 209 of the network maintenance unit 200, it is possible to realize a cooperative signature of the agent. .

The configuration and operation of the preferred embodiment of the network maintenance system according to the present invention have been described above. However, such example embodiments are merely exemplary of the present invention,
It does not limit the invention in any way. It will be readily apparent to those skilled in the art that various modifications can be made in accordance with the particular application without departing from the spirit of the invention.

[0028]

As is apparent from the above description, the network maintenance system of the present invention has the following practically significant effects. First, when a failure occurs due to a factor that needs to be analyzed, not only the state at the time of occurrence or after the occurrence, but also the state before that is grasped, and the cause and the countermeasure method before the maintenance person takes action Can be automatically determined to quickly resolve the network failure. The reason is that in the network maintenance method according to the present invention, a maintenance event can be checked in advance against past data, and it can be confirmed whether a similar failure has occurred in the related network in the past.

Second, in the case of a network composed of a plurality of sub-networks, in a network having the same maintenance range, a fault that has occurred in another sub-network in the past can be determined by the knowledge, experience, or past Similar to the report in, a means of automatically recognizing and determining its identity,
The steps and actions leading up to problem solving can be simplified.
The reason is that the information that the agent can recognize is
This is because the information is always stored in the maintenance record database memory, and the contents can be compared by the event inference unit.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an individual network maintenance unit according to a network maintenance system of the present invention.

FIG. 2 is a block diagram showing a configuration of a network maintenance unit according to the network maintenance method of the present invention.

FIG. 3 is a diagram showing a network abstraction as a conceptual diagram.

FIG. 4 is a pattern diagram showing the concept of matrixing event information from an abstracted network.

FIG. 5 is a pattern diagram when a failure occurs.

FIG. 6 is a pattern diagram immediately after a failure occurs.

FIG. 7 is a pattern diagram at the time of temporary restoration by detour.

FIG. 8 is a pattern diagram after restoration.

FIG. 9 shows event information when a failure occurs.

FIG. 10 shows event information immediately after a failure occurs.

FIG. 11 shows event information at the time of temporary recovery due to detour.

FIG. 12 shows event information after recovery.

FIG. 13 is a diagram for explaining state transition and agent information management of NW1 events in time series.

FIG. 14 is a diagram for explaining state transition and agent information management of NW2 system events in chronological order.

FIG. 15 is a diagram illustrating an example of network maintenance according to the related art.

FIG. 16 is a diagram showing a network remote maintenance example according to the related art.

[Explanation of symbols]

 Reference Signs List 100 Individual network maintenance unit 101, 109 I / O checker 102, 107, 108 Timer 103 Device setting checker 104 Flow information memory 105 Maintenance setting storage unit 106 Agent unit 110 Maintenance information auxiliary control unit 200 Network maintenance unit 205 Maintenance setting database 206 Event inference Unit 207 maintenance record database memory 208 maintenance program control unit 209 event detection unit 212 data transmission / reception buffer 213 network flow information memory 214 common bus 900 existing network function 910 existing PC function 920 network

 ──────────────────────────────────────────────────続 き Continued on the front page F-term (reference) LL09

Claims (5)

[Claims] In a network maintenance system for maintaining a network to which a plurality of network devices are connected, when a failure occurs due to a factor requiring analysis, a state at or after the failure and a state before the failure occur It also comprises an individual network maintenance unit provided in each of the network devices to automatically determine the cause and remedy before a maintenance person takes action, and solves the network failure. Network maintenance method to use. 2. The network maintenance according to claim 1, further comprising a network maintenance unit connected to a plurality of networks including the individual network maintenance unit and transmitting and receiving data to and from each of the individual network maintenance units. method. 3. The individual network maintenance unit includes a timer for measuring a time for capturing data captured from the network device and a maintenance information auxiliary control unit for holding and analyzing the captured data for a predetermined time. The network maintenance method according to claim 1 or 2, wherein 4. A maintenance program control unit connected by a common bus, a maintenance setting database storing a maintenance program from the maintenance program control unit, and a maintenance program matching and inference result of a maintenance event. 4. The network maintenance system according to claim 1, further comprising an event inference unit to be sent to and received from the unit and a maintenance record database memory for recording and storing the maintenance event. 5. The network maintenance unit includes a data transmission / reception buffer, and sends a maintenance program stored in the maintenance setting database to the individual network maintenance unit via the network. 5. The network maintenance method according to 3 or 4.