JPH09284281A - Network monitor system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the network monitor system in which a fault location is quickly analyzed and located by using lots of alarms simultaneously from a communication equipment. SOLUTION: In the network monitor system provided with plural communication equipments (i) (i=1-N) and a network monitor 10 monitoring the operating state of the plural communication equipments, a communication line connecting the plural communication equipments has plural communication layers. When an optional communication equipment or communication line is faulty, the communication equipment that is faulty or affected by a fault informs a fault end point number of plural communication layers to the network monitor. The network monitor 10 definitely specifies a fault line number based on a cross reference table stored in an associate memory (CAM 16) storing the cross reference table between combinations of fault end point numbers and fault line numbers based on the fault end point number informed from the communication equipment.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a network monitoring system capable of analyzing and specifying a failure location by using failure alarms of various communication layers from a plurality of communication devices,
In particular, the present invention relates to a network monitoring system that uses a associative memory to quickly identify a failure point.

[0002]

2. Description of the Related Art Conventionally, a system shown in FIG. 5 is generally known as a system for monitoring the operating state of a communication device and a communication line of each communication layer terminated by the communication device. Explaining the schematic operation of FIG. 5, N communication devices 1 to
N are connected to each other via a communication line 62, and are also connected to a network monitoring device 50 via a line 61 for information transfer, and when a communication failure occurs, any target communication device i ( A fault alarm is notified to the network monitoring device 50 from i = 1 to N). The network monitoring device 50 has a processor (CPU) 51, a main memory (RAM) 52, an I / O port 53 related to connection with each communication device, a data bus 54, and an address bus 55 inside, and the communication device i ( i = 1 to
The failure point is identified based on the failure alarm notified from N).

Now, a conventional method for identifying a failure point in a communication network composed of a plurality of communication devices and a plurality of communication lines will be described in detail. FIG. 2 shows an example of communication lines connecting a plurality of communication devices and a termination example of the communication lines. More specifically, four communication devices 1 to 1 are shown.
4 and three communication lines of layer 1 (# 1-1 to 1-3) and two communication lines of layer 2 (# 2-1 to # 2-2). Here, the layer in the communication line is, for example, layer 1 is a physical line and layer 2 is an AT.
M (Asynchronous Transfer Mode) VP (Virtual
Path) line is a layer on the communication protocol, and a plurality of layer 2 VP lines can communicate on the layer 1 physical line. In existing communication networks,
Since the telephone system is mainly used as a communication terminal, it was possible to deal with the network configuration using only physical lines, but in the future, the introduction of the ATM system technology that has a virtual path layer and a virtual channel layer in the communication network for making the terminal multimedia. Is indispensable, and in this case it is clear that the number of layers of communication lines in a communication network will increase.

A failure example of a communication line in a network configuration having a plurality of communication layers will be described below. In FIG. 2, for example, when the layer 1 communication line # 1-1 fails,
The termination points # 1-11 and 1-12 of the line are in a failure state. FIG.
The communication network example has two layers, layer 1 and layer 2, and when the communication line of the lower layer, layer 1, fails, all the communication lines of the upper layer, layer 2, fail. . This is because the upper layer can be established only when the lower layer is functioning normally. This is also clear from the fact that when bit transfer cannot be performed in layer 1, information transfer in layer 2 cannot be performed. Therefore, layer 1
The two communication lines # 2-1, 2-2 of the upper layer, which is the upper layer, also fail due to the failure of the communication line # 1-1 of the communication line # 1-1. -12, and communication line # 2-2
The terminal points # 2-21 and 2-22 of are also in a failure state. This allows
A total of 6 terminal points # 1-11, 2-11, 2-21 of communication apparatus 1, terminal points # 1-12 of communication apparatus 2, and terminal points # 2-12, 2-22 of communication apparatus 4
A fault alarm at one end point can be sent to the network monitoring device 50.

In the conventional network monitoring system, the fault line number is specified by the following method using these fault alarms. The network monitoring device 50 holds the relationship between layers and the relationship between termination points as a database (not shown: may be built in a memory in the network monitoring device or externally provided). Incidentally, when a failure alarm is received from the above six termination points,
First, the terminal point of the lowest layer among the layers that have received a failure alarm in each communication device is specified. In the present case, the lowest layer among the layers that have received the failure alarm in each communication device is the layer 1 in the communication device 1 and the communication device 2,
It is layer 2 in the communication device 4. Therefore, in the communication device 1, the termination point # 1-11 of layer 1, in the communication device 2 the termination point # 1-12 of layer 1, and in the communication device 4 the termination point # 2-12 of layer 2.
And specify # 2-22.

After that, the communication line number having the termination point is further specified, and the lowest layer among them is the cause of the failure. In the present case, the above-mentioned four termination points # 1-11,
From # 1-12, # 2-12, # 2-22, we can see that the three lines # 1-1, # 2-1, # 2-2 that have them as termination points are out of order. It is possible to specify that # 1-1, which is the lowest layer line, is the cause of the failure. In the conventional technology, the above-described specific processing is performed by making full use of the database to form a logic by software.

[0007]

However, in the above-mentioned prior art, since the process of specifying the faulty line is performed only by software, it takes time to specify the faulty point for which immediateness is required for maintenance and operation. It was a big problem in maintaining and operating the communication network. For example, the processor (CPU) 51 has a processing capability of 1 MIPS,
If 1k Step is required for the failure location identification process,
Only by processing the above six fault alarms, the fault location can be specified in about 6 ms. But,
Assuming an ATM-based communication network that will increase in the future, a maximum of 4096 VP lines may be accommodated in one physical line, in which case 8,000 failure alarms will be sent simultaneously to the network monitoring device. Therefore, it takes about 8 seconds to specify the failure location.

[0008] Furthermore, a large-scale communication network with several tens of communication devices, or one VP line with up to 650,000 V lines
Considering the case of accommodating a C (Virtual Connection) line, the conventional method described above has a problem in that the time required to specify a failure point exceeds a practical range. An object of the present invention is to solve the above-mentioned problems in the prior art, and to provide a network monitoring system capable of quickly analyzing and specifying a failure location by using a large number of failure alarms simultaneously notified from a communication device. To provide.

[0009]

In order to achieve the above object, a network monitoring system of the present invention comprises: (A) a plurality of communication devices i (i = 1 to N) and operating states of the plurality of communication devices. In a network monitoring system comprising a network monitoring device (10) for monitoring the network monitoring device, and a communication line (22) connecting the network monitoring device (10) to the plurality of communication devices and to the plurality of communication devices, The communication line (22) that connects the communication devices to each other has a plurality of communication layers, and when any communication device or communication line fails, the communication device affected by the failure is The network monitoring device (10) is notified of the failure termination point numbers of the plurality of communication layers, and the network monitoring device (10)
An associative memory that stores a correspondence table of a combination of a failure termination point number and a failure line number, and a correspondence table stored in the associative memory (CAM16) based on the failure termination point number notified from the communication device ( It is characterized in that it has means for uniquely specifying the faulty line number from (see FIG. 3). Also,

(B) The means for uniquely specifying the faulty circuit number refers to the correspondence table (FIG. 3), and the faulty circuit number corresponding to the combination of the fault termination point numbers notified from the communication device. When there is no faulty line number corresponding to the combination of faulty termination point numbers notified from the communication device (step 43 in FIG. 4; NO), only the uppermost layer portion of the correspondence table is specified. And the search process is performed again (steps 45, 46 in the same step), and the search process is repeated by sequentially lowering the masking layers one by one until the faulty line number can be specified (steps 47, 48, 49, 46). ), It is characterized by comprising means for estimating a faulty line even for a combination of faulty termination point numbers which is not assumed.

[0011]

BEST MODE FOR CARRYING OUT THE INVENTION According to the present invention, a network monitoring apparatus is provided with an associative memory (CAM), and a communication apparatus notifies the network monitoring apparatus in advance of the content addressable memory (CAM). A correspondence table between the failure termination point number and the failure line number is stored, and the failure alarm sent to the network monitoring device when any communication device or communication line fails causes the above-mentioned problem in the associative memory (CAM). The correspondence table is searched to instantly identify the faulty communication line number.

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of a network monitoring system showing an embodiment of the present invention, in which a processor (CPU) 11, a main memory (RAM) 12 and an I / O related to connection with a communication device are provided inside the network monitoring device 10. O port 13, data bus 1
4, an address bus 15, communication devices 1 to N are mutually connected via a communication line 22, and a line 2 for information transfer is provided.
5 is connected to the network monitoring device 10 via
It is the same as the conventional configuration. The present embodiment is characterized in that an associative memory (CAM) 16 is provided inside the network monitoring device 10. The associative memory (CAM) 16 and the processor (CPU) 11 include a data transfer line 17 and a control line 1.
8 are connected. And associative memory (CAM) 16
The correspondence table of the fault termination point numbers and the fault communication line numbers notified from the communication devices 1 to N to the network monitoring device 10 is written therein. When a failure alarm is received, the failure point is instantly specified by searching the corresponding failure communication line number from the correspondence table using these failure termination point numbers.

FIG. 3 is an example of a correspondence table of the fault termination point numbers and the fault communication line numbers stored in the associative memory (CAM) 16. In the figure, ○ in the failure end point number column
Means failure, and blank means normal. Therefore, in the failure pattern number # 1, the failure end point numbers are # 1-11, # 1-12, #
If the termination points of 2-11, # 2-12, # 2-21, # 2-22 are faulty, the failure of the communication line of the faulty communication line number # 1-1 and the failure pattern The number # 2 means that the communication line with the faulty communication line number # 2-1 has failed if the termination points with the faulty termination point numbers # 2-11 and # 2-12 are faulty. .

Next, a specific fault location specifying process will be described with reference to FIGS. In the communication network configuration example of FIG. 2 described above, for example, when a failure occurs in the layer 1 communication line # 1-1, the termination point # 1-1 of the communication line # 1-1.
1 and # 1-12 are in a failure state. Also, because of this failure, the two communication lines # 2-1 and # 2-2 of the layer 2 which is the upper layer of the failure also fail, so the termination point # of the communication line # 2-1 #
2-11, # 2-12, and communication line # 2-2 termination points # 2-21, # 2
-22 is also in a failure state. Therefore, the communication device 1 gives a failure alarm to the termination points # 1-11, # 2-11, # 2-21, the communication device 2 gives a failure alarm to the termination point # 1-12, and the communication device 4 gives a termination alarm # 2-12. , # 2
-22 failure alarms, that is, a total of 6 failure alarms, are sent to the network monitoring device 1 via the information transfer line 21 of FIG.
Sent to 0.

In the network monitoring device 10, these six failure alarms are received by the I / O port 13 and then transferred to the processor (CPU) 11. In this embodiment, the information included in the failure alarm includes "failure termination point number", "failure cause", "importance", etc., but the processor (CPU)
In FIG. 11, only the “fault termination point number” is extracted from these and the associative memory (CA) is extracted via the data transfer line 17 of FIG.
M) Send to 16 In the present case, associative memory (CAM)
The fault termination point numbers sent to 16 are # 1-11, # 2-11, # 2-
There are six, 21, # 1-12, # 2-12, and # 2-22. At the same time, the processor (CPU) 11 has an associative memory (CAM).
“Search” is designated as the operation mode for 16 via the control line 18. Since the associative memory (CAM) 16 stores a correspondence table (relationship between failure termination point numbers and failed communication line numbers) as shown in FIG. 3 in the present embodiment, failures transferred from the processor (CPU) 11 are stored. Terminal point number sequence (# 1-
11, # 2-11, # 2-21, # 1-12, # 2-12, # 2-22) corresponded to the top row of the correspondence table. It is possible to instantly and uniquely identify a certain thing with one search. Therefore, assuming that the search time of the associative memory (CAM) 16 is about several tens of ns per time, the specification of the failure point according to this embodiment is completed in about several tens of ns.

In FIG. 3, ten end points (# 1-11, # 1-12, # 1-21, # 1-22, # 1-31, # 1) corresponding to the end points in FIG. -3
2, # 2-11, # 2-12, # 2-21, # 2-22) is shown. Generally, if 10 termination points independently fail, there are 2 to 10 10 types of failure patterns, and an associative memory (CAM) capable of storing such failure patterns is not possible at present and the realization of this method is not possible. It may seem difficult, but in reality it can be achieved due to the following circumstances.

That is, in general, a plurality of communication lines in the same layer independently fail, but there is some regularity between layers. For example, if a communication line of a certain physical layer is out of order, a maximum of 4
All 096 VP lines will also be out of order. As described above, since the occurrence of failures is regular, the associative memory (CA
The number of failure patterns stored in M) is greatly limited.
If only the limited (presumed) failure patterns are stored in this way, the capacity of the current associative memory (CAM) can be practically used. Most of the failure alarm patterns notified from the communication device are associative memory (CA).
It should match the expected failure pattern stored in M), but of course, it may also occur.

Next, a method of coping with the case where the failure alarm pattern notified from the communication device does not match the assumed failure alarm pattern will be described. If the failure alarm pattern notified from the communication device is different from the expected one for some reason, the associative memory (CAM) may not be able to identify the failed communication line. In such a case, only the uppermost layer portion of the failure pattern is masked and the associative memory (CAM) is searched again, and if it cannot be further specified, the layer to be masked again is lowered by one and the search is performed.
By repeating this work, it is possible to deal with an unexpected failure pattern to some extent.

A coping method in this case will be described with reference to the flowchart of FIG. First, a failure alarm reception process from the communication device is performed (step 41), and the failure patterns of all layers stored in the associative memory (CAM) correspondence table are searched by the failure termination point number included in the failure alarm (step 41). 42). As a result of the search, if there is a matched failure pattern (step 43; YES), the failed communication line number is specified, and the process ends (step 44). The above is exactly the same as the normal processing described above.

If the matched failure pattern does not exist in the associative memory (CAM) correspondence table (step 43;
NO), only the uppermost layer portion of the failure pattern is masked (step 45), and the retrieval process by the associative memory (CAM) is performed again (step 46). If a matching failure pattern is found in the correspondence table as a result of the search (step 47; YE
S), thereby identifying the faulty communication line number and ending the processing (step 44).

If the matching failure pattern does not exist in the correspondence table (step 47; NO), it is judged whether or not the failure pattern is further masked (step 4).
8), if you lower one layer and mask
If there is no failure pattern stored in advance in the associative memory (CAM) (step 48; NO), the process is terminated. That is, for example, if there are 10 layers and only failure patterns up to layer 3 are stored in advance in the associative memory (CAM), masking up to layer 4 is performed, and if there is no matching pattern, Even if the search is further performed, it is determined that the failure pattern does not exist, and the processing ends. When the masking process is further performed (step 48; YES), the layer to be masked is further lowered by one (step 49), and then the process returns to step 46 again to repeat the above process.

Next, a specific example thereof will be described with reference to FIGS. 3 and 4. In FIG. 3, it is assumed that the network monitoring system has received a failure alarm including failure termination point numbers # 1-11, # 1-12, # 2-11 from the communication device (step 41).
Since the termination points # 1-11, 1-12 of layer 1 should be faulty, all termination points should be faulty in layer 2, but in this case, only termination point # 2-11 of layer 2 is It indicates that it is out of order. Therefore, this failure pattern cannot actually exist, and there is a possibility that an error will occur as a failure pattern that does not exist in the above method (steps 42 and 43; NO). Therefore, in the present embodiment, as described above, the fault termination point number of Layer 2, which is the highest layer, that is, # 2-11 in this case is masked (step 45).

In this case, the failure pattern of only layer 1, the failure pattern of layers 1 and 2, ..., The failure patterns of layers 1 to N-1 are additionally stored in the associative memory (CAM). I'll do it. In the case of only two layers as shown in FIG. 2, the failure pattern of only layer 1 is stored. Then, the layer 2 failure termination point number # 2-11
Is masked (step 45), and a search is performed using the fault termination point numbers # 1-11, 1-12 only in layer 1 (step 46), and it is highly possible that the communication line number # 1-1 is faulty. Can be estimated (step 47; YES). Note that the associative memory (CA
When the correspondence table of M) is searched, if the failure terminal point number of the layer higher than the layer n of the correspondence table is ignored,
As described above, the failure pattern of layer 1 only, layer 1
Failure pattern consisting of 2 and ..., Layers 1 to N-1
It is not necessary to additionally store failure patterns consisting of.
In this way, it is possible to estimate the cause of the failure to some extent by making the above measures for the failure pattern that does not normally occur (when the number of communication layers is three or more, step 48, step 49, step 46). ，
Repeat step 47 until the determination in step 47 is YES).

As described above, according to the present embodiment,
An associative memory (CAM) is provided inside the network monitoring device, and a failure line number can be instantly searched by previously writing a correspondence table of the failure alarm notification pattern and the failure line number in the associative memory (CAM). Therefore, it is possible to significantly reduce the time required to identify the failure point as compared with the conventional technique. Specifically, assuming that the search time of the associative memory (CAM) is about tens of ns per time, the specification of the failure point according to the present embodiment is completed in only about tens of ns, which is 6 ms of the conventional technique.
Compared with, the time can be shortened by about 2 digits. Also, assuming an ATM communication network, a maximum of 40 per physical line
96 VP lines may be accommodated, and as many as 8,000 fault alerts will be sent to the network monitoring system at the same time.
In the conventional technology, it takes about 8 seconds to specify the failure point, but in the present embodiment, the capacity of the associative memory (CAM) is 4 seconds.
If k words x 84 bits, these 8,000 fault alarms are also completed by one search, so the fault location can be specified in several tens of ns, and the time can be shortened by about 5 digits compared to the conventional technology. it can. Furthermore, the present invention is even more effective when considering a large-scale communication network having several tens of communication devices or accommodating as many as 650,000 VC lines in one VP line. There is no end.

[0025]

According to the present invention, even if a large number of failure alarms are simultaneously notified from a plurality of communication devices constituting a network, the failure alarms can be used to quickly analyze and identify the failure location. It is possible to realize a network monitoring system that can

[Brief description of drawings]

FIG. 1 is a configuration diagram of a network monitoring system showing an embodiment of the present invention.

FIG. 2 is a diagram illustrating a plurality of communication devices, a communication line example connecting the communication devices, and a communication line termination point example.

FIG. 3 is a diagram (correspondence table) showing an example of a relationship between a fault termination point number and a fault line number stored in an associative memory.

FIG. 4 is a diagram showing an example of a flowchart for identifying a failure location when an unexpected failure pattern occurs.

FIG. 5 is a diagram showing a configuration example of a conventional network monitoring system.

[Explanation of symbols]

10, 50: Network monitoring device, 11, 51: Processor (CPU) of network monitoring device, 12, 52: Main memory (RAM) of network monitoring device, 13, 53: I / O port related to connection with communication device , 14, 54: a data bus for transferring data between respective constituent elements in the network monitoring device, 15, 55: an address bus for transferring an address necessary for data transfer, 1
6: Associative memory (CAM), 17: Processor (CP)
U) and associative memory (CAM) data transfer line, 18:
A control line for specifying the operation mode of the associative memory (CAM),
21, 61: A line for transferring information between the network monitoring device and the communication device, 22, 62: A communication line for communicating between the communication devices

Claims (2)

[Claims] 1. A plurality of communication devices, a network monitoring device that monitors operating states of the plurality of communication devices, and communication that connects the network monitoring device and the plurality of communication devices and between the plurality of communication devices. In a network monitoring system including a line, the communication line connecting the communication devices to each other has a plurality of communication layers, and when any communication device or communication line fails, the communication is in failure or due to a failure. The affected communication device notifies the network monitoring device of the fault termination point numbers of the plurality of communication layers, and the network monitoring device stores the combination of the fault termination point numbers and the correspondence table of the fault circuit numbers. It is characterized by having a memory and means for uniquely specifying the failure line number from the correspondence table stored in the associative memory based on the failure termination point number notified from the communication device. Web monitoring system to collect. 2. The means for uniquely identifying the faulty line number refers to the correspondence table and identifies the faulty line number corresponding to the combination of the fault termination point numbers notified from the communication device. When there is no faulty line number corresponding to the combination of faulty termination point numbers notified from the communication device, only the uppermost layer portion of the correspondence table is masked and the search process is performed again. 2. A means for estimating a faulty line even for an unexpected combination of fault termination point numbers by successively lowering masking layers one by one until it can be specified, and repeating the search process. The described network monitoring system.