JPH07129482A - Network monitoring equipment - Google Patents

Abstract

(57) [Summary] [Purpose] In a network management system, the number of queries for management information is reduced to reduce the communication load and achieve efficient network management. A management target device having a mechanism for reflecting a plurality of management information diagnosis results in one representative management information, and a network management computer 1 capable of transmitting and receiving the management information between the management target device and the management target device. By configuring the network with the above, the communication load is reduced, and it is possible to manage a plurality of management information, thereby improving the efficiency of network management.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer network management system, and more particularly to a network monitoring device for collecting and managing management information held by each computer on a network.

[0002]

2. Description of the Related Art In conventional network management systems,
For example, "TCP / IP and OSI network management"
(Software Research Center, Inc.), each computer on the network independently collects and holds a plurality of network management information.

On the other hand, the computer for network management requests each computer to transmit the individual management information held by the computer and collects the relevant management information.

For example, LAN (Local Area Netwo)
rk) and other TCP / IP protocols (Transmiss
ion Control Protocol / Internet Protocol
In the field of communication using), SNMP (Simple Ne
A protocol for transmitting and receiving network management information called twork Management Protocol) and standard management information MIB-II (Managem) collected by a computer.
ent Information Base) is defined, and the network management computer in the network monitors each computer in the network by making a transmission request for each piece of MIB information possessed by each computer. For some events, each computer has a mechanism in which alarm information called a trap is given to the network management computer.

The management mechanism based on the network management protocol SNMP and the network management information MIB-II, which have been the industry standard for TCP / IP communication, will be described below with reference to the drawings.

FIG. 1 is a block diagram schematically showing a conventional network system to which the present network management system is applied. In FIG. 1, 1 is a computer for network management, 2 is a computer in the same area, and 3 is a computer.
Is a remote computer. The network management computer 1 monitors a computer 2 in the same area connected via a transmission medium 5 and a remote computer 3 connected via a router 4 and a transmission medium 6.

The monitoring is performed by exchanging messages according to the SNMP protocol which is an industry standard protocol (communication procedure) for managing the TCP / IP network and the like. The computer 1 for network management sends inquiry messages (Get-Request) 7, 7'to the computers 2, 3 via the transmission media 5, 6 and sends response messages (Get-Response) 8, 8 '. To receive.

FIG. 2 is a block diagram showing a network management mechanism between the network management computer 1 and the computer 2.

In FIG. 2, the computer 2 manages itself, for example, management information 14 indicating the state of the interface, and management information 1 indicating the total number of frames received.
5, the management information 16 indicating the number of error frames is held in the management information base 13. When the computer 2 receives the inquiry message 7 via the network, the agent process 12 parses this message, extracts the necessary management information from the management information base, and sends this management information to the network management computer 1 by the response message 8. Send to. When the network management computer 1 receives the management information, the network management application 11 analyzes it and manages the network.

[0010]

However, in the conventional network management system, when trying to collect fine-grained management information for each computer in real time, only the number of pieces of management information to be collected for each computer is used. However, there has been a problem that the number of regular inquiries becomes very large because it is necessary to periodically issue a request for transmission of management information. Therefore, in reality, the detailed management information is not regularly collected, and most of the time, it is only to check the active state of each computer.

For example, in the case of MIB-II management information using the SNMP protocol described above, 171 types of management information are defined. Furthermore, when defined in a table format, there may be a plurality of management information of the same type. On the other hand, in the network management system using SNMP, the computer for network management has a mechanism of collecting management information one by one with respect to the above-mentioned individual information. Therefore, when trying to manage a plurality of pieces of management information for each computer, the network management computer has a number of computers per second (the number of computers in the network) x (number of pieces of management information per computer) x 1 / (average inquiry). You will have to make an inquiry about (interval). However, the number of inquiries is limited due to the limitation of the processing capacity of the computer for network management, and it is necessary to reduce the number of management information or set the inquiry interval longer in reality. In addition, such a frequent inquiry puts a heavy communication load on the communication medium connected to the network management computer, and incurs a large communication cost when connected to another area via a public network. There was also the problem that it would be applied.

An object of the present invention is to provide a network monitoring apparatus capable of managing a plurality of pieces of management information while reducing the number of times management information is inquired as described above.

[0013]

According to the present invention, in a network monitoring device of an intra-regional communication network or a wide-area communication network in which a plurality of computer systems are connected by a communication medium, a plurality of pieces of network management information collected in a computer are stored. A plurality of computers having means for processing and retaining as representative management information which is simple single management data, and means for returning the representative management information in response to an inquiry from another computer in the communication network, A network management computer having a means for inquiring representative management information to each computer and a means for analyzing the representative management information, and collecting and analyzing the representative management information of each computer via a communication medium. It is possible to perform the network management operation by Are those that form.

[0014]

According to the present invention, in a plurality of computers in a network, a plurality of network management information collected in the computers are processed and retained as representative management information which is simple single management data. On the other hand, the network management computer periodically inquires the representative management information of each computer. When a response to this is returned, the network management computer can know that each computer is normally operating by analyzing the representative management information.

Further, when the computer is in an incommunicable state, or when there is a failure on the communication path leading to the computer, the network management computer recognizes this event because there is no response from the computer. It becomes possible to do. Furthermore, if there is any problem with the computer inquired from the network management computer, information representative of the management information having the problem is stored in the representative management information, and this is returned at the time of response, A computer for network management is known to have a problem. The network management computer identifies the management information having the problem from this representative management information, and uses this as a trigger to perform management operations such as inquiring of the management information newly having the problem with the computer. It becomes possible to do.

[0016]

Embodiments of the present invention will now be described in detail with reference to the drawings.

In this embodiment, a case will be described in which the present invention is applied to management by a network management protocol SNMP and network management information MIB-II, which are industry standards in TCP / IP communication.

FIG. 3 is a block diagram showing the outline of the network monitoring apparatus according to this embodiment.

In this embodiment, the representative management information (MIB-) is added to the conventional standard network management mechanism as shown in FIG.
The management mechanism according to II) is added.

In FIG. 3, 21 is a representative management information processing unit, 22 is a representative management information extracting unit, 23 is an extended MIB,
Reference numeral 4 is representative management information.

The extended MIB 23 stores a non-standard MIB and holds the representative management information 24 therein. The representative management information extraction unit 22 regularly monitors the management information 14, 15, and 16, analyzes each management information, and checks whether each value indicates an abnormal value. In the case of indicating an abnormal value, a management information identifier uniquely set for identifying the management information is stored in the representative management information 24. On the other hand, in a normal case, the representative management information stores a character string (such as “Normal”) indicating that the representative management information is normal. For example, when the management information 14 indicating the state of the interface changes to “down”, the management information 15 indicating the total number of received frames or the management information 16 indicating the number of error frames exceeds a certain value. The management information extraction unit determines that there is an abnormality, and stores the identifier indicating the management information having the abnormality as the representative management information 24. The management information 14, 15 and 16 are prioritized in descending order of importance, and when a plurality of management information is determined to be abnormal at one time, the identifier of the management information with the highest priority is the representative management information. Is stored as

In the present embodiment, one identifier of the management information with the highest priority is stored as the representative management information, but instead of the object with the highest priority, the latest one or the cause of the abnormality occurrence is clarified. You may use the oldest one at the time of the anomaly that seems to be most likely to be useful. Alternatively,
When an abnormality occurs, several pieces may be stored in the order of occurrence or several pieces may be stored in the order of importance.

On the other hand, the network management computer 1
Then, the representative management information processing unit 21 for analyzing the representative management information and delivering it to different management applications 11 'according to the obtained management information identifier is provided, and the representative management is periodically performed using the SNMP protocol. The information 24 is inquired.

The operation of the representative management information processing unit 21 of the network management computer is shown in the flowchart of FIG.

In step 100 of FIG. 4, the representative management information is inquired, and if it is determined in the next step 102 that there is a response to the inquiry, the process proceeds to the next step 104 to analyze the response contents and inquire. If there is no response to, the no-response management application in step 106 is executed.

After the analysis of the response content in step 104, it is determined in step 108 whether the response content is normal. As a result, if normal, go to step 110 and wait for the monitoring time and return to step 110. If the response contents are not normal,
In the next step 112, the management information identifier is checked, in step 114 it is passed to the individual management application, and in step 116 each individual management application is executed.

With the configuration as described above, the network management computer 1 normally makes a periodic monitoring inquiry only to the representative management information of each computer.
It is possible to confirm that the network management information that each computer has is a normal value, and if an error occurs, the management information identifier can be obtained from the representative management information. It is possible to collect and deal with more detailed management information by starting an inquiry to the management information and related information from.

[0028]

As described above, according to the present invention, in a network management system, a plurality of pieces of network management information are processed by a computer on the management target side, an abnormality is analyzed, and the representative management information is held. Since the network management computer can receive and process this representative information, the load on the network management computer during normal operation can be reduced, and the network management computer and the managed computer can be exchanged during normal operation. This has an excellent effect that the load of management information can be significantly reduced, and thus the load on the network and the communication cost can be reduced.

[Brief description of drawings]

FIG. 1 is a block diagram showing an outline of a conventional network system.

FIG. 2 is a block diagram showing a network management mechanism by a conventional network management computer.

FIG. 3 is a block diagram showing a network management mechanism to which the network monitoring apparatus according to the present embodiment is applied.

FIG. 4 is a flowchart showing the operation of the representative management information processing unit of this embodiment.

[Explanation of symbols]

 1 ... Network management computer 2 ... Computer in same area 11, 11 '... Network management application 12 ... Agent process 13 ... Management information base 21 ... Representative management information processing unit 22 ... Representative management information extraction unit 23 ... Extended MIB 24 ... Representative management information

Claims (1)

[Claims] 1. In a network monitoring device for an intra-regional communication network or a wide-area communication network in which a plurality of computer systems are connected by a communication medium, a plurality of network management information collected in a computer is processed and a simple single A plurality of computers having means for holding as representative management information which is management data, and means for returning the representative management information in response to an inquiry from another computer in the communication network, and representative management information for each computer And a network management computer having the representative management information analysis means, and the network management operation is performed by collecting and analyzing the representative management information of each computer via a communication medium. A network monitoring device characterized by being capable.