JPH11340991A - Network monitor system using multicast - Google Patents

Abstract

PROBLEM TO BE SOLVED: To attain reduction of the load on a network monitor device and the volume of monitoring traffic on a network without damaging a fault detecting function by preparing a multicast packet of polling, transmitting it to a repeater, copying the packet and transmitting it to the second repeater. SOLUTION: A manager 1 transmits polling for confirming the existence of a monitor object device or reading an active state to a repeater (first step) 21 as a packet to a multicast address allocated for network monitor. The repeater 21 copies the multicast packet and transmits it to repeaters (second step) 22 and 23. Similarly, the repeater 22 copies the multicast packet and transmits it to repeaters (third step) 24 and 25. Thus, the multicast packet of polling is delivered to all the monitoring object devices (terminals) participating in the multicast group.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a network monitoring method for monitoring the operation state of LAN equipment, and more particularly to a multicast for reducing the load on a network monitoring device and the amount of monitoring traffic on a network without impairing a failure detection function. The present invention relates to a network monitoring method that uses a network.

[0002]

2. Description of the Related Art In a conventional network monitoring method, all polling for confirming existence and polling for reading data for each agent connected to a network are individually performed by a manager. There is a problem that a load is imposed, and various distributed management systems have been proposed to solve this problem.

One example of a conventional distributed management system is disclosed in
FIG. 4 shows a configuration diagram described in JP-A-172435. This conventional distributed management system includes one manager 2 and a plurality of agents 5. The manager 2 monitors a plurality of agents 5 connected to the network, and here comprises a monitoring information setting means 3 and an agent management table 4. The agent 5 is a terminal (monitor) monitored by the manager 2 connected to the network, and includes a polling unit 6, an agent management table 4, a monitoring table 7, and the like.

[0004] The conventional distributed management system having such a configuration operates as follows.

That is, the monitoring information setting means 3 of the manager 2 divides the agents connected to the network into groups, sets the agents in the groups, the representative agents, and the priorities in frames, and transmits the frames to the representative agents. 5 sets the addresses and priorities of the agents in the group set in the frame received in the monitoring table 7 and transmits and saves them to the other agents 5 in the group. Is sent and there is no response or Agent 5
When the abnormality notification is received in a frame from, the manager 2 is notified of the abnormality of the agent 5 by the frame.

Further, the monitoring information setting means 3 of the manager 2
Are grouped for the agents 5 connected to the network, and the agents 5 in the group and the representative agent are set in a frame and transmitted to the representative agent. When the address of the agent 5 and the determined priority are set in the monitoring table 7 and transmitted to the other agents 5 in the group to be stored, a frame is transmitted to the agent 5 set in the monitoring table 7 and there is no response. Alternatively, when the manager 2 receives an abnormal notification in a frame from the agent 5,
Is notified of the abnormality of the agent 5 by a frame.

At this time, the agent 5 sets the monitoring table 7
And transmits a frame to the agent 5 with the highest priority, and when there is no response, determines the agent 5 with the next priority as the representative agent and sends it to all the other agents 5 and the manager 2 in the group. I am trying to notify you.

When the agent 5 refers to the monitoring table 7 and transmits a frame to the agent 5 having a higher priority, and there is a response that has not been answered, the agent 5 having a higher priority becomes the representative agent when the agent 5 is restored. The return is notified to all other agents 5 and managers 2 in the group.

Also, when a frame in which an access request is set from the manager 2 is received, when the password and access right authentication in the frame are OK, the data is set in the frame and transmitted to the own agent to the network. When the difference between the received data and the specified data exceeds a predetermined value or is equal to or less than a predetermined value, the manager 2 is notified of an abnormality.

Therefore, each agent 5 has a polling function, the agent performs polling, and notifies the manager 2 only when a state change or an abnormality occurs, thereby making it possible to balance the load on the network.

[0011]

However, this prior art has the following problems.

The first problem is that in the conventional network monitoring, polling is performed on all monitored devices by unicast, so that the load on the network monitoring device and the amount of monitoring traffic on the network are large.

The second problem is that the polling is executed by the representative agent and the result is notified to the manager, so that the processing is complicated, and it is necessary to provide each agent with a special polling function. There is a lot of traffic.

In view of the above, the present invention has been made in view of the above, and in a network monitoring device for monitoring the operation state of a LAN device, the load of the network monitoring device and the monitoring on the network are maintained without impairing the failure detection function. An object of the present invention is to provide a network monitoring method using multicast that reduces the amount of traffic.

[0015]

A network monitoring system using multicast according to the present invention is a network monitoring system for monitoring the operation state of a LAN device. In the network monitoring system, a manager creates a multicast packet for polling in the manager and connects it in a tree-like manner. The first-stage relay device, which copies the received multicast packet and transmits it to the second-stage relay device;
Similarly, the transmission to the next stage is repeated in order to make the polling multicast packet reach all the monitoring target devices participating in the multicast group.

The network monitoring system using multicast according to the present invention is a network monitoring system for monitoring the operation state of a LAN device. In the network monitoring system, a manager creates a multicast packet for polling and connects the manager to the manager in a tree form. The first-stage relay device copies the received multicast packet, transmits the copied multicast packet to the second-stage relay device, and repeats the transmission to the next stage in the same manner. A polling multicast packet is delivered to all the monitored devices participating in the multicast group, and each of the relay devices receiving the poll in the multicast packet unicasts a response to the poll to a manager. -It is characterized by transmission in packets.

Further, the manager records response information of the monitored device to the polling in a management table.

Further, the multicast packet is:
It is characterized in that polling for confirming the existence of the monitoring target device and polling for reading the operation status are packets addressed to a multicast address allocated for network monitoring.

Further, each of the relay devices is equipped with a standardized multicast function.

[0020]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a configuration diagram of a network monitoring system using multicast according to the present invention.

In the network monitoring system using multicast according to the present invention, a manager 1 that executes polling for confirming the existence of a monitored device and polling for reading the operation state is connected to a tree corresponding to multicast. Relay devices 21, 22, 23, 24, 25,
26 and 27.

The manager 1 includes a management table 11 and operates as follows. In FIG. 1, first,
Manager 1 and each relay device 21, 22, 23, 24, 2
5, 26, 27 join a multicast group for network monitoring. The manager 1 transmits polling for confirming the existence of the monitoring target device and polling for reading the operation status to the relay device (first stage) 21 as a packet addressed to the multicast address allocated for network monitoring. The relay device 21 copies the received multicast packet and transmits it to the relay device (second stage) 22 and the relay device (second stage) 23. Similarly, the relay device 22 copies the received multicast packet and transmits it to the relay device (third stage) 24 and the relay device (third stage) 25. Similarly, the relay device 23 copies the received multicast packet and transmits it to the relay device (third stage) 26 and the relay device (third stage) 27. In this way, the multicast packet for polling reaches all the monitoring target devices (terminals) participating in the multicast group.

Each of the relay devices 21, 22, 23, 24, 25, 2 that has received the polling by the multicast packet
6 and 27 transmit a response to the polling to the manager 1 as a unicast packet. The manager 1 records information of the managed device that has received the response in the management table 11.

As a result, the load on the network monitoring apparatus and the amount of monitoring traffic on the network are reduced as compared with the case where polling is performed from the network monitoring apparatus to all monitored apparatuses by unicast.

Next, the overall operation of this embodiment will be described in detail with reference to the block diagram of FIG. 1 and the flowchart of FIG.

First, the manager 1 performs a multicast
Join the group (step A1 in FIG. 2). At this time, a notification of participation in the multicast group is transmitted to the subordinate relay device 21 by multicast. Each relay device to be monitored also sends a multicast notification of participation in the multicast group. Next, the manager 1 creates a polling packet (step A).
2), the packet is transmitted to the relay device 21 by multicast (step A3). Then, a polling packet is created again after a predetermined polling cycle (step A2). Further, when a response to the polling sent from each of the relay devices to be monitored by unicast is received (step A5), the information is registered in the management table 11 (step A6).

On the other hand, the relay device 21 first joins the multicast group (step B1). At this time, notification of participation in the multicast group is transmitted to the manager 1 and the subordinate relay devices (22, 23) by multicast. Next, when the relay device 21 receives the polling packet from the manager 1 by multicast (step B2), the multicast distribution is performed by referring to the member list of the multicast group that is periodically exchanged between the relay devices. It is determined whether it is necessary (step B3). If there is a relay device that has joined the multicast group, a polling packet is transmitted to the subordinate relay devices 22 and 23 by multicast as needed (step B4). Thereafter, a response to the polling is transmitted to the manager by unicast (step B5).

Next, a specific example will be described. As shown in FIG. 3, for example, the management table 11 has items of group number, host address, and response time. The group number is a number for identifying a multicast group. The monitoring target device can be divided into a plurality of multicast groups and monitored. For example, a multicast group can be classified according to importance, and the polling cycle can be changed for each group. The host address is the address of the monitoring target device, and can be known from the source address included in the polling response packet. The response time records the time at which the polling response packet was received. If this time is not updated for a certain period of time or more, it is considered that an abnormality has occurred in the monitored device or network. When the manager 1 transmits a polling packet by multicast and receives a response to the polling, the manager 1 registers the information in the management table 11. If the response time of the management table 11 is not updated for a certain period of time or more, the occurrence of a failure is notified to the network administrator.

[0030]

The first effect is that, in the conventional network monitoring, polling is performed on all the devices to be monitored by unicast. However, since this polling is replaced by multicasting, the load on the network monitoring device and the monitoring on the network are changed. The traffic volume can be reduced.

The second effect is that since there is no need for complicated processing such as polling performed by the representative agent and notification of the result to the manager, it is only necessary to implement a standardized multicast function. Is that the load on the agent can be reduced as compared with.

A third effect is that it is not necessary to mount a special polling function in each agent, but only a standardized multicast function, so that agents can be realized at low cost.

[Brief description of the drawings]

FIG. 1 shows a configuration diagram of a network monitoring system using multicast according to the present invention.

FIG. 2 shows a flowchart of a network monitoring method using multicast according to the present invention.

FIG. 3 shows a management table of a network monitoring method using multicast according to the present invention.

FIG. 4 shows a configuration diagram of a conventional technique.

[Explanation of symbols]

 1, 2 manager 3 monitoring information setting means 4 agent management table 5 agent 6 polling means 7 monitoring table 11 management table 21, 22, 23, 24, 25, 26, 27 relay device

Claims (5)

[Claims] In a network monitoring method for monitoring the operation state of a LAN device, a manager creates a polling multicast packet and sends it to a first-stage relay device connected to the manager in a tree form. The second relay device copies the received multicast packet, transmits it to the second relay device, repeats the transmission to the next stage in the same manner, and joins the polling multicast packet to the multicast group. A network monitoring method using multicast, which is sent to all monitored devices. 2. In a network monitoring method for monitoring the operation state of a LAN device, a manager creates a polling multicast packet and transmits it to a first-stage relay device connected to the manager in a tree-like manner. The second relay device copies the received multicast packet, transmits it to the second relay device, repeats the transmission to the next stage in the same manner, and joins the polling multicast packet to the multicast group. Each relay device that has received the polling in the multicast packet, sends a response to the polling in a unicast packet to the manager. The network monitoring method used. 3. The network monitoring system using multicast according to claim 1, wherein the manager records response information of the monitoring target device to the polling in a management table. 4. The method according to claim 1, wherein the multicast packet is a packet addressed to a multicast address assigned for polling for monitoring the existence of the device to be monitored and polling for reading the operation status for network monitoring. A network monitoring system using multicast according to any one of claims 1 to 3. 5. The relay device according to claim 1, wherein each of the relay devices implements a standardized multicast function.
A network monitoring method using multicast according to any one of claims 1 to 4.