JPH1041966A - Node health check method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow a network monitor system to cope with a communication state of nodes by reducing a time for network monitor polling and preventing application of polling to a node whose communication is available. SOLUTION: A network monitor system 1 sends node monitor information denoting a node of a monitor object to a master node 2 directly connected to the system 1 and the master node based on the node monitor information conducts health check polling periodically to subordinate nodes 3A, 3B, 3C being monitor objects, discriminates availability of communication of the subordinate nodes and informs the result to the network monitor system. The network monitor system accesses the subordinate nodes whose communication is available and stops access to the subordinates nodes whose communication is not available based on the information from the master node.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] The present invention relates to a node health check method for monitoring a plurality of nodes in a network.

[0002]

2. Description of the Related Art Conventionally, this kind of node health check function has been used for the purpose of improving the efficiency of a polling operation as disclosed in, for example, Japanese Patent Application Laid-Open No. 1-293043. FIG. 3 shows the flowchart.

In FIG. 3, at the time of start, addresses of all slave stations connected to the network to be polled are registered in advance in the polling table of the master station. A predetermined value is set in the counter (step S401), and at the same time, polling for all slave stations is started based on the address registered in the polling table, and the address of the operating slave station is determined based on the response to polling of each slave station. Is created (step S402).

Next, based on the created polling list table, polling of the operating slave station is started (step S403), and the value of the counter is decremented each time polling is completed once (step S40).
4).

Next, it is determined whether or not the value of the decremented counter is 0 (step S405). If the value of the counter is not 0 (NO in step S405), the process returns to step S403 to poll the operating slave station. repeat. When the value of the counter becomes 0 (step S405)
YES), and returns to step S401 to perform polling for all slave stations in the same manner.

[0006]

However, according to this conventional example, there are the following problems. Since the efficiency of the polling operation is solved by the counter, the polling of the slave station which does not respond temporarily but recovers thereafter is not performed until the counter becomes zero. Therefore, if the initial value set in the counter is too large, there is a disadvantage that the polling operation for the recovered slave station is not performed for a while. Also,
If the value is too small, the effect of improving the efficiency of the polling operation is not so high, and it is difficult to set an appropriate value.

Even when the slave station is temporarily busy and responds to the master station that it cannot respond, the polling is not performed until the counter is initialized next time. I don't know the status. This is inefficient, and if the master station can know the state of the slave station, it is possible to take appropriate measures against it.

An object of the present invention is to provide a network monitoring system (NM) in which one specific node in a network constantly monitors the communication state of another node in the same network.
By notifying the node communication status information to S), it is possible to eliminate unnecessary useless polling to a communication disabled node in the network, and to shorten the network monitoring polling time as compared with the conventional method.

[0009]

SUMMARY OF THE INVENTION The present invention is a node health check method for monitoring a plurality of subordinate nodes via a main node directly connected to a network monitoring system.
The network monitoring system sends node monitoring information indicating a monitoring target node to the main node, and the main node
Based on this node monitoring information, the health check periodically polls the subordinate node to be monitored, determines whether communication of the subordinate node is possible or not, and notifies the network monitoring system of the result. Based on this notification, access is made to the communicable subordinate node, and access to the non-communicable subordinate node is stopped.

[0010] The node monitoring information from the network monitoring system includes information on the inter-node polling cycle and the response waiting monitoring time, and the main node determines the polling cycle and the response waiting monitoring time for the subordinate node based on the information. decide.

[0011]

One node (main node) directly connected to the network monitoring system monitors the communication state of another node (subordinate node) in the network, and changes in the communication state of the subordinate node (response possible, response impossible). , No response) to the network monitoring system, thereby preventing the network monitoring system from polling a node that cannot communicate with the network monitoring system and shortening the network monitoring polling time. In addition, it is possible to prevent polling leakage from a node that has become communicable. The network monitoring system can respond to the situation by notifying the network monitoring system of whether the uncommunicable node has made an unresponsive response or no response.

[0012]

Next, an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 shows an example of a network configuration to which the node health check method according to the present invention is applied. In FIG. 1, a network monitoring system (hereinafter, referred to as NMS) 1 is directly connected to one of a plurality of nodes having the same configuration as a main node 2, and the other nodes are connected to the main node 2 as subordinate nodes 3. It is connected. These NM
S1, the main node 2 and the subordinate node 3 are connected by a transmission path 4, respectively, and constitute a network as a whole. In the example of the figure, three nodes are subordinate nodes 3A and 3B.
3C, the communication state is monitored by the main node 2.

An operation example will be described with reference to FIG. First,
A node health check start command is sent from the NMS 1 to the main node 2. At this time, a list of nodes in the network (hereinafter, referred to as a target node list) is also sent to the main node 2.

When a subordinate node is added or deleted according to the target node list, the main node 2 monitors the new network configuration only by changing this list on the NMS 1 before starting the node health check. Can be. Also, at the start of the node health check, information such as the inter-node polling cycle and the monitoring time for waiting for a response from the subordinate node can be sent to the main node 2, and the main node 2 can be operated in accordance therewith.

In the main node 2, each subordinate node 3
A, 3B, and 3C are polled, and the communication status of each subordinate node 3A, 3B, and 3C is read. In the example of FIG. 2, when a health check is first performed on the subordinate node 3A, there is a response from the subordinate node 3A that can respond, so the main node 2 simply stores the fact and responds to the NMS 1. Does not give any notice.

Next, when a health check is performed on the subordinate node 3B, if there is a response that cannot be responded from the subordinate node 3B, the main node 2 stores the fact and responds to the subordinate node 3B. Since the communication state has changed, the subordinate node 3B notifies the NMS 1 that it cannot respond. Upon receiving this, the NMS 1 stops accessing the subordinate node 3B.

Next, when the main node 2 performs a health check on the subordinate node 3C, if there is no response from the subordinate node 3C even when the timeout occurs, the main node 2
Stores the fact and notifies the NMS 1 that the dependent node 3C has not responded because the communication state of the dependent node 3C has changed. NMS1 which received this
Stops access to the subordinate node 3C.

Thereafter, the main node 2 periodically performs a health check on the subordinate modes 3A, 3B, and 3C.
When a health check is performed on the subordinate node 3C that has not received a response, and there is a response that can be responded from the health check, the main node 2 stores that fact and the communication state of the subordinate node 3C has changed. , Notifies the NMS 1 that the subordinate node 3C can respond. The NMS 1 receiving this restarts the access to the subordinate node 3C.

[0020]

As described above, according to the present invention, the following effects can be obtained. By accurately grasping whether each node in the network is in a communicable state or in a non-communicable state, the network monitoring system polls communicable nodes, and polls non-communicable nodes. Without doing so, it is possible to efficiently check the health of nodes in the network. Since the network monitoring system does not perform polling for nodes that cannot communicate, it is possible to prevent an increase in useless packets generated in the network.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an example of a network configuration to which a node health check method according to the present invention is applied.

FIG. 2 is a sequence diagram of an operation according to the configuration of FIG. 1;

FIG. 3 is a flowchart of a conventional example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Network monitoring system 2 Main node 3A, 3B, 3C Subordinate node 4 Transmission line

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Moritaka Tomikawa 1-403 Kosugi-cho, Nakahara-ku, Kawasaki-shi, Kanagawa Prefecture Within Nippon Electric Telecom System Corporation (72) Inventor Masayuki Shimizu Kosugi-cho, Nakahara-ku, Kawasaki-shi, Kanagawa 1-chome 403 NEC Telecom System Corporation

Claims (2)

[Claims] 1. A node health check method for monitoring a plurality of subordinate nodes via a main node directly connected to a network monitoring system, wherein the network monitoring system indicates a node to be monitored to the main node from the network monitoring system. The monitoring node sends monitoring information, and based on the node monitoring information, the main node periodically polls the subordinate node to be monitored for a health check, determines whether communication of the subordinate node is possible or not, and performs the network monitoring. Notify the system, the network monitoring system accesses the subordinate nodes that can communicate based on the notification,
A node health check method characterized by stopping access to a subordinate node that cannot communicate. 2. The node monitoring information from the network monitoring system includes information on an inter-node polling cycle and a response waiting monitoring time, and the main node performs a polling cycle and a response waiting monitoring on a subordinate node based on the information. 2. The node health check method according to claim 1, wherein a time is determined.