JP2018152672A - Monitoring method, monitoring device, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To distribute monitoring loads.SOLUTION: A monitoring method for a network including a plurality of nodes includes the steps of classifying the plurality of nodes according to a hop number with a monitoring device on a communication path with a minimum hop number with the monitoring device communicatively connected with the network for monitoring the network, and allocating a monitorable monitoring destination node so that monitoring loads are balanced among nodes with the same hop number at the time of setting a monitoring relation among the nodes so that a node at a side with a smaller hop number monitors a node at a side with a larger hop number.SELECTED DRAWING: Figure 9

Description

  The present invention relates to a monitoring method, a monitoring device, and a program.

  There is a network system including a plurality of network devices (hereinafter also referred to as “nodes”) and a network controller (hereinafter also referred to as “monitoring server”) that monitors and controls each network device. The monitoring server performs monitoring by directly communicating with all the nodes to be monitored by, for example, a so-called server / client method.

JP 2011-130144 A JP 2013-47922 A

  With the recent increase in the scale of networks, the number of nodes to be monitored by the monitoring server has increased, and the load on the monitoring server has increased. If the increase in load is dealt with by increasing the number of monitoring servers or adopting high-spec (high-performance) monitoring servers, there is a problem that the cost associated with the facilities increases.

  An object of the present invention is to provide a technique capable of distributing a monitoring load.

  In one aspect, a computer included in a monitoring device that is communicably connected to a network including a plurality of nodes and that monitors the network is connected to the monitoring device in a communication path with a minimum number of hops between the monitoring device and the computer. When classifying the plurality of nodes according to the number of hops between them, and when setting the monitoring relationship between the nodes so that the node with the smaller hop number monitors the node with the larger hop number, And a process of allocating monitoring destination nodes that can be monitored so that the monitoring load is balanced between nodes having the same hop count.

  In one aspect, the monitoring load can be distributed.

FIG. 1 shows an example of a network system according to the embodiment. FIG. 2 shows an example of a tree. FIG. 3 is a diagram illustrating a hardware configuration example of an information processing apparatus (computer) applicable to the monitoring server and the node. FIG. 4 is an explanatory diagram of monitoring cost information. FIG. 5 shows an example in which layers and monitoring costs for a plurality of nodes in the network system shown in FIG. 1 are determined. FIG. 6 shows an example of allocation of child nodes in the example of FIG. FIG. 7 is a diagram illustrating an example of a method of allocating child nodes between parent nodes. FIG. 8 is a diagram illustrating an example of a method of allocating child nodes between parent nodes. FIG. 9 is a flowchart illustrating an example of a monitoring process executed by the CPU of the monitoring server (information processing apparatus operating as the monitoring server). FIG. 10 is a flowchart illustrating an example of processing as a parent node of a node. FIG. 11 is a flowchart illustrating a processing example of the CPU of the monitoring server when communication disconnection is detected.

  Hereinafter, a monitoring method, a monitoring device, and a program according to embodiments will be described with reference to the drawings. The configuration of the embodiment is an exemplification, and the present invention is not limited to the configuration of the embodiment.

<Network configuration>
FIG. 1 shows an example of a network system according to the embodiment. The network system includes a monitoring server 1 that monitors the network and a plurality of nodes 2 included in the network. The monitoring server 1 is an example of a “monitoring device”.

  The plurality of nodes 2 are network devices monitored and controlled by the monitoring server 1. The node 2 (network device) includes a terminal device and a relay device. The terminal device includes, for example, a sensor terminal having a communication function called a personal computer, a workstation, a server machine, a tablet terminal, a smartphone, a sensor node, or the like. The relay device includes a router, a layer 2 / layer 3 switch, a HUB, and the like. However, the network device may include network devices other than the above examples. The topology between the nodes 2 can be appropriately set as long as each node 2 is connected to the monitoring server 1 directly or indirectly (through another node 2). The number of nodes 2 can be set as appropriate.

  The monitoring server 1 monitors and controls a plurality of nodes 2. As a protocol used for monitoring, Simple Network Management Protocol (SNMP), ping, telnet, or the like can be applied. However, protocols other than those described above may be applied for monitoring.

  The monitoring server 1 in the embodiment forms a tree having the monitoring server 1 as a vertex (root). FIG. 2 shows an example of a tree. The tree is used as a route for transferring monitoring information (monitoring information) from the end of the tree toward the vertex.

  The node 2 located at the first hop from the monitoring server 1 is set as a “representative node”. Nodes other than the representative node are set to “general nodes”. The general node is connected to the representative node or another node.

  In the example of the tree shown in FIG. 2, two representative nodes # 1 and # 2 are set. The first hop to which the representative node belongs is called “layer 1”. The second and subsequent hops from the monitoring server 1 are called layers 2, 3,. In the example of FIG. 2, general nodes a to e are illustrated, and the general node a and the general node d belong to the layer 2. General node b, general node c, and general node e belong to layer 3. Each arrow illustrated in FIG. 2 is directed from the child node to the parent node.

  The representative node forms a parent-child relationship with the monitoring server 1, and the general node forms a parent-child relationship with the representative node or another general node. In the example illustrated in FIG. 2, there is a node 2 that is general nodes a to e. The general node a is a child node having the representative node # 1 as a parent node, and the general node d is a child node having the representative node # 2 as a parent node. The general node a is a parent node of the general node b and the general node c, and the general node d is a parent node of the general node e.

The monitoring server 1 generates a tree (node parent-child relationship), and transmits monitoring control data to each node 2. The monitoring control data includes data indicating a parent-child relationship for each node 2. For example, the monitoring control data includes the following information.
-Information indicating one or more monitored child nodes (information indicating monitored nodes)
-Upper layer node 2 or monitoring server 1 for notifying the monitoring result (information indicating the node transmitting the monitoring result)
Information indicating the monitoring target item, for example, information indicating an alarm (alarm) or state of the monitoring target.

  Each node 2 monitors the child node according to the monitoring control data. The monitoring includes two types of “status collection” for collecting information of monitoring results from the child node and “event notification” for receiving information transmitted spontaneously by the child node. Each node 2 adds the monitoring result of its own node to the monitoring result from the child node and transmits it to the parent node.

<Hardware configuration>
FIG. 3 is a diagram illustrating a hardware configuration example of an information processing apparatus (computer) applicable to the monitoring server 1 and the node 2. As an example, the information processing apparatus 10 includes a central processing unit (CPU) 11, a main storage device 12, an auxiliary storage device 13, a communication interface (communication IF) 14, and an input connected to each other via a bus B 1. A device 15, an output device 16, and a sensor 17 are included.

  The main storage device 12 is used as a program development area, a work area for the CPU 11, a data and program storage area, a communication data buffer area, and the like. The main storage device 12 is formed of, for example, a random access memory (RAM) or a combination of a RAM and a read only memory (ROM).

The auxiliary storage device 13 is used as a storage area for data and programs. The auxiliary storage device 13 is, for example, a hard disk drive (HDD), a solid state drive (SSD), a flash memory, an electrically erasable programmable read-only memory (EEPRO).
M). Each of the main storage device 12 and the auxiliary storage device 13 is an example of “storage device”, “storage medium”, “memory”, and “storage unit”.

  The communication IF 14 manages communication processing. For example, a network interface card (NIC) is used for the communication IF 14. The input device 15 is, for example, a key, a button, a pointing device (such as a mouse), a touch panel, or a voice input device (microphone). The output device 16 is, for example, a display, a printer, a speaker, a lamp, or the like.

  The CPU 11 loads the program stored in the auxiliary storage device 13 to the main storage device 12 and executes it. The information processing apparatus 10 operates as the monitoring server 1 by executing the program. The CPU 11 performs processing for generating the above-described tree and transmitting monitoring control data based on the tree to each node 2. In addition, the monitoring result at each node 2 is received, and the node 2 and the network are controlled through analysis (monitoring) of the monitoring result. The CPU 11 can operate as a “classification unit” and an “allocation unit” by executing the program.

The CPU 11 is an example of a “control device”, “control unit”, “controller”, and “processor”. The CPU 11 is also called an MPU (Microprocessor) or a processor. The CPU 11 is not limited to a single processor, and may have a multiprocessor configuration. A single CPU connected by a single socket may have a multi-core configuration. At least a part of the processing performed by the CPU 11 may be executed by a multi-core or a plurality of CPUs. At least a part of the processing performed by the CPU is performed by a processor other than the CPU, for example, a dedicated processor such as a digital signal processor (DSP), a graphics processing unit (GPU), a numerical operation processor, a vector processor, or an image processing processor. Also good.

Further, at least a part of the processing performed by the CPU 11 may be performed by an integrated circuit (IC) or other digital circuits. Further, the integrated circuit and the digital circuit may include an analog circuit. Integrated circuit is LSI, Application Specific Integrated Circuit
(ASIC) and programmable logic device (PLD). The PLD includes, for example, a field-programmable gate array (FPGA). At least a part of the processing performed by the CPU 11 may be executed by a combination of a processor and an integrated circuit. The combination is called, for example, a microcontroller (MCU), a SoC (System-on-a-chip), a system LSI, a chip set, or the like.

<Monitoring server processing>
Next, generation and transmission processing of a tree and monitoring control data in the monitoring server 1 will be described. For example, the auxiliary storage device 13 of the information processing apparatus 10 that operates as the monitoring server 1 stores topology information, node information, monitoring cost information, tree information, and the like as shown in FIG.

The topology information indicates the connection state of each node 2. The node information includes the network address (Internet Protocol (IP) address, Media Access Control (MAC)) of each node 2
Address), device type, and information indicating the device configuration. The monitoring cost information includes information used for calculating a cost that occurs when each node 2 is monitored. The tree information includes information indicating layers, nodes belonging to each layer, and parent-child relationships of the nodes. These pieces of information are used when the CPU 11 generates a tree and monitoring control data by executing a program.

<< Determination of monitoring cost for each node >>
The monitoring server 1 weights each node 2 to be monitored with resources (amount of use of the CPU 11 and memory (main storage device 12 and the like)) used for monitoring. The weight indicates the load applied to the monitoring server 1 such as the usage time and usage of the CPU 11 and the memory (main storage device 12 or the like) used for monitoring and the data collection time. The weight is called “monitoring cost”.

  FIG. 4 is an explanatory diagram of monitoring cost information. As an example, the monitoring cost is determined by a combination of monitoring cost elements as shown in the table of FIG. In the example of FIG. 4, “alarm / status collection number”, “collection interface”, “collection complexity”, and “collection time” are illustrated as monitoring cost elements. The monitoring cost element may be at least one selected from the above, and elements other than the above may be adopted.

  “Number of alarms / states collected” indicates the number of alarms and states to be monitored (an example of the number of information items to be monitored). The monitoring cost increases in proportion to the increase in the number of collections. For example, the collection number is set as the cost value as it is. If the number of collection is 15, for example, the cost value is 15.

  The “collection interface” indicates a monitoring information collection method. For example, there are a case where monitoring information is collected using an SNMP TRAP command and a case where monitoring information is collected using an SNMP GET command. In GET, the SNMP agent returns a response (monitoring information) in response to a request from the SNMP manager. In TRAP, the SNMP agent voluntarily transmits information (monitoring information) to the SNMP manager. Since the GET procedure is more complicated than TRAP, the monitoring cost is also increased. For example, the cost value for GET is set to 10, the cost value for TRAP is set to 2, the cost value for ping is set to 1, and the cost value for telnet is set to 20.

  “Collecting complexity” indicates the logic (number of procedures: number of command executions, for example) from the start to the end of collection. The cost value increases as the number of procedures increases. For example, the number of command executions can be set as a cost value as it is.

  “Collecting time” indicates, for example, the time from request transmission of monitoring information to response reception. The longer the collection time, the higher the monitoring cost. For example, the cost value is set to 1 when the collection time is 0.1 seconds, and the cost value is set to 5 when the collection time is 1 second. However, the cost value for each monitoring cost element can be set as appropriate, and may be set so that the cost value increases as the monitoring load increases.

The monitoring server 1 uses the node information of each node 2 to be monitored to calculate a monitoring cost value when monitoring information is collected from each node. For example, the following formula can be applied.
Monitoring cost = Σ (i + j + k) x (h)
However, i is a cost value of the collection interface, j is a cost value of the collection complexity, and k is a cost value of the collection time. h is the number of alarm / status collections for the combination of i, j and k. However, the monitoring cost may be calculated using a formula other than the above formula.

<< Determination of monitoring layer for each node >>
The monitoring server 1 uses the topology information to set the number of hops of the shortest path from the monitoring server 1 to each node 2 as the “monitoring layer” to which each node 2 belongs. The node corresponding to the first hop path from the monitoring server 1 is layer 1.

  As described above, the layer 1 node is referred to as a “representative node”. The second hop is layer 2, and the third and subsequent hops are layer 3. Nodes after layer 2 are called “general nodes”. FIG. 5 shows an example in which layers and monitoring costs for a plurality of nodes 2 in the network system shown in FIG. 1 are determined.

  The node 2 (monitoring source) of each layer monitors the node 2 (monitoring destination) of the immediately lower layer. For example, the layer 1 node 2 monitors the layer 2 node. The monitoring source layer is called a parent layer, the parent layer node 2 is called a parent node, the monitoring destination layer is called a child layer, and the child layer node 2 is called a child node. The number of child nodes monitored by one node 2 may be two or more.

<< Determine the monitoring destination of each node >>
In the parent layer, the nodes in the child layer are determined so that the monitoring cost is distributed among the parent nodes. For example, a description will be given using the example shown in FIG. Consider assigning child nodes (general node c, general node d, and general node e) of layer 3 to parent nodes (general node a and general node b) of layer 2. In this case, the allocation is performed so that the sum of the monitoring costs of the monitoring target child nodes of the general node a and the total monitoring cost of the monitoring target child nodes of the general node b are close (balanced) (see FIG. 6). ).

  As an example of the result, the monitoring server 1 determines the general node c and the general node d (total value of monitoring costs: 1500) as the monitoring destination (child node) of the general node a. In addition, the monitoring server 1 determines a general node c and a general node d (total monitoring cost: 1500) as monitoring destinations (child nodes) of the general node a.

7 and 8 are diagrams showing an example of a method for allocating child nodes between parent nodes. First, the monitoring server 1 takes out (i) a node (parent node group) belonging to the parent layer and a node (child node group) belonging to the child layer. 7 and 8 are examples, and are different from the topology shown in FIG.

  For example, as shown in FIG. 7, the monitoring server 1 takes out data of a parent node group of a parent layer (layer 2) and a child node group of a child layer (layer 3). The parent node group includes a parent node A, a parent node B, and a parent node C, and the child node group includes a child node H, a child node I, a child node J, and a child node K. Assume that the monitoring costs of child node H, child node I, child node J, and child node K are 1000, 2000, 3000, and 4000, respectively.

  The monitoring server 1 generates (ii) a parent candidate table corresponding to each child node. As shown in FIG. 8, the parent candidate table T1 includes a list of parent node candidates (nodes in the immediately upper layer on the shortest path from the monitoring server 1). In other words, the parent candidate table T1 stores identification information of one or more parent node candidates corresponding to the identification information of the child node. FIG. 8 shows parent candidate tables T11, T12, T13, and T14 corresponding to the child node H, child node I, child node J, and child node K. As shown in FIG. 8, the parent candidate table T1 may store the monitoring cost of the child node.

The monitoring server 1 (iii) generates a “parent node monitoring table” indicating which child node each parent node monitors according to the following conditions.
(Rule 1) A child node (first child node) having only one parent node candidate (also referred to as a parent candidate) is unconditionally placed under the monitoring of the parent node.
(Rule 2) A child node (second child node) having a plurality of parent candidates checks a parent node monitoring table as a parent candidate, and selects a parent having the smallest total cost.

  In the example shown in FIG. 8, the monitoring server 1 determines the parent nodes in order of young numbers. However, as long as the result of allocation is balanced, the order of determination of the parent nodes can be changed as appropriate. The parent node candidate for the child node H is one of the parent nodes A. Therefore, the monitoring server 1 registers the child node H in the parent node monitoring table T2 according to the rule 1.

  For example, as shown in FIG. 8, the parent node monitoring table T2 stores the identification information of each child node and the total value (total cost) of the monitoring cost of the child node in association with the identifier of the parent node. . FIG. 8 shows parent node monitoring tables T21, T22, and T23 corresponding to the parent node A, parent node B, and parent node C. In each of the parent candidate table T1 and the parent node monitoring table T2, the parent node candidate and the monitoring cost of each parent node may be stored.

  For the child node I, there are a parent node A, a parent node B, and a parent node C as parent node candidates. The monitoring server 1 refers to the total cost of the parent node candidate and registers the child node I in the parent node monitoring table T2 of the parent node candidate having a small total cost value (rule 2). At this time, if there are a plurality of parent node monitoring tables T2 having the same total cost value, the child node I is registered in one parent node monitoring table T2 according to a predetermined priority. In this example, the child node I is registered in the parent node monitoring table T22 of the parent node B having a smaller node number among the parent node B and the parent node C that are candidates for the parent node. However, the priority order can be set as appropriate.

  The parent node candidate for the child node J is one of the parent nodes A. Therefore, the monitoring server 1 registers the child node J in the parent node monitoring table T21 according to the rule 1. The monitoring server 1 updates the total cost value of the parent node monitoring table T21 to 4000 (1000 + 3000).

The monitoring server 1 registers the child node K in the parent node monitoring table T23 corresponding to the parent node C according to the rule 2. As a result, the monitoring costs for the child nodes H, I, J, and K are allocated in a balanced manner between the parent nodes A, B, and C. That is, the load for monitoring the child nodes H, I, J, and K is distributed among the parent nodes A, B, and C. The parent candidate table T1 and the parent node monitoring table T2 are generated and stored in at least one of the main storage device 12 and the auxiliary storage device 13.

<Processing by CPU>
FIG. 9 is a flowchart illustrating an example of a monitoring process executed by the CPU 11 of the monitoring server 1 (the information processing apparatus 10 operating as the monitoring server 1). The process shown in FIG. 9 is started upon input of a predetermined initial trigger such as activation of the monitoring server 1. However, the processing start condition in FIG. 9 may be other than the above.

  In 001, the CPU 11 of the monitoring server 1 determines the monitoring cost of each node 2 (see FIG. 4 and the like). In 002, the CPU 11 of the monitoring server 1 determines each node monitoring layer (see FIG. 5). In 003, the CPU 11 of the monitoring server 1 takes out the node 2 belonging to the parent layer and the node belonging to the child layer (see FIG. 7).

  In 004, the CPU 11 of the monitoring server 1 generates a parent candidate table T1 related to the child node extracted in 003. The processing of 005 and 006 loops as many times as the number of extracted child nodes. In 005, the CPU 11 of the monitoring server 1 determines whether there is one parent node candidate. In 006, when it is determined in 005 that there is only one parent node candidate, the CPU 11 determines the parent node candidate as the parent node and registers it in the corresponding parent node monitoring table T2.

  The processing of 007 and 008 loops for the number of child nodes having a plurality of candidate parent nodes. Further, 007 loops for each parent node candidate in the parent candidate table T1. In 007, the CPU 11 of the monitoring server 1 selects a parent node candidate having a small total cost. In 008, the CPU 11 of the monitoring server 1 determines the parent node candidate selected in 007 as a parent node and registers it in the parent node monitoring table T2.

  The processing from 003 to 008 is also executed for the remaining layers. That is, the extraction in the process of 003 is performed starting from the lowest layer. For example, in the example of FIG. 5, nodes having layer 2 and layer 3 as a parent layer and a child layer, respectively, are extracted. When the allocation for this is completed, the process for the one higher layer is performed.

  In the example of FIG. 5, nodes having layer 1 and layer 2 as a parent layer and a child layer, respectively, are extracted (003), and the processes of 004 to 008 are performed. At this time, the total value of the monitoring cost of the child node hanging from each child node and the monitoring cost of the own node is used as the monitoring cost of the child node. The processing from 003 to 008 is repeated until layer 1 is taken out as a parent layer.

  In this way, the tree information (FIG. 2) finally having the monitoring server 1 as a vertex is generated by the monitoring server 1, and at least one of the main storage device 12 and the auxiliary storage device 13 (hereinafter referred to as “memory”). Is remembered. In 009, the CPU 11 of the monitoring server 1 transmits monitoring control data to each node 2 (representative and general nodes). The monitoring control data is transmitted from the communication IF 14 to each node 2.

  In the information processing apparatus 10 operating as each node 2, the monitoring control data is received by the communication IF 14 and stored in the memory (at least one of the main storage device 12 and the auxiliary storage device 13). The CPU 11 of each node 2 monitors information to be monitored (such as an alarm or status) using the monitoring control data, and stores the monitoring result (also referred to as monitoring information) in the memory.

  FIG. 10 is a flowchart illustrating a processing example of the node 2 as a parent node. The process 1 in FIG. 10 starts when the CPU 11 of the node 2 receives a trigger for periodic polling. In 101, the CPU 11 collects monitoring information (information indicating the monitoring result of the monitoring target in the child node) from each monitoring target child node. The collection is performed by transmitting a monitoring result transmission request to each child node and receiving a response.

  In 102, the CPU 11 of the node 2 adds the monitoring information of itself (own node) to the monitoring information collected from each child node, and generates notification information to the parent node (notification destination). In 103, the CPU 11 of the node 2 determines whether or not the communication with the notification destination (parent node) is normal. When it is determined that the communication is normal, the CPU 11 transmits notification information to the parent node (104). When it is determined that the communication is not normal, the CPU 11 saves (stores in the memory) the notification information (105). The saved data is transmitted to the parent node when communication with the parent node is restored.

  Process 2 shown in FIG. 10 is started when monitoring information is received from a child node as an event. In 111, the CPU 11 of the node 2 adds the monitoring information of itself (own node) to the monitoring information collected from each child node, and generates notification information to the parent node (notification destination). The process 111 is the same as the process 102. Thereafter, the process proceeds to 103.

  The monitoring server 1 receives the notification information including the monitoring information from the representative server and the general nodes below it from each representative server by executing the processing 1 and the processing 2 shown in FIG. Can do. The monitoring server 1 controls the network and each node 2 using the received monitoring information.

  FIG. 11 is a flowchart illustrating a processing example of the CPU 11 of the monitoring server 1 when the communication disconnection is detected. In 201, the CPU 11 of the monitoring server 1 receives notification information from each representative node to the monitoring server 1.

  In 202, the CPU 11 of the monitoring server 1 identifies the node X that is disconnected. For example, referring to FIG. 6 as an example, when the representative node # 1 becomes unable to communicate with the general node a by an existing method such as a life check, a notification of “disconnection with the general node a” is sent to the monitoring server 1. Send. In 202, the CPU 11 of the monitoring server 1 determines whether a communication disconnection notification has been received. Here, the CPU 11 specifies the general node a as the node X.

  In 203, the CPU 11 of the monitoring server 1 identifies the node Y monitored by the node X. The identification of the node Y can be performed using tree information (monitoring control data for the node X). The CPU 11 identifies the general node c and the general node d, which are child nodes of the node X (general node a), as the node Y. In this way, the CPU 11 detects all nodes 2 downstream of the node X on the tree as nodes Y.

  In 204, the CPU 11 of the monitoring server 1 uses the information (path information) of the tree excluding the node 2 (representative node # 1) that can no longer communicate with the node X and the node Y, and the parent node of the node X and the node Y Is newly determined. As the method for determining the parent node, the method described with reference to FIGS. 7 and 8 is used.

In 205, the CPU 11 of the monitoring server 1 transmits monitoring control data to the nodes X, Y, and the parent node. As a result, even if communication between nodes occurs, the tree can be repaired and monitoring can be continued based on the tree information indicating the transfer route of the monitoring information to the monitoring server 1.

<Effects of Embodiment>
In the embodiment, a CPU 11 (an example of a computer) included in a network monitoring server 1 (an example of a monitoring device) including a plurality of nodes 2 performs the following processing.

  The plurality of nodes are classified according to the number of hops with the monitoring server 1 in the communication path having the minimum number of hops with the monitoring server 1 connected to the network so as to be communicable.

  When setting the monitoring relationship between nodes so that the node with the smaller hop count monitors the node with the larger hop count (setting the parent-child relationship), the monitoring load is balanced between the nodes with the same hop count The monitoring destination node (child node) that can be monitored is allocated as described above.

  According to the embodiment, the monitoring server 1 generates a tree that becomes a transfer route of monitoring information, and transmits monitoring control data based on the tree to each node 2 that becomes a parent node. Thereby, the monitoring load of the node 2 can be distributed to the parent node. Therefore, the load on the monitoring server 1 is reduced or reduced. In other words, an increase in the load on the monitoring server 1 due to an increase in the number of monitoring targets can be suppressed. Therefore, the monitoring server 1 can collect monitoring information from each node to be monitored without increasing the number of monitoring servers or adopting a high-performance monitoring server. That is, it is possible to cope with an increase in the number of nodes to be monitored while suppressing an increase in equipment cost.

  In addition, in each layer, the total cost of the child nodes hanging from the parent node is balanced between the parent nodes (so that there is no bias), so that load imbalance does not occur between the parent nodes of the same layer, Smooth collection of monitoring information can be performed.

  Furthermore, when communication disconnection occurs in the tree path, information indicating communication disconnection is transmitted from the node 2 to the monitoring server 1, and the monitoring server 1 reconstructs the tree. As a result, even if communication disconnection occurs, monitoring (collection of monitoring information) can be continued using the reconstructed tree. The configurations described in the embodiments are examples and can be combined as appropriate.

DESCRIPTION OF SYMBOLS 1 ... Monitoring server 2 ... Node 10 ... Information processing apparatus 11 ... CPU
12 ... Main storage device 13 ... Auxiliary storage device

Claims (6)

A computer included in a monitoring device connected to a network including a plurality of nodes so as to be able to communicate with the network,
A process of classifying the plurality of nodes according to the number of hops with the monitoring device in the communication path of the minimum number of hops with the monitoring device;
When setting the monitoring relationship between nodes so that the node with the smaller number of hops monitors the node with the larger number of hops, the monitoring load can be balanced between the nodes with the same number of hops. A process for allocating a monitoring target node,
A program that executes In the process of performing the allocation,
A plurality of nodes on the side with a smaller number of hops are candidates for a parent node,
A plurality of nodes on the side with a larger number of hops as child nodes monitored by a parent node,
For the first child node connected to one parent node candidate in the communication path among the child nodes, determine the one parent node candidate as a parent node;
Among the child nodes, the second child node connected to the plurality of parent node candidates in the communication path has a small total monitoring cost of the child nodes monitored by each of the plurality of parent node candidates The program according to claim 1, wherein the computer executes a process of determining one of a plurality of parent node candidates as a parent node. The computer according to claim 1, further causing the computer to execute a process of determining the monitoring cost of each of the plurality of nodes based on at least one of the number of information items to be collected, a collection method, a collection complexity, and a collection time. program. Based on the result of the allocation, information indicating the monitoring target node, information indicating the node transmitting the monitoring result, and information indicating the monitoring target item are transmitted to the monitoring node among the plurality of nodes. Processing,
Processing to receive the monitoring result of the node located at the first hop in the communication path and the monitoring result of each node downstream of the node located at the first hop from the node located at the first hop in the communication path. The program according to claim 1 or 2, wherein the program is executed. In a network monitoring method including a plurality of nodes,
Classifying the plurality of nodes according to the number of hops with the monitoring device in the communication path with the minimum number of hops between the monitoring device and the monitoring device connected to the network so as to be communicable;
When setting the monitoring relationship between nodes so that the node with the smaller number of hops monitors the node with the larger number of hops, the monitoring load can be balanced between the nodes with the same number of hops. Execute allocation of monitoring nodes
A monitoring method characterized by that. In a monitoring apparatus that is connected to a network including a plurality of nodes so as to be communicable and monitors the plurality of nodes,
A classification unit that classifies the plurality of nodes according to the number of hops between the monitoring device and a communication path with a minimum number of hops between the monitoring device;
When setting the monitoring relationship between nodes so that the node with the smaller number of hops monitors the node with the larger number of hops, the monitoring load can be balanced between the nodes with the same number of hops. An allocation unit that allocates a monitoring target node,
Including monitoring equipment.

Images