JP6045410B2 - Network device and state change notification method - Google Patents

Description

  The present invention relates to a network device and a state change notification method for detecting a state change of a monitoring target and notifying a state value indicating a state after the change.

The network device is a device in which an SNMP (Simple Network Management Protocol) agent function is incorporated. The SNMP agent function detects a state change in the device by polling by software, and uses Trap which is an event notification function. This is a function for notifying the remote SNMP manager of the contents of the state change via the network.
SNMP is an application layer protocol for monitoring and controlling communication devices (for example, routers, switches, servers, etc.) connected to a TCP / IP network via the network. In 1990, IETF (Internet) (Engineering Task Force).
Specifically, SNMP defines communication between an SNMP management system in which an SNMP manager operates and an SNMP management target system in which an SNMP agent operates.

An SNMP management system generally includes a management application and an SNMP manager.
On the other hand, an SNMP management target system generally includes a management target resource, a management target object for managing a data structure called MIB (Management Information Base) in which the management target resource is abstracted, and an SNMP agent. .
Here, the management application, the managed object, the SNMP manager, and the SNMP agent are realized by software.
In particular, the SNMP manager and the SNMP agent are called protocol stacks and are often part of the OS.

SNMP is defined by several types of PDUs (Protocol Data Units). In order to reduce the amount of communication between the SNMP management system and the SNMP management target system, the content of the state change on the SNMP management target system is called Trap. In many cases, the SNMP agent notifies the SNMP manager using the type of PDU (see, for example, Patent Document 1).
Specifically, when the managed object of the SNMP managed system monitors the managed resource by polling and detects a change in the status of the managed resource, the status value (managed resource status is stored in the MIB). In many cases, the SNMP agent employs a method in which the SNMP agent gives the trap the status value updated by the managed object and notifies the SNMP manager.

In such a method, the time detection granularity of state change is determined by the polling cycle for the managed resource in the managed object. To reduce the time detection granularity of state change, increase the number of polls per unit time. It is necessary to let
However, in the SNMP management target system, not only the management target object operates, but also applications other than the management target object often operate, and the CPU resource of the SNMP management target system competes among a plurality of applications. .
For this reason, when the number of polling times per unit time is increased, the CPU resource that requires the managed object increases, and the operation of applications other than the managed object may become difficult.
This tendency is remarkable especially when the SNMP management target system is configured by an embedded system having a poor calculation capability.

Therefore, since it is not easy to shorten the polling period for the managed resource, conventionally, the abandoning of the polling period for the managed resource is abandoned, and the detection of the polling period less than the allowable use amount of the CPU resource is performed. I am trying not to.
Alternatively, the representative value within the polling period is used by support such as latching of the state change by the hardware constituting the management target resource.

If you give up shortening the polling cycle for a managed resource, the CPU resources used by applications other than the managed object will increase. Needless to say, there are situations where the managed object cannot detect a change in the status of the managed resource. To increase.
For example, in the case where the state takes a binary value of “0” and “1”, the state is “0” at the first polling timing, and the state is “0” → “1” before the next polling timing arrives. → Under the situation of changing like “0”, the managed object cannot detect the status change of the managed resource.

On the other hand, when the state change is latched by the hardware that constitutes the management target resource, it is specified in advance which state is “0” or “1”, and when the value becomes significant, The state change is latched by the wear.
For example, if the value “1” is defined to be significant, the state is “0” at the first polling timing, and the state is “0” → “1” → “0” until the next polling timing arrives. Under the situation of changing as “”, “1” is latched, so that the managed object can detect “0” → “1” at the next polling timing.
However, when a change occurs a plurality of times such as “0” → “1” → “0” → “1” during the polling period, for example, the states are “0”, “1”, and “2”. When a change such as “0” → “1” → “2” → “1” occurs, the intermediate state is lost, so all the state changes of the managed resources Cannot be detected.

JP 2000-134203 A (paragraph number [0024], FIG. 1)

Mark A. Miller, "SNMP Internetwork Management", published by Shosuisha Inc., January 20, 2001, pages 124-128

  Since the conventional network device is configured as described above, it is difficult to shorten the polling cycle for the managed resource in the managed object, and a situation in which the state change of the managed resource cannot be detected occurs. There was a problem such as that.

  The present invention has been made to solve the above-described problems, and provides a network apparatus and a state change notification method capable of notifying the state change of a managed resource without causing an increase in CPU resources. For the purpose.

The network device according to the present invention writes a state value indicating a state after the change into the state history table every time the state change is detected by the state change detection unit that detects the state change of the monitoring target. In addition, the state value writing unit for storing the state value writing history for the state history table in the state history table, and the writing history stored in the state history table for each predetermined period, If there is a new write, the state value reading unit that reads a new state value from the state history table and the state value read by the state value reading unit are notified to the management system. A state value notification unit, and the state history table is configured by a ring buffer capable of storing a maximum of N state values. If the number of new state values written to the queue exceeds N and some state values are overwritten with other state values before they are read, the state value reading unit is currently stored in the state history table. The N state values read are read, and the state value notification unit notifies the management system of the N state values read by the state value reading unit .

According to the present invention, the write history stored in the state history table is referred to every predetermined cycle to check whether there is a new write of the state value. state value reading unit from the history table read new state value, and a status value notification unit configured to notify the management system state values read by the state value reading unit, the state history table of the N state values at the maximum Consists of ring buffers that can be stored, the number of new writes of state values to the state history table exceeds N, and some state values are overwritten with other state values before they are read In this case, the state value reading unit reads the N state values currently stored in the state history table, and the state value notification unit manages the N state values read by the state value reading unit. Since it is configured to notify the stem, without increasing the CPU resource, there is an effect that can be notified without fail status change monitor.

BRIEF DESCRIPTION OF THE DRAWINGS It is a block diagram which shows the network management system with which the network device by Embodiment 1 of this invention is applied. It is a schematic diagram which shows the structural example in case the SNMP management object system (network apparatus) by Embodiment 1 of this invention is a station side optical line termination | terminus apparatus in a media converter. It is a flowchart which shows the processing content (state change notification method) of the network device by Embodiment 1 of this invention. It is explanatory drawing which shows an example of the status history table. It is explanatory drawing which shows the reading log | history of the state value stored in MIB17. It is explanatory drawing which shows an example of the status history table. It is a block diagram which shows the network device by Embodiment 3 of this invention.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a network management system to which a network device according to Embodiment 1 of the present invention is applied.
In FIG. 1, an SNMP management target system 1 is connected to an SNMP management system 3 via an IP network 2, detects a state change in the apparatus by polling by software, and uses Trap, which is an event notification function, This is a network device having an SNMP agent function for notifying the SNMP management system 3 of the contents of the state change.
The SNMP management system 3 is a device that manages the contents of the state change notified from the SNMP management target system 1.

The management target resource 11 of the SNMP management target system 1 has a state change detection unit 12, a state value writing unit 13, and a state history table 14.
The state change detection unit 12 monitors the state of the monitoring target and detects the state change (for example, a failure of a communication line connected to the managed resource 11 or a failure of a communication device constituting the managed resource 11). Perform the process.

Each time a state change is detected by the state change detection unit 12, the state value writing unit 13 writes a state value indicating the state after the change into the state history table 14, and writes a state value writing history to the state history table 14 as a state. The process of storing in the history table 14 is performed.
The state history table 14 includes a ring buffer capable of storing up to N state values.
The ring buffer is composed of N state value storage areas. When the state value is written in order from the first state value storage area and the state value is written up to the last state value storage area, the first state value is stored again. State values are written (overwritten) in order from the state value storage area.
Further, the state history table 14 stores, as a state value writing history, final write position information indicating a state value storage area in which the state value was last written (a state value storage area in which the latest state value is written). .

The SNMP management target object 15 of the SNMP management target system 1 has a state value reading unit 16 mounted thereon.
The state value reading unit 16 compares the write history stored in the state history table 14 of the management target resource 11 with the read history stored in the MIB 17 for each polling cycle instructed by the SNMP management system 3, It is confirmed whether or not a new value has been written. If there is a new write, a process of reading a new state value from the state history table 14 is performed.
The MIB 17 is a memory that stores a state value reading history by the state value reading unit 16.
Note that the last read position information indicating the state value storage area where the state value was read last is stored as the state value reading history.

The SNMP agent 18 of the SNMP management target system 1 is equipped with a state value notification unit 19 and terminates the SNMP protocol.
The state value notification unit 19 performs processing of notifying the SNMP management system 3 via the IP network 2 of the state value read by the state value reading unit 16 using Trap which is an event notification function.

In the example of FIG. 1, the SNMP management target system 1 is configured by a computer, each of the management target resource 11, the SNMP management target object 15, and the SNMP agent 18 is realized by software, and the CPU of the SNMP management target system 1 is Although it is assumed that the management target resource 11, the SNMP management target object 15 and the SNMP agent 18 execute processing programs, the management target resource 11, the SNMP management target object 15 and the SNMP agent 18 are dedicated. It may be configured by hardware.
For example, the state change detection unit 12, the state value writing unit 13, the state value reading unit 16, and the state value notification unit 19 are configured by a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like, and a state history table 14 and MIB 17 may be constituted by a storage device such as a RAM or a hard disk.

The SNMP manager 21 of the SNMP management system 3 terminates the SNMP protocol, receives the status value notified from the SNMP agent 18 of the SNMP management target system 1, and outputs the status value to the management application 22. .
The management application 22 is an application that manages the state value output from the SNMP manager 21.

  In the example of FIG. 1, it is assumed that each of the SNMP manager 21 and the management application 22 of the SNMP management system 3 is realized by software, but each of the SNMP manager 21 and the management application 22 has dedicated hardware ( For example, it may be configured by a semiconductor integrated circuit on which a CPU is mounted or a one-chip microcomputer.

FIG. 2 is a schematic diagram showing a configuration example in a case where the SNMP management target system (network device) according to Embodiment 1 of the present invention is an optical line terminal (OLT) in the media converter.
In FIG. 2, the OLT 31 corresponds to the SNMP management target system 1 in FIG. 1, and includes OSUs (Optical Subscriber Units) 32 and 33, a multiplexing module 34, and a control unit 35.
The OSUs 32 and 33 are line terminators connected to transmission lines 36 and 37 such as Ethernet (registered trademark), for example, and optical signals are transmitted to the ONUs (Optical Network Units) 38 and 39 via the transmission lines 36 and 37. Send and receive.

For example, when the OSUs 32 and 33 receive the optical signals (frames) transmitted from the ONUs 38 and 39, the multiplexing module 34 multiplexes these frames and transmits them to a host device (not shown), while the frames transmitted from the host device. Is received, the frame is separated, and the separated frames are transmitted to the ONUs 38 and 39.
The control unit 35 is a processing unit that controls the processing contents of the OSUs 32 and 33 and the multiplexing module 34.
The ONUs 38 and 39 are subscriber side optical line terminators, and transmit and receive optical signals to and from the OSUs 32 and 33 via the transmission paths 36 and 37.

In the example of FIG. 2, an example in which there are two OSUs, ONUs, and transmission lines is shown. However, this is only an example, and one or three or more may be used. .
Note that the media converter uses one transmission line for one OLT and one ONU.
When the SNMP management target system 1 in FIG. 1 is the OLT 31 in FIG. 2, the OSUs 32 and 33 and the multiplexing module 34 are responsible for the management target resource 11, and the control unit 35 is responsible for the SNMP management target object 15, MIB 17 and SNMP agent 18. .

Next, the operation will be described.
FIG. 3 is a flowchart showing the processing contents (state change notification method) of the network device according to the first embodiment of the present invention.
The management application 22 of the SNMP management system 3 instructs the SNMP management target system 1 to detect a state change, read a state value, and the like via the SNMP manager 21.
The SNMP management target system 1 performs state change detection, state value reading, and the like under the instruction of the SNMP management system 3.

Specifically, it is as follows.
The state change detection unit 12 of the management target resource 11 monitors the state of the monitoring target, and if the state of the monitoring target has changed, detects the state change (step ST1 in FIG. 3).
For example, when the SNMP management target system 1 is the OLT 31 shown in FIG. 2, a failure in the transmission lines 36 and 37, a failure in the OSUs 32 and 33, and the like are detected.

Each time the state change detection unit 12 detects a state change, the state value writing unit 13 performs a process of writing a state value indicating the state after the change into the state history table 14 (step ST2).
Here, FIG. 4 is an explanatory diagram showing an example of the state history table 14.
In the example of FIG. 4, the state history table 14 includes five state value storage areas for storing state values, and also includes an area for storing state value writing history.
As the state value write history, the last write position information indicating the state value storage area where the state value was last written (the state value storage area where the latest state value is written) is stored in the state history table 14. In this case, the state value writing unit 13 writes the state value indicating the changed state in the state value storage area next to the state value storage area indicated by the final writing position information.
For example, if the state value storage area indicated by the final write position information is the state value storage area (2), the changed state is written to the state value storage area (3), and the state value storage area indicated by the final write position information is If it is the state value storage area (3), the changed state is written in the state value storage area (4).
If the state value storage area indicated by the final write position information is the state value storage area (5), the changed state is written to the state value storage area (1).

When the state value writing unit 13 writes the state value indicating the state after the change into the state history table 14, the state value writing unit 13 updates the final writing position information, which is the state value writing history stored in the state history table 14 (step ST3). ).
That is, the state value writing unit 13 stores the last write position information, which is a state value write history stored in the state history table 14, as a state value storage area (a plurality of state values written this time). In this case, the position information is updated to the position information indicating the last written state value storage area.

The state value reading unit 16 of the SNMP management target object 15 stores the write history stored in the state history table 14 of the management target resource 11 and the MIB 17 at every polling cycle instructed by the SNMP management system 3 (step ST4). The read histories that have been read are compared to check whether or not there is a new write of the state value (step ST5).
Here, FIG. 5 is an explanatory diagram showing the reading history of the state values stored in the MIB 17.
When the state value reading unit 16 stores the last read position information indicating the state value storage area where the state value was last read as the state value reading history, the state value storage indicated by the last read position information is stored. If the area is different from the state value storage area indicated by the final writing position information stored in the state history table 14, it is determined that there is a new writing of the state value.

When the state value reading unit 16 confirms that a new state value is written, the state value reading unit 16 reads a new state value from the state history table 14 and outputs the state value to the SNMP agent 18 (step ST6).
For example, if the state value storage area indicated by the final read position information is the state value storage area (2) and the state value storage area indicated by the final write position information is the state value storage area (4), the state history table 14 adds a new one. As state values, the state values written in the state value storage areas (3) to (4) are read.
If the state value storage area indicated by the final read position information is the state value storage area (2) and the state value storage area indicated by the final write position information is the state value storage area (5), the state history table 14 As state values, the state values written in the state value storage areas (3) to (5) are read.
It should be noted that the state value storage area indicated by the final read position information is the state value storage area (3) and the state value storage area indicated by the final write position information is the state value storage area (3). If the state value storage areas indicated by the information are the same, it is determined that a new state value has not been written, and the state value reading process is not performed.

When the state value reading unit 16 reads a new state value from the state history table 14, the state value reading unit 16 updates the last read position information, which is the state value reading history stored in the MIB 17 (step ST7).
That is, the state value reading unit 16 stores the last read position information, which is the state value reading history stored in the MIB 17, in the state value storage area where the current state value has been read this time (if a plurality of state values have been read, The position information indicating the state value storage area read in (2) is updated.

  When the state value notification unit 19 of the SNMP agent 18 receives the state value from the state value reading unit 16, the state value is sent to the SNMP manager of the SNMP management system 3 via the IP network 2 using Trap, which is an event notification function. 21 is notified (step ST8).

The SNMP manager 21 of the SNMP management system 3 receives the state value notified from the SNMP agent 18 of the SNMP management target system 1 and outputs the state value to the management application 22.
The management application 22 manages the state value output from the SNMP manager 21.

As is apparent from the above, according to the first embodiment, the write history stored in the status history table 14 and the read history stored in the MIB 17 are stored for each polling cycle instructed from the SNMP management system 3. In comparison, it is confirmed whether or not there is a new writing of the state value. If there is a new writing, a state value reading unit 16 for reading a new state value from the state history table 14 is provided, and a state value notification unit 19 is provided. Is configured to notify the SNMP management system 3 via the IP network 2 of the state value read by the state value reading unit 16 using Trap, which is an event notification function. Without incurring an increase in resources, it is possible to notify the SNMP management system 3 of the state change of the monitoring target without omission. Achieve the results.
That is, even in a situation where a plurality of state changes occur at intervals shorter than the polling cycle, the SNMP management system 3 can be notified of the state values indicating the contents of the plurality of state changes without omission, without particularly shortening the polling cycle. There are effects that can be achieved.

Embodiment 2. FIG.
In the first embodiment, an example in which the state history table 14 is configured by a ring buffer capable of storing a maximum of N state values has been described. In this case, from the state value storage area at the head of the ring buffer, When the status values are written in order and the status values are written up to the last status value storage area, the status values are written again in order from the top status value storage area.
Therefore, when (N + 1) or more state values are written in the state history table 14 between a certain polling timing and the next polling timing, some state values are read by the state value reading unit 16. Lost by being overwritten with other status values before.
For example, when (N + 2) state values are written in the state history table 14 between a certain polling timing and the next polling timing, before the two state values are read by the state value reading unit 16. Lost by being overwritten with other state values.

In the second embodiment, when the state value reading unit 16 writes (N + 1) or more state values to the state history table 14 between a certain polling timing and the next polling timing, (N + 1) It is detected that more than one state value has been written, and N state values currently stored in the state history table 14 are read.
Here, detection that (N + 1) or more state values are written is performed as follows, for example.

In the second embodiment, an example in which the number of state value storage areas in the state history table 14 is 2 ⁿ −1 will be described. This is an example of N = 2 ⁿ −1.
FIG. 6 is an explanatory diagram showing an example of the state history table 14. The state history table 14 of FIG. 6 is an example of n = 4 and includes eight state value storage areas (0) to (7).
It is assumed that the final write position information stored in the state history table 14 is constituted by a read reset (ReadReset).

When the state change detection unit 12 detects a state change, the state value writing unit 13 of the managed resource 11 writes the state value indicating the state after the change to the state history table 14 as in the first embodiment. Unlike the first embodiment, the last write position information stored in the state history table 14 is always updated to a value obtained by incrementing the last write position.
Specifically, for example, when nine state changes are detected between a certain polling timing and the next polling timing, the state value storage area (0) at the head of the ring buffer in FIG. After the state value is written and the state value is written up to the last state value storage area (7), the state value is written again in the head state value storage area (0).
At this time, in the first embodiment, the final write position information is updated to indicate the state value storage area (0).
On the other hand, in the second embodiment, since the final write position information is incremented 8 (= 9-1) times, it is updated to indicate the state value storage area (8).

The state value reading unit 16 reads the last write position information stored in the state history table 14, and if the logical product of the read value and “inverted (2 ⁿ −1)” is non-zero, the state value reading unit 16 It is determined that the value writing has already made a round in the state value storage area of the ring buffer. That is, it is determined that state values exceeding the number (2 ⁿ −1) of state value storage areas in the state history table 14 have been written between a certain polling timing and the next polling timing.
When the state value reading unit 16 determines that the writing of the state value has already made a round of the state value storage area of the ring buffer, the state value reading unit 16 stores (2 ⁿ −1) state values currently stored in the state history table 14. Read.
When the state value reading unit 16 reads (2 ⁿ −1) state values, the state value notification unit 18 uses the event notification function Trap to convert (2 ⁿ −1) state values to IP. Notification is made to the SNMP manager 21 of the SNMP management system 3 via the network 2.

Thereby, even if a state value exceeding the number (2 ⁿ −1) of state value storage areas in the state history table 14 is written between a certain polling timing and the next polling timing, Only the (2 ⁿ −1) state values can be notified to the SNMP manager 21 of the SNMP management system 3.

Embodiment 3 FIG.
FIG. 7 is a block diagram showing a network device according to Embodiment 3 of the present invention. In the figure, the same reference numerals as those in FIG.
In the third embodiment, it is assumed that the SNMP management target system 1 that is a network device is configured as a complex of two units.
That is, the main signal conduction card 41 (first card) as one unit and the monitoring control card 42 (second card) as one unit are connected by a mechanism capable of hot-swapping. Assume something. Since the mechanism capable of hot-swapping is a known mechanism, detailed description thereof is omitted.

In the example of FIG. 7, the main signal continuity card 41 implements the management target resource 11, and the monitoring control card 42 implements the SNMP management target object 15, the MIB 17, and the SNMP agent 18.
When the SNMP management target system 1 in FIG. 7 is the OLT 31 in FIG. 2, the OSUs 32 and 33 and the multiplexing module 34 bear the management target resource 11 in the main signal conduction card 41, and the control unit 35 performs monitoring control. It is responsible for the SNMP managed object 15, MIB 17 and SNMP agent 18 in the card 42.

The basic operation of the SNMP management target system 1 is the same as that of the first embodiment, but in this third embodiment, the main signal conduction card 41 and the monitoring control card 42 are hot-plugged from the network device 1. Connected with possible mechanisms. Alternatively, the connection between the management target resource 11 and the SNMP management target object 15 may also allow hot-swap.
Moreover, each card | curd can detect attachment / detachment mutually.
For example, in a system using a media converter, hot insertion / removal may be allowed between cards in order to increase or decrease the number of OSUs or to add a multiplexing function. . That is, each card can be attached to and detached from a plurality of slots of the OLT motherboard.

In the third embodiment, when the state value reading unit 16 reads a new state value from the state history table 14 and detects that the main signal continuity card 41 is removed from the network device 1, a new state value is obtained. Stop reading the status value.
As a result, it is possible to prevent notification of a state value read in an unauthorized situation where the main signal conduction card 41 is not connected to the network device 1.

In addition, when the monitoring control card 42 is inserted into the network device 1 after the monitoring control card 42 is removed from the network device 1, the state value reading unit 16 stores it in the state history table 14 in the immediately following polling cycle. By referring to the last written position information, the state value storage area where the state value was last written is specified, and the monitoring control card 42 is removed from the network device 1 and inserted into the network device 1 During this period, even if there are a plurality of new state values written to the state history table 14, only the last written state value is read.
When the state value reading unit 16 reads only the last written state value, the state value notifying unit 18 uses the event notification function Trap to send only the last written state value via the IP network 2. Notify the SNMP manager 21 of the SNMP management system 3.

As a result, while the supervisory control card 42 is removed from the network device 1, a state change event that cannot be confirmed as being correct is skipped to prevent an uncertain state value from being notified to the SNMP management system 3. be able to.
In this case, the state value written before the last written state value is discarded.

Note that a system using a media converter is generally provided for companies. In the service for enterprises, it is required to manage so that the conduction start time is minimized, and tracking the detailed state change of the management target resource 11 is important as a predictive detection of a failure that occurs later.
When a sign is detected, a detailed diagnosis is required, and cost is required for such a diagnosis, so unnecessary signs need to be eliminated.
Therefore, the SNMP management target system 1 disclosed in the third embodiment is suitable for a system using a media converter.
The third embodiment can be said to be a trap issuing method considering prevention of reissuance of traps that have already been output when the monitoring control card 42 is replaced.

  In the present invention, within the scope of the invention, any combination of the embodiments, or any modification of any component in each embodiment, or omission of any component in each embodiment is possible. .

  DESCRIPTION OF SYMBOLS 1 SNMP management object system (network apparatus), 2 IP network, 3 SNMP management system, 11 Management object resource, 12 State change detection part, 13 State value writing part, 14 State history table, 15 SNMP management object, 16 State value Reading unit, 17 MIB, 18 SNMP agent, 19 state value notification unit, 21 SNMP manager, 22 management application, 31 OLT, 32, 33 OSU, 34 multiplexing module, 35 control unit, 36, 37 transmission path, 38, 39 ONU, 41 Main signal conduction card (first card), 42 Supervisory control card (second card).

Claims (6)

A state change detection unit for detecting a state change of a monitoring target;
Each time a state change is detected by the state change detection unit, a state value indicating the state after the change is written in the state history table, and a state value writing history for the state history table is stored in the state history table A value writing unit;
Every predetermined cycle, the writing history stored in the state history table is referred to check whether there is a new writing of the state value. If there is a new writing, a new one is added from the state history table. A state value reading unit for reading a state value,
A status value notifying unit for notifying the management system of the status value read by the status value reading unit ,
The state history table is configured by a ring buffer capable of storing up to N state values, and the number of new writes of state values to the state history table exceeds N, and some state values Is overwritten with other state values before the is loaded,
The state value reading unit reads N state values currently stored in the state history table,
The state value notification unit notifies the management system of the N state values read by the state value reading unit.
A network device. The state change detection unit, the state value writing unit and the state history table is mounted on the first card, the state value reading unit and the state value notification unit are mounted on a second card,
The first card and the second card, according to claim 1 Symbol placement of network devices, characterized in that it is connected capable of hot plug mechanism. The state value reading section when reading a new state value from the state history table, when the first card is removed from the network device, characterized in that stops loading the new state value The network device according to claim 2 . The state value reading unit, after the second card is removed from the network device, when the second card is inserted into the network device, a new write multiple state values for the state history table Even if there are, only the last written state value is read.
The state value notification unit, the network apparatus according to claim 2 or claim 3, wherein a state value read by the state value reading unit and notifies to the management system. The state value notification unit uses the Trap is an event notification function, claim from claim 1, wherein the notifying the state value read by the state value reading unit to the management system via the network network device according to any one of the four. A state change detection unit that detects a state change of a monitoring target;
Each time a state change is detected in the state change detection processing step, the state value writing unit writes a state value indicating the state after the change to the state history table, and writes the state value writing history to the state history table. A state value writing process step to be stored in the state history table;
The state value reading unit refers to the writing history stored in the state history table for each predetermined period to check whether there is a new writing of the state value, and if there is a new writing, A state value reading process step for reading a new state value from the state history table;
A state value notifying unit for notifying the management system of the state value read in the state value reading processing step ;
The state history table is configured by a ring buffer capable of storing up to N state values, and the number of new writes of state values to the state history table exceeds N, and some state values Is overwritten with other state values before the is loaded,
The state value reading unit reads N state values currently stored in the state history table;
A state change notifying method comprising: notifying the management system of the N state values read by the state value reading unit .

Images