JP5322987B2 - Failure detection device - Google Patents

Description

  The present invention relates to a failure detection device that detects a failure location in a network composed of a plurality of switch devices.

  Conventionally, failure management of a network in which switch devices are connected is performed by an operator monitoring an alarm that is issued from the switch device and indicates a failure or state change that has occurred on the network or device. For example, the switch device detects that the state of the port of its own device has changed from the transfer enabled state to the transfer disabled state as a link down. Then, by issuing the port identifier or state detected by the switch device as an alarm to the management device, the operator can identify the location of the failure and take measures against the failure.

  However, the failure management based on the alarm from the switch device has a problem that the response to the failure is delayed when a silent failure occurs. Here, the silent failure is a state where a failure has occurred but the information cannot be grasped by the operator. Specifically, this occurs because the switch device does not issue an alarm in spite of the failure of customer communication.

  In order to deal with such a silent failure, in conventional failure management, the switch device at the end of the network transmits an Eth-CC frame periodically (for example, one frame per second), so that the network Checking communication. Here, the Eth-CC frame is a test frame for detecting a failure between apparatuses. As a result, when the switch device at the end of the network does not receive the Eth-CC frame within a predetermined time, it is understood that a failure has occurred in the network. Note that ETH-CC is a function for detecting a failure by transmitting and receiving a test frame between devices. 1731 (Ethernet (registered trademark) OAM).

  In Patent Document 1, MIB (Management Information Base) information is acquired by polling from each of adjacent switch devices on the network, and MIB information of adjacent switch devices is compared, so that an abnormality between the switch devices can be detected. A technique for determining the presence or absence is disclosed. The MIB is information that is disclosed by a communication apparatus managed by SNMP so as to notify the state of the own apparatus to the outside, and is defined by RFC1156 and RFC1213.

JP 2006-174375 A

However, because the above-described failure detection by the Eth-CC frame is determined by whether or not the switch device at the end of the network has received the frame, it is not possible to detect the presence or absence of a failure on the network. However, there was a problem that the location of the failure could not be identified.
In addition, when performing failure management using the technique described in Patent Document 1, each of the switch devices configuring the network needs to include a processing unit that analyzes MIB information. Therefore, there is a problem that the method described in Patent Document 1 cannot be used when performing failure management of a network configured by a conventional switch device that does not include a processing unit that analyzes MIB information.

The present invention has been made to solve the above-described problem, and is a failure detection device for detecting a failure location in a network composed of a plurality of switch devices, wherein the plurality of switch devices constituting the network Link information indicating the amount of transmission data to the adjacent switch device, which is another switch device connected adjacent to the switch device, and the amount of received data from the adjacent switch device is read for each of the adjacent switch devices. A read unit; first link information which is link information read by the read unit from a first switch device among the plurality of switch devices; and a second switch device which is an adjacent switch device of the first switch device. And the second link information which is the link information read by the reading unit from The amount of data transmitted from the first switch device to the second switch device indicated by the link information is received from the first switch device in the second switch device indicated by the second link information. When the data amount is larger than a predetermined ratio, or the transmission data amount from the second switch device to the first switch device indicated by the second link information is indicated by the first link information, A failure has occurred in the link between the first switch device and the second switch device when the amount of data received from the second switch device in the first switch device is greater than a predetermined rate. A link failure detection unit that determines that the link failure is detected.

  Further, in the present invention, the link information read by the reading unit includes adjacent switch identification information that identifies the adjacent switch device, and the link failure detection unit is selected from the link information read from the first switch device. The adjacent switch identification information indicating the second switch device is extracted, and the adjacent switch identification information indicates the first switch device out of the link information read from the second switch device. It is desirable to extract what is shown and to determine whether or not the link between the first switch device and the second switch device has failed using the extracted two link information.

  Further, in the present invention, a registration unit that registers the link information read by the reading unit in a link information storage unit, a transmission data amount indicated by the link information that the registration unit previously registered in the link information storage unit, and the registration The difference between the transmission data amount indicated by the link information newly registered in the link information storage unit and the reception data amount indicated by the link information previously registered by the registration unit in the link information storage unit and the registration unit A difference calculation unit that calculates a difference from the received data amount indicated by the link information newly registered in the link information storage unit, and the reading unit periodically reads the link information from each of the switch devices, The difference of the transmission data amount from the first switch device calculated by the difference calculation unit to the second switch device is calculated by the difference calculation unit. Transmission data from the second switch device to the first switch device calculated by the difference calculation unit when the difference is greater than a predetermined ratio by the difference in the received data amount from the first switch device in the switch device When the difference in amount is larger than the difference in the amount of received data from the second switch device in the first switch device calculated by the difference calculation unit, the first switch device and the first switch It is desirable to determine that the link between the two switch devices has failed.

  In the present invention, when the difference between the transmission data amount and the reception data amount of the first switch device calculated by the difference calculation unit is both zero, the port of the switch device is out of order. It is desirable to include a port failure detection unit that determines that

According to the present invention, the failure detection device compares the transmission data amount and the reception data amount included in the link information read from each of the switch devices with the transmission data amount and the reception data amount of the adjacent switch device, thereby determining the failure. Judgment of presence or absence of failure and identification of failure location. Thereby, even when a silent failure occurs in the network, the failure location can be specified.
Further, according to the present invention, it is not necessary for each of the switch devices to include a processing unit that analyzes link information.

It is a schematic block diagram which shows the structure of a network monitoring system provided with the failure detection apparatus by one Embodiment of this invention. It is a figure which shows the information which the MIB information storage part of a failure detection apparatus memorize | stores. It is a flowchart which shows operation | movement of a failure detection apparatus. It is a figure which shows the difference information of a switch apparatus in case a port has failed. It is a figure which shows the difference information of the switch apparatus in case a link has failed.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a schematic block diagram illustrating a configuration of a network monitoring system including a failure detection apparatus according to an embodiment of the present invention.
The network monitoring system includes a failure detection device 100 and a plurality of switch devices 200-1 to 200-5 (hereinafter, the switch devices 200-1 to 200-5 are collectively referred to as the switch device 200). Network.

The failure detection apparatus 100 includes a switch information storage unit 101, a reading unit 102, a timing unit 103, a counting unit 104, a registration unit 105, an MIB information storage unit 106 (link information storage unit), a difference calculation unit 107, and a failure detection unit 108 ( A link failure detection unit and a port failure detection unit).
The switch information storage unit 101 stores the name of the switch device 200 configuring the network and the IP address in association with each other.
The reading unit 102 periodically polls each of the switch devices 200 (for example, once every 10 seconds) to request MIB information (link information) for each port. Read information. Here, the MIB information indicates information such as the cumulative number of received packets (received data amount), the cumulative number of transmitted packets (transmitted data amount), and the link state. In the present embodiment, standard MIB information defined by RFC (Request For Comments) is used as the MIB information. Therefore, according to the present embodiment, information of the same format can be read even if the vendor and device specifications of the switch device 200 are different.
The timer 103 measures the elapsed time from the time when the reading unit 102 performed polling.
The counting unit 104 counts the number of polling performed by the reading unit 102.
The registration unit 105 registers the MIB information read by the reading unit 102 from the switch device 200 in the MIB information storage unit 106 in association with the name of the switch device 200 that is the read source and the number of times of polling counted by the counting unit 104. To do.

FIG. 2 is a diagram illustrating information stored in the MIB information storage unit of the failure detection apparatus.
The MIB information storage unit 106 stores the MIB information read by the reading unit 102. Specifically, as shown in FIG. 2, the MIB information storage unit 106 associates the number of polling executions, the switch device name, and the port number with the cumulative number of received packets, the cumulative number of transmitted packets, the link state, and the adjacent switch. The device name (adjacent switch identification information) and the port number of the adjacent switch device are stored. Here, “adjacent switch device (adjacent switch device)” indicates another switch device 200 that is directly connected to one switch device 200. For example, referring to FIG. 1, the switch device 200-1 and the switch device 200-2 are adjacent to each other, but the switch device 200-1 and the switch device 200-3 are not adjacent to each other. In other words, the case where the first switch device 200 among the plurality of switch devices 200 is connected to the second switch device 200 without passing through the other switch device 200 is shown.

Further, when the failure detection unit 108 determines that there is a failure, the failure detection unit 108 identifies the location where the failure has occurred, and notifies the operator of the location where the failure has occurred by voice or screen.
The polling execution count indicates the number of polling executions by the reading unit 102 counted by the counting unit 104.
The switch device name indicates the name of the switch device 200 from which the reading unit 102 has read the MIB information, and the switch device 200 can be uniquely specified by the name.
The port number is the port number of the switch device 200.
The combination of the number of times of polling, the switch device name, and the port number is the primary key of the MIB information storage unit 106, and the MIB information is uniquely determined by these combinations.

The difference calculation unit 107 calculates the difference between the MIB information read last by the reading unit 102 and the MIB information read last time by the reading unit 102 from the MIB information stored in the MIB information storage unit 106.
The failure detection unit 108 uses the MIB information difference calculated by the difference calculation unit 107 to determine whether there is a failure in a link between two adjacent switch devices 200 or a port of the switch device 200.

In addition, the switching devices 200-1 and 200-5 at the end of the network periodically transmit and receive Eth-CC frames via the switching devices 200-2 to 200-4. At this time, it is assumed that the communication frequency by the Eth-CC frame is higher than the polling frequency by the failure detection apparatus 100. Also, adjacent switch devices 200 transmit and receive LLDP (Link Layer Discovery Protocol) packets to each other. Here, the LLDP packet is a packet for notifying the adjacent node of the name and setting information of the switch device 200. Note that LLDP is a layer 2 protocol for exchanging device identification information, setting information, and the like, and is defined in IEEE 802.1AB.
Further, when receiving the polling signal from the failure detection device 100, the switch device 200 detects the failure detection of the cumulative received packet number, the cumulative transmission packet number, the link state, and the information of the adjacent switch device 200 for each port of the own device. Transmit to device 100.

  In the failure detection apparatus 100 having such a configuration, the reading unit 102 reads MIB information for each adjacent switch apparatus 200 from each of the switch apparatuses 200 configuring the network. The MIB information includes at least the number of transmitted packets to the adjacent switch device 200 and the number of received packets from the adjacent switch device 200. Next, the failure detection unit 108 includes first link information, which is link information read by the reading unit 102 from the first switch device 200 among the plurality of switch devices 200, and a switch device adjacent to the first switch device 200. A failure occurs in the link between the first switch device 200 and the second switch device 200 using the second link information that is the link information read by the reading unit 102 from the second switch device 200. It is determined whether or not.

Specifically, the failure detection unit 108 indicates that the second link information indicates the amount of data transmitted from the first switch device 200 to the second switch device 200 indicated by the first link information. It is determined whether or not the amount of received data from the first switch device 200 in the device 200 is larger than a predetermined ratio. The failure detection unit 108 also includes the first switch device 200 in which the amount of transmission data from the second switch device 200 to the first switch device 200 indicated by the second link information is indicated by the first link information. It is determined whether or not the received data amount from the second switch device 200 is greater than a predetermined rate. When the failure detection unit 108 indicates that at least one of the two determination results is greater than a predetermined ratio, the link between the first switch device 200 and the second switch device 200 has failed. It is determined that
As a result, the failure detection apparatus 100 can identify a failure location even when a silent failure occurs in the network.

Next, the operation of the failure detection apparatus 100 according to the present invention will be described.
FIG. 3 is a flowchart showing the operation of the failure detection apparatus.
First, the reading unit 102 of the failure detection device 100 reads the IP address and name of the switch device 200 that configures the monitoring target network from the switch information storage unit 101 (step S1). Next, the reading unit 102 performs MIB polling by transmitting a polling signal with the read IP address as a destination (step S2). When the reading unit 102 performs polling, the counting unit 104 increments the polling execution count stored in the internal memory by one. Note that the initial value of the number of times of polling stored in the counting unit 104 is “0”. In addition, the time measuring unit 103 starts measuring the elapsed time from the current time (step S3).

When the switch device 200 receives the polling signal from the failure detection device 100, the switch device 200 uses the accumulated number of received packets, the accumulated number of transmitted packets, the link state, and the name and port number of the adjacent switch device 200 as a failure as MIB information. It transmits to the detection apparatus 100.
When the failure detection device 100 receives the MIB information from the switch device 200, the reading unit 102 of the failure detection device 100 reads the MIB information from the switch device 200 (step S4).

  Next, the registration unit 105 reads the polling count stored in the counting unit 104, the name of the switch device 200 from which the MIB information is read, and the port number of the switch device 200 as a main key. The MIB information is registered in the MIB information storage unit 106 (step S5).

Next, the difference calculation unit 107 reads the MIB information newly registered this time and the MIB information registered last time from the MIB information storage unit 106. Specifically, the difference calculation unit 107 reads the MIB information associated with the largest polling execution number and the MIB information associated with the second largest polling execution number from the MIB information storage unit.
Next, the difference calculation unit 107 calculates the difference between the previously registered cumulative transmission packet number and the newly registered cumulative transmission packet number for each combination of the switch device name and the port number. The difference calculation unit 107 calculates the difference between the previously registered cumulative received packet number and the newly registered cumulative received packet number for each combination of the switch device name and the port number (step S6).
The MIB information storage unit 106 stores only the MIB information associated with the polling execution count “1” at the first execution time. Therefore, the difference calculation unit 107 reads the MIB information and outputs it to the failure detection unit 108 without calculating the difference.

Next, the failure detection unit 108 selects one combination of the switch device 200 and the port number, and executes the processes of steps S8 to S13 shown below. Note that the failure detection unit 108 executes the processes in steps S8 to S13 for all combinations of switch devices 200 and port numbers (step S7).
First, the failure detection unit 108 selects the name of the switch device 200 selected in step S7 (first switch device: hereinafter referred to as switch device 200-A) and the port selected in step S7 (hereinafter referred to as port A). The difference in the number of received packets and the difference in the number of transmitted packets associated with the port number are read out from the difference calculation unit 107. Next, the failure detection unit 108 determines whether or not both of the read difference in the number of received packets and the difference in the number of transmitted packets are “0” (step S8).

Here, FIG. 4 is a diagram illustrating the difference information of the switch device when the port is out of order.
If the failure detection unit 108 determines that both the difference in the number of received packets read and the difference in the number of transmission packets are both “0” (step S8: YES), the failure detection unit 108 determines that the port A is faulty. The screen notifies the operator that a failure has occurred in port A (step S9).
The reason why it is determined that a failure has occurred in port A when the difference in the number of received packets and the difference in the number of transmitted packets are both “0” as shown in FIG.
As described above, since the communication frequency of the Eth-CC frame by the switch device 200 is higher than the polling frequency by the failure detection device 100, when the port A of the switch device 200-A is normal, the switch device 200 While the detection apparatus 100 performs polling twice, one or more Eth-CC frame communications are performed.
On the other hand, as described above, when the port A of the switching device 200-A is out of order, it is predicted that the port A will not accept transmission / reception of an Eth-CC frame. Therefore, if the switch device 200-A has never transmitted and received an Eth-CC frame while the failure detection device 100 performs polling twice, the port A of the switch device 200-A has failed. It turns out that there is a high possibility that

On the other hand, when the failure detection unit 108 determines that at least one of the difference in the number of received packets read and the difference in the number of transmission packets is not “0” (step S8: NO), the failure detection unit 108 is adjacent to the switch device 200-A. The difference in the number of received packets and the difference in the number of transmitted packets of the switch device 200 (second switch device: hereinafter referred to as switch device 200-B) are read from the difference calculation unit 107 (step S10).
Hereinafter, a specific method for reading the difference in the number of received packets and the difference in the number of transmitted packets of the switch device 200-B will be described.

  The failure detection unit 108 includes an “adjacent switch device” (that is, the switch device 200-B) and an “adjacent switch device port number” (hereinafter, “the adjacent switch device 200-B”) associated with the combination of the switch device 200-A and the port A. Read from the MIB information storage unit 106. Then, the failure detection unit 108 reads from the difference calculation unit 107 the difference in the number of received packets and the difference in the number of transmission packets associated with the switch device 200-B and the port B. At this time, among the MIB information stored in the MIB information storage unit 106, the MIB information in which “switch device name” indicates the switch device 200-B and “port number” indicates port B is “adjacent switch device”. As shown in FIG.

  That is, the failure detection unit 108 extracts, from the MIB information read from the switch device 200-A, the “adjacent switch device” indicating the switch device 200-B. Further, the failure detection unit 108 extracts information indicating that the “adjacent switch device” indicates the switch device 200-A from the link information read from the switch device 200-B.

  Next, the failure detection unit 108 determines whether or not the difference in the number of transmitted packets of the switch device 200-A is significantly larger than the difference in the number of received packets of the switch device 200-B (step S11). . Here, “remarkably large” indicates that the difference in the number of transmitted packets of the switch device 200-A is larger than the difference in the number of received packets of the switch device 200-B by a predetermined ratio (for example, 10 times) or more. When the failure detection unit 108 determines that the difference between the difference in the number of transmitted packets of the switch device 200-A and the difference in the number of received packets of the switch device 200-B is not significant (step S11: NO), the failure detection unit 108- It is determined whether or not the difference in the number of transmitted packets of B is significantly larger than the difference in the number of received packets of the switching device 200-A (step S12).

When the failure detection unit 108 determines that the difference between the difference in the number of transmitted packets of the switch device 200-B and the difference in the number of received packets of the switch device 200-A is not significant (step S12: NO), the failure detection unit 108- It is determined that there is no failure between A and the switch device 200-B.
On the other hand, the failure detection unit 108 determines that the difference in the number of transmitted packets of the switch device 200-A is significantly larger than the difference in the number of received packets of the switch device 200-B (step S11: YES), or the switch When it is determined that the difference in the number of transmitted packets of the device 200-B is significantly larger than the difference in the number of received packets of the switch device 200-A (step S12: YES), the switch device 200-A and the switch device 200-B It is determined that the link connecting the two has failed (step S13).

FIG. 5 is a diagram illustrating the difference information of the switch device when the link is broken.
Here, as shown in FIG. 5, when the number of transmitted packets from the switch device 200-A to the switch device 200-B is significantly larger than the number of received packets from the switch device 200-A in the switch device 200-B, or When the number of transmitted packets from the switch device 200-B to the switch device 200-A is significantly larger than the number of received packets from the switch device 200-B in the switch device 200-A, the switch device 200-A and the switch device 200- The reason for determining that the link with B has failed will be described.

When the link between the switch device 200-A and the switch device 200-B is normal, even if some packet loss occurs, the number of transmitted packets of the switch device 200-A and the received packets of the switch device 200-B Numbers are expected to be approximate values. Similarly, it is predicted that the number of transmitted packets of the switch device 200-B and the number of received packets of the switch device 200-A are approximate values.
On the other hand, when the link between the switch device 200-A and the switch device 200-B is broken, each switch device 200 is adjacent to each other even though it can transmit a packet to the adjacent switch device 200. Packets cannot be received from the switch device 200. Therefore, it is predicted that the number of transmitted packets will be significantly larger than the number of received packets.
Therefore, when the number of transmitted packets is significantly larger than the number of received packets, it can be seen that there is a high possibility that the link between the switch devices 200 is broken.

If the failure detection unit 108 has not executed the above-described steps S8 to S13 for the combinations of all switch devices 200 and port numbers, the failure detection unit 108 returns to step S7 and sets a new switch device 200-A. Make a selection.
On the other hand, when the failure detection unit 108 executes the above-described processing for all combinations of switch devices 200 and port numbers, the reading unit 102 has a constant time measured by the time measuring unit 103 from step S3. It is determined whether or not the time (periodical MIB polling execution time) has elapsed (step S14). If the reading unit 102 determines that the time measured by the time measuring unit 103 has passed a certain time (step S14: YES), the reading unit 102 returns to step S1 and performs the next MIB polling.

  On the other hand, when the reading unit 102 determines that the time measured by the time measuring unit 103 has not passed the predetermined time (step S14: NO), the failure detection apparatus 100 is operated by an operator or the like, an interruption process, or the like. Then, it is determined whether a processing end request is input from the outside (step S15). If the failure detection device 100 determines that an end request has not been input from the outside (step S15: NO), the failure detection device 100 returns to step S14 and continues to determine whether a fixed time has elapsed. On the other hand, when it is determined that the termination request is input from the outside (step S15: YES), failure detection apparatus 100 ends the process.

As described above, according to the present embodiment, the failure detection device 100 identifies the failure location based on the MIB information read from each of the switch devices 200. Thereby, even when a silent failure occurs in the network, the failure location can be specified.
Further, according to the present embodiment, the failure detection apparatus 100 reads and analyzes MIB information. Therefore, it is not necessary for each of the switch devices 200 configuring the network to include a processing unit that analyzes the MIB information, and it is possible to specify a failure location for the network configured by the conventional switch device.

Further, according to the present embodiment, the switch device 200 holds information on the adjacent switch device 200 by transmitting and receiving LLDP packets, and the failure detection device 100 reads information on the adjacent switch device as MIB information. Thereby, the failure detection apparatus 100 does not need to manage the network configuration in advance, and can specify the failure location by storing at least the name and IP address of the switch device 200 configuring the network. .
Further, according to the present embodiment, the failure detection apparatus 100 identifies a failure location based on a difference in MIB information obtained by two polls. Therefore, it is not necessary to perform accurate synchronization processing for each of the switch devices 200, and failure point identification processing can be easily performed.

As described above, the embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to the above, and various design changes and the like can be made without departing from the scope of the present invention. It is possible to
For example, in the present embodiment, a case has been described in which the reading unit 102 reads information on the adjacent switch device 200 from each of the switch devices 200 and the failure detection unit 108 identifies a failure location based on the information. Not limited to. For example, the switch information storage unit 101 of the failure detection apparatus 100 may store topology information indicating a network configuration, and the failure detection unit 108 may specify a failure location based on the topology information.

  Further, in the present embodiment, a case has been described in which the number of transmission / reception packets within a predetermined time is obtained by calculating a difference in MIB information. However, the present invention is not limited to this. For example, transmission / reception within a predetermined time from information other than MIB information The number of packets may be calculated, or a signal indicating the number of transmitted / received packets within a predetermined time may be directly read from the switch device 200.

  In the present embodiment, the case where the failure detection unit 108 confirms the presence / absence of communication in the switch device 200 using the number of transmission packets and the number of reception packets has been described. However, the present invention is not limited to this, and for example, the failure detection unit 108 May confirm the presence / absence of communication in the switch device 200 using the transmission data amount and the reception data amount.

  The above-described failure detection apparatus 100 has a computer system inside. The operation of each processing unit described above is stored in a computer-readable recording medium in the form of a program, and the above processing is performed by the computer reading and executing this program. Here, the computer-readable recording medium means a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, a semiconductor memory, or the like. Alternatively, the computer program may be distributed to the computer via a communication line, and the computer that has received the distribution may execute the program.

  The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

  DESCRIPTION OF SYMBOLS 100 ... Failure detection apparatus 101 ... Switch information storage part 102 ... Reading part 103 ... Time measuring part 104 ... Counting part 105 ... Registration part 106 ... MIB information storage part 107 ... Difference calculation part 108 ... Failure detection part 200, 200-1-200 -5, 200-A, 200-B ... switch device

Claims (4)

A failure detection device for detecting a failure point in a network composed of a plurality of switch devices,
A transmission data amount from each of the plurality of switch devices constituting the network to an adjacent switch device which is another switch device connected adjacent to the switch device and a reception data amount from the adjacent switch device are indicated. A read unit that reads link information for each adjacent switch device; and
Of the plurality of switch devices, the first link information that is read by the read unit from the first switch device, and the read unit from the second switch device that is an adjacent switch device of the first switch device. The amount of transmission data from the first switch device to the second switch device indicated by the first link information using the second link information that is the link information read by the second link information is the second link information. When the amount of received data from the first switch device in the second switch device indicated by the information is greater than a predetermined ratio, or from the second switch device indicated by the second link information The amount of data transmitted to the switch device is received data from the second switch device in the first switch device, which is indicated by the first link information. A link failure detection unit that determines that a failure has occurred in a link between the first switch device and the second switch device when the amount is greater than a predetermined ratio. Fault detection device. The link information read by the reading unit includes adjacent switch identification information that identifies the adjacent switch device,
The link failure detecting unit extracts, from the link information read from the first switch device, the adjacent switch identification information indicating the second switch device, and from the second switch device. From the read link information, the information indicating that the adjacent switch identification information indicates the first switch device is extracted, and using the extracted two link information, the first switch device and the second switch are extracted. The failure detection device according to claim 1, wherein it is determined whether or not a link with the device is broken. A registration unit for registering link information read by the reading unit in a link information storage unit;
The difference between the transmission data amount indicated by the link information previously registered by the registration unit in the link information storage unit and the transmission data amount indicated by the link information newly registered by the registration unit in the link information storage unit, and the registration unit A difference calculation unit that calculates a difference between the received data amount indicated by the link information previously registered in the link information storage unit and the received data amount indicated by the link information newly registered by the registration unit in the link information storage unit;
With
The reading unit periodically reads the link information from each of the switch devices,
The difference of the transmission data amount from the first switch device calculated by the difference calculation unit to the second switch device is obtained from the first switch device in the second switch device calculated by the difference calculation unit. The difference calculation unit calculates a difference in transmission data amount from the second switch device to the first switch device calculated by the difference calculation unit when the difference is greater than a predetermined ratio by the difference in received data amount The link between the first switch device and the second switch device fails when the first switch device is larger than the difference in the amount of received data from the second switch device by a predetermined ratio or more. The failure detection device according to claim 1, wherein the failure detection apparatus determines that the failure has occurred. Port failure detection unit that determines that the port of the switch device is faulty when the difference between the transmission data amount and the reception data amount of the first switch device calculated by the difference calculation unit is both zero The failure detection apparatus according to claim 3, further comprising:

Images