JPH09321789A - Network system with duplexed routers and fault countermeasure method for it - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the availability of a network and to avoid the increase of the load on the CPU of a computer by reducing the time for switching a route from an active router to a stand-by router to be shorter than that at the time of using RIP(routing information protocol) in a network system with duplexed routers. SOLUTION: One of the duplexed routers is set to be an active router and the other is set to be a stand-by router and they are assigned with respectively the same logic address (IP(internet protocol) address) and different physical addresses (MAC(medium access control) addresses), and the communication controller 11 of a computer (a1) on a network holds this logical address and physical address. Then when the communication controller 11 communicates with another computer (b-1) through routers α and β, the physical address of the active router is used for communication. On the other hand, when an ARP (address resolution protocol) response message is not transmitted to the active router α in spite of requesting it from the active router α, the router is switched to the stand-by router β by judging the fault to be generated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a network system in which a router is duplicated and a failure countermeasure method for a network system in which a router is duplicated, and is a system for enhancing reliability by duplicating a router. Moreover,
The present invention relates to a network system suitable for use in a network system that needs to be operated without interruption.

[0002]

2. Description of the Related Art In recent years, there has been an increasing number of cases in which a router is used as a device for relaying data between networks and selecting a route when communicating between computers connected to different networks.

Generally, as a protocol for a router to select a route (referred to as "routing protocol"), static routing, which holds routing information as static information, is dynamically exchanged, and routing is performed based on it. There are two types of methods called dynamic routing.

By the way, when two networks are connected by such a router, if one router connects the networks, if the router fails, communication between the networks can be performed. Disappear.

In order to improve the reliability of communication between networks against such a failure of the router, the two networks are connected to each other, and normally, communication between networks is performed through one router. In some cases, if a failure occurs in this router and it becomes impossible to relay the message, the route may be switched to the other router.

As described above, when the router is duplicated and a spare router is used when there is a failure,
Dynamic routing is inevitably adopted as the routing protocol.

A typical example of this dynamic routing is RIP (Routing Information Pr
otocol). RIP is a protocol that defines a route control method in which each computer and router connected to a network system exchanges route information with each other, retrieves only the shortest route information from the exchanged route information, and retains it. The router performs message communication by selecting a route from its own route information. In addition,
The RIP standard is defined in RFC (Request For Comments) 1058 issued by the United States NIC (Network Information Center), which is one of Internet-related organizations.

In RIP, the router has one cycle (30 seconds).
RIP to other routers connected to the network for each
The route information called a message is distributed. Then, each router selects an optimum route and performs routing.

However, even when the router fails, the information is held for six cycles (180 seconds) after receiving the RIP message, and the operation is the same as in the normal state. Therefore, even if there is a spare router and switching is performed at the time of failure, switching is not performed for at least 6 cycles (180 seconds), and communication cannot be performed during that period.

[0010]

The RIP described in the above prior art is the most widely used routing protocol to date as the dynamic routing of TCP / IP. However, even if the router is duplicated, it takes at least 180 seconds until the route is switched after the failure of the router, and the communication is interrupted during that time.

As a method of shortening the time from the occurrence of a failure in the router to the switching of the route and the time of interruption of communication, for example, the following method is considered. .

First, the logical addresses of all the routers on the network are associated with the destination network address (the one address is associated with the entire network) that arrives at the router, and the computer is preliminarily calculated. Register in the memory of. For one destination network, select a specific one from the registered routers.
Messages are relayed using two routers, and at the same time, from the computer to this router, the ICMP echo request message (TCP standardized by RFC792) is sent.
/ IP, which is defined by a command called Ping) is periodically transmitted.

Then, the router waits for a certain period of time to send an ICMP echo reply message in response to the ICMP echo request message from the router.
If the P echo reply message is not received, it is determined that a failure has occurred in that router, the relay router is switched to another router registered in the memory in advance, and message communication is continued.

According to this method, for example, the ICMP echo request message is transmitted every 15 seconds, and after the transmission, the ICMP echo reply message is waited for 10 seconds. If the ICMP echo reply message is not received even after waiting 10 seconds, it is determined that the router that has sent the ICMP echo request message has failed, and the relay router is switched to another router. In this way, the route switching time will be 25
Can be finished in seconds.

However, this method has a problem that the load of the CPU of the computer is increased because the CPU interprets and executes the communication program mounted on the memory of the computer.

The present invention has been made to solve the above problems, and its purpose is to switch a route from an active router to a backup router in a network system having dual routers. Switching time is R
A network system in which the router is duplicated and a failure countermeasure method for the network system in which the router is duplicated, which is shorter than when IP is used, improves the network availability, and does not increase the CPU load of the computer To provide.

[0017]

In order to achieve the above object, the configuration of the invention relating to the network system in which the router of the present invention is duplicated is such that one or more networks are connected via the router, and In a network system in which the router is duplicated, a computer connected to the network has a communication control device, and the computer communicates by the communication control device that it has, and the duplication is also performed. The routers are assigned the same logical address and different physical addresses, one of which is the working router and the other one of which is the backup router, and the communication control device uses the logical address of the duplicated router. This communication control device has means for holding the physical address and
When communicating with another computer via the router, the physical address of the working router is used for communication, and a request message for periodically requesting the physical address from the logical address is transmitted to the duplicated router. Whether or not a response message, which is a response to the transmitted request message, is transmitted from the router is monitored for each duplicated router, and when the response message is not transmitted from the router, Assuming that the router has failed, if it is determined that the working router has failed, the physical address of the backup router is used when communicating with other computers via the router. It is intended to switch to.

More specifically, in the network system in which the router is duplicated, the communication control device is
A logical address and a physical address assigned to the duplicated router are received from a computer and set as means for holding the logical address and the physical address of the duplicated router. is there.

Further in detail, in a network system in which a router is duplicated, the communication control device is
The logical address and physical address assigned to the duplicated router are obtained from a response message which is a response to a request message for obtaining a physical address from the logical address transmitted to each duplicated router. , And these are set as means for holding the logical address and physical address of the duplicated router.

More specifically, in a network system in which the routers are duplicated, when a failure occurs in the working router, the spare router is set as a new working router, and the working router is replaced with a new one. As a backup router, the communication control device has means for holding a logical address and a physical address of a duplicated router, and a logical address and a physical address assigned to the new working router and the new backup router. The address and the address are respectively set.

More specifically, in the network system in which the router is duplicated, the physical address is a MAC (Media Access Control) address and the logical address is an IP (Internet Protocol).
l) address, and the request message for the physical address from the logical address is ARP (Addres
s Resolution Protocol) request message, and the response message as a response is an ARP response message.

In order to achieve the above object, the configuration of the invention relating to the failure countermeasure method of the network system in which the router is duplicated according to the present invention is such that one or more networks are connected via the router and In a failure countermeasure method for a network system having dual routers, a computer connected to the network has a communication control device,
This computer communicates by means of a communication control device each of which has the same logical address and a different physical address, with one of the duplicated routers being a working router and the other being a standby router. Is assigned to the communication control device, and the communication control device has means for holding a logical address and a physical address of the duplicated router. When the communication control device communicates with another computer via the router, Communicate using the physical address of the working router,
A request message for periodically requesting a physical address from a logical address is sent to a duplicated router, and whether or not a response message, which is a response to the sent request message, is sent from the router, Each duplicated router is monitored, and when a response message is not sent from the router, it is determined that a failure has occurred in that router. If it is determined that the working router has a failure, When communicating with another computer via the router, the physical address of the spare router is used for communication.

More specifically, in the failure countermeasure method for a network system in which the router is duplicated, the communication control device receives a logical address and a physical address assigned to the duplicated router from a computer. Then, they are set as means for holding the logical address and physical address of the duplicated router.

More specifically, in a failure countermeasure method for a network system in which a router is duplicated, the communication control device duplicates a logical address and a physical address assigned to the duplicated router. Obtained from the response message which is a response to the request message for requesting the physical address from the logical address transmitted to each of the duplicated routers, and uses them as means for holding the logical address and the physical address of the duplicated router. This is set.

More specifically, in a failure countermeasure method for a network system having dual routers, when a failure occurs in the working router, the spare router is set as a new working router, and the working router is replaced. As a new spare router, the communication control device is a means for holding a logical address and a physical address of a duplicated router, and is a logic assigned to the new working router and the new spare router. The address and the physical address are set respectively.

More specifically, in a failure countermeasure method for a network system having dual routers, the physical address is a MAC (Media Access Control) address and the logical address is an IP (Internet P
request message for requesting a physical address from the logical address is ARP
(Address Resolution Protocol) Request message, the response message that is the response is ARP
It is made to be a response message.

[0027]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments according to the present invention will be described below with reference to FIGS.

[General Matters Regarding TCP / IP] First, in order to facilitate understanding of the embodiment of the present invention, FIG.
The general items of TCP / IP will be briefly described with reference to FIGS. First, the concept of the IP address will be described with reference to FIG. FIG. 2 is a diagram schematically showing the form of each class of IP address.

In the TCP / IP communication protocol, an IP address is a network assigned to an entity (also called a "node") such as a computer (in the TCP / IP terminology, called "host") participating in communication. It is a unique address.

The IP address is represented by a 32-bit integer and is composed of a network ID and a host ID, as shown in FIG. Network ID
Is the ID that identifies the network, and the host ID
Is an ID for identifying a host on the network.

Classes are provided in the IP address depending on the form of the network to be used. In FIG. 1, classes A to C are shown in FIG.
The class to which the IP address belongs is identified by the pattern of the first bit of the network ID.

As can be seen from the figure, the class A has a larger number of bits of the host ID, and therefore is used in a large-scale network. On the other hand, the C class has a host ID of only 8 bits, so it can be said that it is suitable for a small-scale network.

As described above, the IP address is a 32-bit bit string, but in order to express this, a so-called octet display is used in which it is divided into 8-bit units and expressed in decimal. It is becoming popular.
For example, in a binary bit string, (10000010 0
0000001 00000100 0000100
The IP address represented by 0) is "130.1.4.
It is written as "8". In the present specification, this octet display will be used hereinafter.

When it is desired to identify one network with an IP address, this is called a network address, and the host ID in the IP address is represented as 0. For example, “130.1.4.
8 ”belongs to class B, the lower 16 bits are the host ID, and the network address is“ 130.
It becomes 1.0.0 ".

Next, the MAC address will be described.

The above-mentioned IP address was an address assigned to each node, but rather than this IP address,
The 48-bit address allocated from a lower position is a MAC (Media Access Control) address. Therefore, by comparison, it can be said that the IP address is a logical address and the MAC address is a physical address.

As a well-known model for explaining the communication protocol, OSI (Open Systems Interconnec
There is a reference model. Here, with reference to FIG. 3, the communication protocol in TCP / IP according to the OSI reference model and the relationship between the above IP address and MAC address will be outlined. Figure 3 shows the OSI reference model and TC
It is a schematic diagram which shows the relationship of the communication protocol of P / IP.

The OSI reference model is a model in which communication functions are hierarchically represented as shown in FIG. 3A. The lower the layer, the more physical layer the physical function is. This layer is responsible for logical functions. Each layer of this model provides a specific service to an upper layer and receives a specific service from a lower layer.

When this OSI reference model is applied to the communication protocol of TCP / IP, it becomes as shown in FIG. 3 (b). In the OSI reference model, communication services such as Telnet (virtual terminal function) and FTP (file transfer) belonging to the upper layers correspond to the fifth to seventh layers.

Below that is the transport layer protocol of the fourth layer, which is TCP (Transmition Control Protocol).
And UDP (User Datagram Protocol). Below that, the protocol of the third network layer is the IP protocol. (This layer is T
In CP / IP terms, it is sometimes called the Internet layer. ) The lowest layer is related to establishment of electrical standards and physical communication paths, and a typical standard is Ethernet (Erthenet). (This layer is sometimes referred to as a network interface layer in the term of TCP / IP.) Also in the description of the present specification, it is assumed that communication is performed by Ethernet.

The mechanism in the case of actual communication will now be described. As shown in FIG. 3 (c), data is passed to a lower layer at the time of transmission, and the header of that layer is used in the lower layer. Information is added, and conversely, at the time of reception, the header of that layer is analyzed and passed to the upper layer.

Specifically, in the lowest layer, the communication target receives an Ethernet frame. At the beginning, there is an Ethernet header, which includes a destination MAC address, a source MAC address, and a frame type indicating the type of the frame. After that, it becomes Ethernet data as seen from this layer. In the upper layer IP protocol, the lower layer Ethernet data is referred to as an IP datagram in this layer, and is composed of an IP header and IP data. Then, in the IP header, the destination IP
The address, the IP address of the transmission source, and the protocol type indicating the type of protocol are included.

Similarly, in the upper layer TCP protocol, the IP data consists of a TCP header and TCP data.

As described above, the specifications are set so that the header of each layer is processed, and the consistency of communication and the modularization of the processing in each layer are achieved.

Next, referring to FIG. 4, ARP (Address Reso
Solution Protocol) Let's talk about the protocol. FIG. 4 is a schematic diagram illustrating the mechanism of ARP.

As described above, in the lower network interface layer of TCP / IP, the address of the communication node is recognized as the MAC address, and in the upper Internet layer, it is recognized as the IP address.
Both the MAC address and the IP address are necessary because the Ethernet frame is assembled when the communication is performed by designating the other party by the IP address. However, normally, a host participating in communication recognizes the IP address of the other party, but does not recognize the MAC address. Therefore, the protocol for knowing the MAC address from the IP address is ARP. This A
There are two types of messages used for RP, an ARP request message and an ARP response message.

Here, it is assumed that there is a system having the network configuration shown in FIG.

(1) Suppose node A wants to communicate with node C. The IP address of C is 160.160.
Since it is 0.3, the IP address of the destination is 160.16
It is 0.0.3.

(2) The node A first receives the MA of the node B.
Broadcast an ARP request message over the network to obtain the C address. In this message, the target IP address of 160.160.0.3 is included.

(3) Nodes B, C, and routers, which are nodes on the same segment (here, the upper part that does not cross the router), receive the broadcast message and analyze the contents.

(4) Node B is the target IP
Since the address 160.160.0.3 and my address do not match, I do nothing. The same applies to routers.

The node C returns the ARP response message to the node A because its own address matches the target IP address of 160.160.0.3. This ARP response message contains the MAC address 0900.0420.1111 of node B.

(5) The node A receives the MAC response of the destination node B according to the returned ARP response message.
You can know the address.

In the above description, the router is also a target for receiving the ARP request message. Router too
Have a pair of IP address and MAC address at both ends,
Note that it may return an ARP response message when applicable. This is because it is necessary to know the MAC address of the router when performing communication beyond the router.

[First Embodiment] A first embodiment of the present invention will be described below with reference to FIGS. 1 and 5 to 11.

(I) Network Configuration and IP Address / MAC Address Allocation First, the network configuration and IP address / MAC address allocation according to the first embodiment of the present invention will be described with reference to FIG. FIG. 5 is a schematic diagram showing the configuration of the network system according to the embodiment of the present invention.

In this embodiment, it is assumed that Ethernet is used as the communication standard of the network interface layer, and each host performs message communication by TCP / IP as a protocol of the upper layer.

In this network system, as shown in FIG. 5, the transmission line A and the transmission line B are the router α and the router β.
Connected through. The computer a1, the computer a2, ..., The computer am are connected to the transmission path A. On one transmission path B, a computer b1, a computer b2,
..., the computer bn is connected. The network on the transmission path A side is also referred to as segment A, and the network on the transmission path B side is also referred to as segment B.

Each computer is composed of a CPU 12, a memory 13, and a communication control device 11, and these respective elements are connected by a system bus 14.

Now, let us say that the network address on the segment A side is 160.160.0.0 and the network address on the segment B side is 160.161.
Let's set it to 0.0.

Then, an IP address and a MAC address are assigned to each computer as shown in Table 1 below.

[0062]

[Table 1]

Further, the router has an I for each terminal (which is called an "interface") of the connected network.
It has a P address and a MAC address. When assigning an IP address, an IP address having the network address of the segment A is assigned to the interface on the segment A side, and an IP address having the network address of the segment B is assigned to the interface on the segment B side.

The IP addresses and MAC addresses of the routers α and β are as shown in Table 2 below.

[0065]

[Table 2]

It should be noted here that, as shown in Table 2 above, the IP addresses assigned to the interfaces of the router α and the router β are the same.

(II) Operation of ARP in this Embodiment Next, the AR in this embodiment will be described with reference to FIGS.
The operation of P will be described. FIG. 6 is a network system according to an embodiment of the present invention, in which the computer a1 to the AR
It is a schematic diagram when a P request message is issued. FIG. 7 is a schematic diagram when the ARP response message is issued from the computer a2 in the network system according to the embodiment of the present invention. FIG. 8 is a schematic diagram showing an example of the routing table. Figure 9 shows ARP
It is a schematic diagram which shows an example of a table.

The general mechanism of the ARP protocol has already been described, but will be described in more detail in order to understand the embodiments of the present invention.

Each of the computers a1 to am and b1 to bn has a routing table RTT and an ARP table ARPT on the memory 13.

The network according to this embodiment has the network configuration shown in FIG. 5 and is assigned an IP address and a MAC address as shown in Tables 1 and 2.

As shown in FIG. 8, the routing table RTT is composed of a destination network address RTT1 and an IP address RTT2 of the relay router.

The meaning of this routing table is that a message to the IP address having the destination network address RTT1 is transmitted to the router having the IP address indicated by the IP address RTT2 of the relay router.

For example, if the computer a1 has the values shown in FIG. 8 in the routing table RTT, the computer a1 has a network address of 16
A message destined for an IP address having 0.161.0.0 has an IP address of 160. In addition, as shown in FIG. 9, the ARP table ARPT contains the IP address and the MA.
It is composed of a pair of C addresses and can be said to be a conversion table of IP addresses and MAC addresses.

For example, in terms of the values shown in FIG. 9, IP
The address ARPT11 is 160.160.0.3,
The MAC address ARPT12 is 3, so the host with the IP address 160.160.0.3 (actually,
It can be seen that the MAC address of the communication control device 11 of the computer holds the MAC address) is 3.

Using this ARPT table, computer a
Suppose one sends a message to the host with IP address 160.160.0.3.

At this time, first, the computer a1 determines that the transmission source I
Create an IP datagram with a P address of 160.160.0.3. Then, referring to the ARP table, I
The MAC address 3 corresponding to the P address 160.160.0.3 is taken out. Next, this address MAC address 3
The IP datagram to be transmitted is stored in the memory 13, and the communication control device 11 is activated for transmission. The communication control device 11, which has received the transmission activation, sends the destination M from the memory 13.
The AC address and the transmission IP datagram are taken in to create an Ethernet frame, and the Ethernet frame created via the transmission path A is transmitted to the computer a3.

Thus, for example, when transmitting from the computer a1 to another host by IP address, it is good if there is a corresponding MAC address in the ARP table, but if there is no corresponding MAC address, as already described. Then, the MAC address is obtained by the ARP protocol.

For example, suppose computer a1 wants to communicate with computer a2. The IP address of the computer a2 is 160.
160.0.2, and this IP address is computer a1
It is assumed that the ARP table could not be searched. At this time, as shown in FIG. 6, the communication control device 11 of the computer a1 broadcasts an ARP request message on the network (however, only segment A),
MAC corresponding to IP address 160.160.0.2
Ask each host for the address.

Each host receives this ARP request message, analyzes it, and does nothing if it is different from its own IP address. Computer a2 is ARP
Since it has the IP address 160.160.0.2 that is the target of the request message, the AR that has written its own MAC address of 2, as shown in FIG.
P Response message is returned. Since the destination IP address of the ARP response message becomes the source IP address of the ARP request message, of course,
The destination of the ARP response message is the computer a1.

Upon receiving the ARP response message, the computer a1 searches for the MAC address (computer a that was written in the received ARP response message).
2) which is the MAC address of the second communication control device is read, and the IP address 160.1 is set in its own ARP table.
It is registered in association with 60.0.2, and thereafter, if necessary, the MAC address is used to assemble an Ethernet frame for communication.

(III) Operation of Network System in this Embodiment Next, the operation of the network system according to this embodiment will be described with reference to FIGS. FIG. 1 is an enlarged view of a main part of a network system according to an embodiment of the present invention.

The feature of the present invention is that the router is duplicated. The router is duplicated and one of them is used as a working one and the other is used as a backup, and when a failure occurs, switching is performed.

The configuration of the network system according to the present invention has already been shown in FIG. 5, of which the router α is the active router and the router β is the standby router. Further, when there is no fear of misunderstanding, the working router α and the standby router β are also described.

Therefore, normally, the working router α is used as the relay router of the segment A and the segment B,
If any failure occurs in this router, the backup router β will be used.

(III-1) Table Held in Communication Control Unit As shown in FIG. 1, the communication control unit 11 has an internal memory 118 in which a duplicate router management table is stored. DRTT, IP address storage table IPAT,
It holds a MAC address storage table MACAT. In addition, normally, the MAC address storage table MACA
T is stored in a ROM (Read Only Memory) 114.

These IP address storage table IPA
T is the own IP address of the computer a1, 160.16
It stores 0.0.1. On the other hand, the MAC address storage table MACAT is the M of the communication control device of the computer a1.
It stores 1 which is the AC address.

The structure of the duplicated router management table DRTT will be described with reference to FIG. FIG.
It is a schematic diagram which shows the structure of the duplication router management table DRTT.

The duplicated router management table DRTT is 2
The IP address storage area DRTT1 is used for handling duplicated routers, and the main components of the configuration are the IP address storage area DRTT1.
And a working router management area DRTT2 for storing information about the corresponding working router and a spare router management area DRTT3 for storing information about the spare router.

Here, the active router management area DRTT2
Let's take as an example. Spare router management area DRT
In T3, the term "working router" may be read as "spare router" in the following description.

Working router MAC address DRTT21
Stores the MAC address assigned to the active router that is the destination from this host.

Working router address registration bit DRTT2
3 is a bit indicating whether or not the MAC address of the active router is registered in the active router MAC address DRTT21, which is 1 when the active router MAC address is already registered in this area and 0 when it is not registered. Is.

The active router ARP response reception management bit DRTT24 is a bit for managing the reception of the ARP response message from the active router, and is 1 in the waiting state for reception from the active router, and the ARP response from the active router. 0 if the message has been received
It is. The working router failure indication bit DRTT25 is a bit for indicating whether or not a failure is detected in the working router, and is set to 1 when a failure in the working router is detected, and is set to 0 when it is not detected.

Since the computer a1 belongs to the segment A, it is routed through the interfaces on the segment A side of the router α and the router β. Therefore, 160.160.1.1 is stored in the IP address DRTT1 which is the area for registering the IP address of the dual router management table DRTT, and the active MAC address DRTT21 which is the area for registering the MAC address of the active router. , M + n + 1, and a spare MA that is an area for registering the MAC address of the spare router.
The C address DRTT31 stores m + n + 3, respectively.

Generally, the IP addresses and MAC addresses of the active router α and the standby router β and the computer a1
The IP address is set by the system administrator in advance as system information before the system starts operating.

Then, the computer a1 fetches them at the time of the start-up process and stores them in the dual router management table DRTT in the internal memory 118 of the communication control unit 11.
And the IP address storage table IPAT.

(III-2) Routing for normal data transmission When the communication control device 11 receives a communication activation from the CPU 12, it uses the active router α for routing. For example, suppose that the user wants to communicate with the computer b1 belonging to the network of the segment B over the router. The IP address of the computer b1 was 160.161.0.1. At this time, as shown in FIG. 4, the MAC address m + n + 1 of the working router α stored in the dual router management table DRTT is used as the destination MAC address of the head of the Ethernet frame. Further, the IP address 160.161.0.1 of the computer b1 is set as the destination IP address of the IP datagram that becomes the data.

The active router α looks at the destination of the transmitted Ethernet frame, finds that it is the Ethernet frame for itself, and analyzes the IP datagram. By looking at the destination IP address belonging to the IP datagram and knowing that the data should be sent to the segment B, this data will be sent to the segment B.

(III-3) Transmission of ARP request message The communication control device 11 has a timer inside and a fixed period T
An ARP request message is transmitted to the active router α and the standby router β for each q.

The main area of the format of the ARP message is a source MAC address, a source IP address, a destination MAC address, and an area for storing a destination IP address.

The ARP request message is sent to the destination M.
The MAC address of the node having the desired IP address is stored in the AC address area. Here, for the active router α, the DRTT MAC address m + n + stored in the internal memory 118 of the communication control device 11 is used.
Therefore, the ARP request message is issued to each of the spare routers β by using the MAC address m + n + 3.

(III-4) Reception of ARP Response Message and Countermeasures for Failure The communication control device 11 transmits the ARP request message to the working router α and the protection router β, and then transmits the transmitted AR.
The timer monitoring is performed for the reception of the ARP response message from the working router α and the protection router β, which are replies to the P request message.

Before the monitoring timer times out, data is transmitted to the computers b1 to bn while continuing to receive the ARP response message from the active router α for the ARP request message transmitted every period Tq. In some cases, the MAC address of the active router α is used to transmit data to the active router α as shown in (III-2).

It is assumed that, in response to the ARP request message transmitted every cycle Tq, there is no reply of the ARP response message from the active router α, and the timer for monitoring the reception of the ARP response message times out.

At that time, it is judged that some failure has occurred in the active router α, and the computer b1 of the segment B is determined.
When transmitting data to the computer bn2, the MAC address m + n + 3 of the backup router β is used, an Ethernet frame is created using this as the destination MAC address, and the Ethernet frame is sent to the transmission path A. That is, this operation corresponds to switching the communication path from the active router α to the standby router β.

In this way, even after the communication path is switched to the backup router β, the communication control device 11 keeps the working router α.
And the periodical transmission of the ARP request message to the backup router β continues.

In response to the ARP request message issued thereafter, when the reply of the ARP response message is received from the working router α before the monitoring timer times out, the failure of the working router α was temporary or When it is determined that the message has been recovered, the relay router for the message is switched from the backup router 54 to the active router 53 again. That is, the MAC address addressed to the active router α is used as the destination address to create an Ethernet frame and communicate with the computers b1 to bn of the segment B. After that, unless a failure occurs again, the communication control device 11
Will transmit data again using the active router α as a relay router.

Even if the ARP request message is sent a fixed number of times, if the ARP response message is not returned, the router judges that it is not recoverable and displays it on the console of the management server of the network to display the system. Contact the administrator and take measures such as replacing the faulty router.

(IV) Configuration and Operation of Communication Control Device According to this Embodiment The operation as the network system of the present invention is as described above, but below, the above operation is realized using FIG. The configuration and operation of the communication control device will be described in more detail. FIG. 11 is a block diagram showing the configuration of the communication control device according to the embodiment of the present invention.

(IV-1) Configuration and General Operation of Communication Control Unit 11 Computers connected to a network are generally connected using a communication control unit. The type that is often used is a type that is added to the expansion slot of a computer, LAN
It is called an adapter, an Ethernet adapter, a LAN board, a network interface card (NIC), or the like. It may also be mounted on the same board as the CPU.

The communication control device 11 according to this embodiment is assumed to be an intelligent type, and the communication control device 11 itself is an MPU.
It has (Micro processing Unit) 111. This MPU operates according to the program stored in the ROM 114. In addition, the communication control device 1 is stored in the ROM 114.
In addition to the operation program No. 1, a MAC address storage table MAT that stores the MAC address assigned to this communication control device is held.

The transmission circuit 115 uses the CSMA / CD (Carr
ier Sense Multiple Access with Collision Detectio
n) method, the common transmission line A via the transceiver 15
To transmit an Ethernet frame.

The internal memory 118 has a dual router management table DRTT and an IP address storage table IPAT for storing the IP address assigned to the own computer. When the communication control unit 11 is activated, the CPU
12 requests the bus scheduler 116 to use the internal bus 119 from the signal line S1, and the bus scheduler 116
Receives permission to use the internal bus 119 from the signal line S2. When the use permission of the internal bus is received, the CPU 12 sets the memory 118 in the write permission state by the signal line S11. Then, the system bus 14 and the interface 11
7 and the internal bus 119, the IP address of the duplicated router and the MAC addresses of the working router and the spare router are written in the IP address registration area and the MAC address registration area in the duplicated router management table DRTT, respectively. Also, the IP address assigned to the computer is written in the IP address storage table IPAT.

(IV-2) Data Transmission Operation of Communication Control Device 11 Next, the data transmission operation of the communication control device 11 of this embodiment will be described.

First, the computer b1 belonging to the segment B
The case of sending a message to the computer bn will be described.
The CPU 12 is a computer b2 connected to the common transmission path B2.
When there is a message to be sent to the computer bn, an IP datagram is created. The IP datagram is composed of an IP header and a data part other than the IP header. The IP header contains a destination IP address, a source IP address, and other management information.

That is, the CPU 12 writes the message to be transmitted in the data section 713, and the IP header contains
The destination IP address, the source IP address that is its own IP address, and other necessary management information are set.

When transmitting an IP datagram, the CPU 12 sends the IP datagram to be transmitted and the destination MA.
The C address is written in the memory 13, and a transmission request is sent to the MPU 111 in the communication control device by the signal line S9.
Here, the source MAC address is the IP shown in R8.
It can be obtained by searching the ARP table ARPT from the address.

MP which received the transmission request by the signal line S9
U111 uses the signal line S4 to generate the bus scheduler 116.
Request the right to use the system bus 114. MPU11
1 receives permission to use the system bus 14 from the bus scheduler 116 via the signal line S3. Then, via the internal bus 119, the interface 117, and the system bus 14, the IP datagram to be transmitted from the memory 13 and the destination MAC address are taken into the internal memory 118, and an Ethernet frame is created. The relationship between the IP datagram and the Ethernet frame is as already shown in FIG.

When the MPU 111 creates an Ethernet frame, the MPU 111 stores the created Ethernet frame in the internal memory 118 and requests the transmission circuit 115 to send it through the signal line S5. Upon receiving the transmission request through the signal line S5, the transmission circuit 115 takes out the Ethernet frame to be transmitted from the internal memory 118 and transmits the Ethernet frame to the common transmission line B via the transceiver 15.

By the way, the above-mentioned processing is the general processing when the Ethernet frame is transmitted to the transmission line.
Next, a description will be given of a peculiar process in the case of transmitting data over a router such as the case of transmitting data from the computer a1 of the segment A to the computer b1 of the segment B.

In such a case, the CPU 12 writes the message to be sent in the data part of the IP datagram, creates the IP datagram in which the IP address of the destination computer for sending the message to the destination IP address is written, and outputs the signal line. A transmission request is sent to the MPU 111 in S10. At the same time, MPU1 indicates that the destination of the IP datagram is a host belonging to segment B.
Notify 11 Upon receiving the transmission request to the host belonging to the segment B, the MPU 111 takes out the IP datagram transmitted from the memory 13 from the memory 13 via the interface 117. Then, referring to the duplicated router management table DRTT shown in FIG. 1, when the communication failure of the working router is not detected, the MAC address of the working router is taken out and written in the destination address of the Ethernet frame.

On the contrary, when the failure of the working router is detected and the communication failure of the protection router is not detected, the MAC address of the protection router is taken out,
Write to the destination address of the Ethernet frame.

Then, the IP datagram fetched from the memory 13 is written in the data section 25 of the Ethernet frame. After creating the Ethernet frame in this way, the MPU 111 starts transmission to the transmission circuit 115 via the signal line S5. Transmission circuit 11 that has received transmission activation
5 transmits the Ethernet frame to the common transmission line A via the transceiver 15. Since the MAC address of the relay router is written in the destination of the transmitted Ethernet frame, the designated relay router receives it and transmits it via the common transmission path B to the destination computer belonging to the segment B. It

(IV-3) Fault Detection Operation of Duplicated Router of Communication Control Device 11 Next, the communication control device 11 according to the present embodiment sets the cycle Tq.
Every time, the operation of transmitting the ARP request frame on the transmission line and detecting the failure of the duplicated router will be described.

(IV-3-1) Transmission of ARP Request Message The ARP request transmission timer 1121 interrupts the MPU 111 from the signal line S7-1 every cycle Tq.
The MPU 111 that received the interrupt from S7-1 sends an ARP
Create request message.

At this time, the source MAC address is
MAC address storage table MACT in ROM 114
MA assigned to its own communication control device
The C address 1 is stored. The source IP address is the IP address storage area IP in the internal memory 118.
IP address 16 assigned to itself by referring to A
Store 0.160.0.1.

On the other hand, even in the case of the ARP request message for the working router α, the ARP
Even in the case of a request message, the duplicated router management table D in the internal memory 118 is stored in the destination IP address.
Referring to the IP address registration area of RTT, the IP address 160.160.
The ARP request message in which 1.1 is written is created.

Then, in order to create an Ethernet frame, the ARP request message is stored in the data section of the Ethernet frame. The header part of the Ethernet frame stores a source MAC address, a destination MAC address and other management information. The source MAC address is the MAC address 1 stored in the MAC address storage table MAT held in the ROM 114. The destination MAC address is the duplicate router management table DRT.
The value stored in T is used. In the present embodiment, the destination MAC of the Ethernet frame for the active router α
As the address, m + n + 1, as the destination MAC address of the Ethernet frame for the backup router β,
m + n + 3 is used.

Then, the assembled Ethernet frame is sent to the transmission circuit 115, the transceiver 15 and the common transmission line A.
To the working router α and the standby router β via the. At the same time, the ARP response reception monitoring timer 1122 is started by the signal line S7-4. As a result, the ARP response reception monitoring timer 1122
It waits for the reception of the ARP response message transmitted from the working router α and the protection router β.

(IV-3-2) Reception of ARP Response Message Next, the operation when the communication control device 11 receives an Ethernet frame will be described.

When the transmission circuit 115 receives the Ethernet frame via the common transmission line 51 and the transceiver 15, it inputs a reception interrupt to the MPU 111 through the signal line S6 and stores the received Ethernet frame in the internal memory 118.
To be stored.

Now, in the management information of the header part of the Ethernet frame, there is an area for storing the frame type, and the type of the frame can be discriminated.
Therefore, by reading this frame type, the data part of the received Ethernet frame is ARPed.
It is possible to judge whether or not it is a message.

When the MPU 111 receiving the reception interrupt reads the frame type of the Ethernet frame stored in the internal memory 118 and determines that this is not the ARP message, the data part of the received Ethernet frame is stored in the memory via the interface 117. 13 is written to notify the CPU 12 that the message is received by the signal line S10.

When the MPU 111 that received the reception interrupt reads the frame type of the Ethernet frame stored in the internal memory 118 and determines that this is an ARP message, first, the MPU 111 sends the ARP message written in the data section. The source IP address is referenced to determine where the ARP message came from.

The source IP address is the IP address 160.160.1.1 assigned to the duplicated router.
, It is either from the working router α or the backup router β.

At this time, it is possible to determine whether the ARP response message is from the working router α or the backup router β, based on the source MAC address of the header part of the Ethernet frame.

When it is the ARP response message from the active router α, the active router ARP response reception management bit DRT24 of the duplicated router management table DRTT described in FIG. 10 is set to 1.

Thereafter, referring to the spare router ARP response reception management bit DRT34, when it is determined that the ARP response message from the spare router has been received, the ARP response reception monitoring timer 1122 is stopped by the signal line S7-3. Let When it is determined that the ARP response from the backup router has not been received, the reception process is terminated without stopping the ARP response reception monitoring and monitoring timer 1122.

On the other hand, it is the ARP from the backup router β.
If it is a response message, the backup router A of the duplicated router management table DRTT described in FIG.
The RP response reception management bit DRT34 is set to 1.

Thereafter, when it is determined that the ARP response message from the active router has been received by referring to the active router ARP response reception management bit DRT24, the ARP response reception monitoring timer 1122 is stopped by the signal line S7-3. AR from working router
When it is determined that the P response has not been received,
The reception process is terminated without stopping the ARP response reception monitoring timer 1122.

That is, the ARP response reception monitoring timer 1122 is stopped for the first time only when the ARP response message is received from both the working router and the protection router.

When the source IP address of the received ARP response message is other than the IP address assigned to the duplicate router, the process relating to the duplicate router is irrelevant. At this time, the received ARP message is sent to the memory 13 via the interface 117.
To the CPU 12 by the signal line S10.
A reception interrupt is put in to inform that the ARP response message has been received.

(IV-3-3) Processing at the time of timeout of ARP response reception monitoring timer 1122 As described above, the communication control device 11 transmits the ARP request message, and then the ARP response reception monitoring timer 1
Although the ARP response message is monitored by 122, when the ARP response message is not received by either the working router or the protection router within the determined time, the ARP response reception monitoring timer 1122 A timeout process will be performed. In other words, it is determined that some failure has occurred in the router that does not receive it.

ARP response reception monitoring timer 1122
When is timed out, the ARP response reception monitoring timer 1122 interrupts the MPU 111 via the signal line S7-2. The MPU 111 that received the interrupt
Duplicated router management table DRTT working router ARP
The values of the response reception management bit DRTT24 and the backup router ARP response reception management bit DRTT34 are checked.

If any of these bits is 0, which indicates that the router has not been received, that router has failed. Therefore, when the active router ARP response reception management bit DRTT24 is 0, the active router failure indicator is displayed. Bit DRTT25 is set to 1, and spare router AR
When the P response reception management bit DRTT34 is 0, the spare router failure indication bit DRTT35 is set to 1
And Then, the CPU 12 is notified that a failure has occurred in any of the routers.

Then, as already described in (III), C
When there is a failure in the working router and no failure in the backup router, the PU 12 performs router switching processing, and thereafter, the backup router performs communication.

The MPU 111 continues to send the ARP response message to the active router and the standby router each time an interrupt is generated from the timer 1122 even after detecting the failure in the router.

Before the ARP response reception monitor timer times out, the AR
When the P response message is received, it is recognized that the failure is recovered, and the failure indication bit of the router is cleared.

(V) Advantages of Using the Duplex Router of this Embodiment In a network system using the duplication router of this embodiment, for example, Tq is 10 seconds, and the timeout time of the ARP request reception timer is 5 seconds. If set, it will take about 15 seconds to switch to the backup router 54 after the failure of the working router 53.

This can greatly reduce the time during which communication cannot be performed on the network, as compared with a system using normal RIP routing.

Further, since the communication control device 11 performs the ARP response message transmission to the router at the cycle Tq and the reception monitoring of the ARP request message from the router, the switching control of the router is performed without imposing a processing load on the CPU 12. Can be done.

[Second Embodiment] A second embodiment of the present invention will be described below with reference to FIGS. 12 to 19.

In the first embodiment, the technique for detecting the fault of the router by using the ARP message and switching the router to secure the safety of the network system has been described. When the router has a fault and cannot be recovered, the faulty router is replaced as described in the first embodiment. At this time, since the MAC address of the router changes, the contents of the duplicate router management table DRTT on the host side must be rewritten. Therefore, normally, the system administrator resets the contents of the duplicated router management table DRTT of each host. This imposes an excessive burden on the system administrator when a large number of hosts are connected to the network.
In some cases, it may be necessary to suspend the work performed on the computer.

The present embodiment is based on the first embodiment, but is intended to automate such work as much as possible while keeping the system in the operating state. That is, in the configuration of the network system of the first embodiment, when the ARP request message is transmitted from the host side and the router is recognized by the ARP response message from the router side, its own duplicate router management table It is intended to register the returned MAC address of the newly exchanged router in the DRTT.

(I) Operation of Network System According to this Embodiment In the description of this embodiment, the operation of the network system will be followed by a flowchart.

First, the outline of the operation of the communication control device 11 according to the present embodiment will be described with reference to FIG. FIG. 12 is a schematic diagram showing an outline of the operation of the communication control device 11 according to the embodiment of the present invention.

The operation of the communication control apparatus shown in FIG. 11 is roughly classified into an Ethernet frame transmission process S0 and an Ethernet frame reception process S1.

In the Ethernet frame transmission process (S0), when the ARP request message is transmitted (S0
1) and when other Ethernet frames are transmitted (S02). The process of (S0) when transmitting this Ethernet frame will be described in detail in (I-1).

When the ARP request message is transmitted, the ARP response reception monitoring timer 1122 shown in FIG. 10 waits for the ARP response message. Then, when the communication control device does not transmit the ARP response reception timer within a fixed time, the timeout process is executed. The processing at the time-out will be described in detail in (I-3).

In the Ethernet frame receiving process (S1), when the ARP response message is received (S1
1) and when other Ethernet frames are received (S12). The process of receiving the Ethernet frame (S1) will be described in detail in (I-2).

(I-1) Ethernet Frame Transmission Process First, the operation of the communication control device 11 when transmitting an Ethernet frame will be described in the order of FIGS. 13 and 14. FIG. 13 is an MPU in the communication control device 11.
11 is a flowchart showing an operation when 111 transmits an Ethernet frame. FIG. 14 is a flowchart showing a process when an ARP request message is transmitted to the duplex router.

As described above, the head portion of the Ethernet frame has a frame type area, and by setting the value, it is possible to discriminate between the ARP message and other Ethernet frames. .

When there is a message to be transmitted, the CPU 12 prepares in the memory 13 the message to be transmitted, the destination MAC address, and the value according to the frame type of the Ethernet frame to be transmitted, and the signal line S11
Then, the MPU 111 starts transmission.

Here, the message prepared by the CPU 12 in the memory 13 is an ARP message or another IP message.
Either of the datagrams.

Upon receiving the transmission activation, the MPU 111 fetches the transmission message, the destination MAC address and the frame type value from the memory 13 and performs the Ethernet frame transmission process as follows.

First, the MPU 111 refers to the value of the frame type fetched from the memory 13 and determines whether the transmission data is an ARP message or not (S21-).
1).

Memory 1 except for ARP message
An Ethernet frame is created from the message captured from No. 3, the destination MAC address, and the frame type (S21-7), and a transmission request is sent to the transmission circuit 115 through the signal line S5 (S21-8).

Then, the transmission circuit which receives the transmission request transmits the Ethernet frame through the transceiver 15 and the common transmission path 51.

On the other hand, when it is determined from the frame type that the message is an ARP message (S21-2), the information in the ARP message is further checked to see if it is an ARP request message (S21-3). .

If the message is an ARP request message, then the destination IP address in the ARP message is looked at and compared with the IP address stored in the duplicate router management table DRTT,
When it is determined that the destination is a duplex router (either the working router or the standby router described in the first embodiment) (S
21-3), go to S21-4.

When the ARP message is not the ARP request or the destination is not the duplex router, the Ethernet frame is created as usual (S2
1-7), and transmits it (S21-8).

When it is determined that the destination is a dual router, the active address registration bit DRTT23 of the dual router management table DRTT is checked to see if the active router address is registered (S21-).
4).

When the working router address is not registered, ARP is used to know the MAC address of the working router.
Since the request message needs to be transmitted, the Ethernet frame is assembled (S21-7) and transmitted (S21-8).

When the active router address is registered, in order to know the MAC address of the active router, A
There is no need to issue an RP request message, and within this communication control device 11, an ARP response message is created using the value of the active router MAC address (S21-
5) The CPU 12 is notified of the created ARP message.

Similarly to the first embodiment, the communication control device 11 of the present embodiment also sets a fixed cycle Tq for the duplicated router.
Then, the ARP request message is transmitted. Next, this process will be described with reference to FIG.

The ARP request transmission timer 113 shown in FIG. 11 counts time, and interrupts the MPU 111 by the signal line S8-1 at regular intervals. This processing is performed in response to this interrupt. When this interrupt comes, the MPU 111 sets the value of the IP address DRTT1 which is the area storing the IP address of the duplicated router management table DRTT to the destination I.
An ARP request message with the P address is created (S18-1).

Then, the working router MAC address registration bit DRTT12 and the spare router MAC address registration bit of the duplicated router management table 13 are checked (S18).
-2, S18-3), if neither of them is registered, the ARP request message is broadcast (S18-9).

Since the IP address of the ARP request message is the IP address stored in the duplicated router management table DRTT, it is expected that a response will be returned with respect to this message. For both routers). Therefore, the working router ARP response reception management bit DRTT24 of the duplicated router management table DRTT is set to 1 (S18-10), which means waiting for reception, and similarly, the backup router AR.
The P response reception management bit DRTT34 is set to 1 which means waiting for reception (S18-10).

After transmitting the Ethernet frame to the transmission line, the dual router ARP response reception monitoring timer 1122 is activated (S18-8), and the ARP response message comes from both the working router and the standby router. To count the timer until.

Even when the MAC address of the working router and the MAC address of the protection router are both registered in the duplicate router management table DRTT, the ARP request message is transmitted to the protection router and the protection router. This is because the MAC address is known, so I
It does not mean to know the MAC address from the P address, but it is intended to judge whether or not there is a failure in the router depending on whether or not the ARP response message is returned.

Therefore, also at this time, an Ethernet frame in which the MAC address of the active router is set is created and transmitted to the active router (S18-4). The working router A of the duplicated router management table DRTT
The RP response reception management bit DRTT24 is set to 1 which means waiting for reception (S18-5).

Similarly, an Ethernet frame in which the spare router MAC address is set is created and transmitted to the spare router (S18-4). Then, the standby router ARP response reception management bit DRTT34 of the duplicated router management table DRTT means waiting for reception 1
(S18-5).

At this time also, the duplicate router ARP response reception monitoring timer 122 is started (S18-8).
Then, the time when the ARP response message is returned from the working router and the protection router is counted.

(I-2) Ethernet Frame Reception Processing Next, the operation of the communication control device 11 when receiving an Ethernet frame will be described in the order shown in FIG. FIG. 15 is a flowchart showing the operation when the MPU 111 in the communication control device 11 receives an Ethernet frame.

When the transmission circuit 115 receives the Ethernet frame from the transceiver 15, the transmission circuit 115 stores the received frame in the internal memory 118 and outputs the M frame through the signal line S6.
A receive interrupt is sent to PU111.

The MPU 111 which has received the reception interrupt starts the reception processing and performs the following processing.

First, the MPU 111 has the internal memory 118.
With reference to the frame type of the received frame stored in, it is determined whether or not it is the Ethernet frame of the ARP message (S19-1). If it is not the Ethernet frame of the ARP message, the data part 25 of the received Ethernet frame is written in the memory 13 via the interface 117, and the signal line S10
Then, a message reception interrupt is sent to the CPU 12 to end the processing (S19-16).

When the received Ethernet frame is an ARP message frame, the information in the ARP message is further checked to see if it is an ARP response message (S19-2).

If the message is an ARP response message, then the sender IP address in the ARP message is checked and compared with the IP address stored in the duplicated router management table DRTT.
When it is determined that the transmission source is the duplex router (either the working router or the protection router described in the first embodiment) (S
19-3), go to S19-4.

When the ARP message is not the ARP response or the transmission source is not the duplex router, the CPU 12 is notified as usual that the Ethernet frame is received (S19-16).

When it is determined that the transmission source is the duplex router, the working address registration bit DRTT23 of the duplex router management table DRTT is checked to see if the working router address is registered (S19). −
4).

When the active router address is not registered, it is necessary to register the MAC address of the active router. Therefore, the MAC address of the active router that is the transmission source of the ARP response message transmitted is set to the duplicate router management. Working router MAC address DR of table DRTT
It registers in TT21 (S19-5).

Then, the duplicate router management table DRT
The working address registration bit DRTT23 of T is set to 1 which means that it has already been registered (S19-6).

After that, the communication control device 11 is
2 is notified that the ARP response message has been received (S19-20).

Next, the MPU 1111 has the spare router reception management bit DR of the dual router management table DRTT.
The TT34 is checked to see if the standby router is waiting to receive the ARP response message (S).
19-7).

When the ARP response message from the spare router has already been received, this means that the ARP response messages from both the working router and the spare router have been received. Therefore, the duplex router ARP response reception monitoring timer 1122 Is stopped (S19-8). When the ARP response message from the spare router is not received, nothing is done, and the ARP response message from the spare router is waited next.

Now, returning slightly, in S19-4, when it is determined that the transmission source is the dual router, the working address registration bit D of the dual router management table DRTT is set.
It is assumed that the RTT 23 is checked and the MAC address of the working router is registered.

At this time, when the source MAC address of the received ARP response message belongs to the working router, it is not necessary to register the MAC address after the working router, and the ARP response message is returned from the working router. In order to indicate that, the active router reception management bit of the duplicated router management table DRTT is set to 1 indicating that it has been received (S19-10).

On the other hand, when the MAC address of the sender of the received ARP response message is not that of the working router, the sender is the spare router. At this time, first, the spare address registration bit DRTT33 of the dual router management table DRTT is checked to see if the MAC address of the spare router is registered (S1).
9-11).

When the MAC address of the spare router is registered, the source MAC address of the received ARP response message is the duplicated router management table D.
It is checked whether or not it matches the spare router MAC address DRTT31 of the RTT (S19-12).

If they match, nothing else is done, so in order to indicate that the ARP response message has been returned from the spare router, the spare router reception management bit DR of the dual router management table DRTT is shown.
The TT34 is set to 1 indicating that it has been received (S1).
9-13).

At this time, the protection router A
Since the RP response message is returned, it indicates that the backup router has not failed, or has recovered even in the past. Therefore, the spare router failure indication bit D of the duplicated router management table DRTT
The RTT 35 is checked (S19-14), the spare router failure indication bit DRT in the duplicated router management table DRTT.
When T35 is 1, it is set to 0 (S19-19).

Next, the MPU 1111 uses the active router reception management bit DRT of the dual router management table DRTT.
By looking at T24, it is checked whether or not the active router is waiting to receive the ARP response (S19-15).

When the ARP response message of the working router has already been received, this means that the ARP response messages of the working router and the protection router have both been received. Therefore, the duplicate router ARP response reception monitoring timer 1122 is set. It stops (S19-8).

In S19-12, it is assumed that the MAC address of the ARP response message and the MAC address of the duplex router do not match. At this time, it indicates that the backup router has been updated and the MAC address has been updated. Therefore, the sender M included in the ARP response
The AC address is registered as a new spare router MAC address in the spare router MAC address DRTT31 of the dual router management table DRTT (S19-1).
8), in order to indicate that the ARP response message has been returned from the spare router, the spare router reception management bit DRTT34 of the duplex router management table DRTT
Is set to 1 indicating that it has been received (S19-1).
3).

Regarding the subsequent failure of the backup router 2
Processing for the spare router failure indication bit DRTT35 in the duplicated router management table DRTT (S19-14, S
19-19) and the processing related to the duplicate router ARP response reception monitoring timer 1122 (S19-15, S19).
-8) is the same as above.

(I-3) Processing when the ARP response reception monitoring timer times out Next, the processing when the ARP response reception monitoring timer times out will be described in the order of FIG.

FIG. 16 is a flowchart showing the processing when the ARP response reception monitoring timer 1122 in the communication control device 11 times out.

As described above, from within the communication control device 11,
An ARP request message is sent to each of the working router and the standby router at regular intervals. The ARP response message that is the reply is
Sometimes it does not come back from one of the routers. At this time, the ARP response reception monitoring timer times out, and the following processing is performed.

When the duplicate router ARP response reception monitoring timer 1122 times out, an interrupt is sent to the MPU 111 by the signal line S7-2. MPU111
When the interrupt is received by the signal line S7-2, the ARP response reception monitoring timer for the duplex router starts time-out processing.

First, the duplicated router management table DRTT
The active router address registration bit DRTT23 is checked to see if the active router MAC address is registered (S20-1).

If the MAC address of the active router is not registered, from the beginning, the router in the system is
It is considered that it does not exist. Therefore, at this time, the spare router ARP response reception management bit DRTT34 of the dual router management table DRTT is set to 0 to cancel the standby router ARP response reception wait (S20-5). Working router ARP response reception management bit DRTT2
4 is set to 0, the waiting for ARP response reception is canceled (S20-6), and this processing is ended.

On the other hand, if the MAC address of the active router is registered, then the active router ARP response reception management bit DRTT24 of the duplicated router management table DRTT is referred to and the ARP response reception standby of the active router is waited for. It is checked whether it is in the state (S20-
2).

When the ARP response of the working router is waiting for reception, the procedure goes to S20-9.

When the ARP response of the working router is not waiting for reception, next, the duplicated router management table DR
The spare router address registration bit DRTT33 of the TT is referred to check whether the MAC address of the spare router is registered (S20-3).

When the MAC address of the spare router is not registered, it is considered that the spare router does not exist in the system. At this time, the spare router A
Cancel the status of waiting for RP response (S20-
5) Then, the state of waiting for reception of the active router ARP response is also canceled (S20-6), and the process is terminated.

When the MAC address of the spare router is registered, the spare router ARP response reception management bit DRTT34 is next checked to see if the standby router ARP response is waiting to be received (( S20-4).

When the ARP response message is received from the spare router, the process is terminated because it is not in the ARP response reception waiting state.

[0218] When the ARP response reception wait state is set, it means that the ARP response message has not arrived, so that the backup router has a failure. At that time, the spare router failure indication bit DRTT
35, it is checked whether or not the value is currently 1 indicating the failure of the spare router (S20-7), 1
If not, it is rewritten to 1 representing the failure (S20-8).

After that, both the standby router and the active router wait for reception, and the processing is terminated (S20-5, S).
20-6).

Although not shown in the flow chart, at this time, the communication control unit informs the CPU 12 that a failure has occurred in the backup router, and when the control returns to the CPU 12, the backup program is set by the network management program. It is also possible to display on the display device that a failure has occurred in the router.

Returning a little, it is assumed that the active router is waiting to receive the ARP response in S20-2.

At that time, next, the spare router address registration bit DR of the duplicated router management table DRTT.
With reference to TT33, it is checked whether or not the spare router MAC address is registered (S20-9).

When the spare router registration address is not registered, it means that the working router has a failure and the spare router as a substitute cannot be used.
Therefore, at this time, the active router failure indication bit DRTT25 of the dual router management table DRTT is set to 1 indicating that a failure has occurred, and the CPU 12 is notified that the active router has failed (S20-1).
7). Upon receiving the notification, the CPU 12 takes measures such as displaying that a failure has occurred on the display device or writing this out in the history of the network.

After that, the standby router and the active router wait for the cancellation of the ARP responses (S20-
5, S20-6) is the same as above.

On the other hand, when the spare router registration address is registered, next, the duplicated router management table D
With reference to the spare router ARP response reception management bit DRTT34 of the RTT, it is checked whether or not the spare router is in the reception waiting state (S20-11).

When the standby router is waiting for reception, it means that both the active router and the standby router have failed. Therefore, the active router failure display bit DRTT25 and the standby router failure display in the duplicated router management table DRTT. The bit DRTT35 is set to 1 indicating that a failure has occurred together (S20-17).

Then, the CPU 12 is notified that a failure has occurred in both duplex routers (S20-17). CP
As described above, the U12 must take appropriate trouble countermeasures such as displaying on the display device and collecting communication history.

After this, the standby router and the active router wait for cancellation of ARP response reception (S20).
-5 and S20-6) are the same as above.

When the standby router ARP response is not waiting for reception, the ARP response is returned from the standby router, so the current situation is that the active router has failed and the standby router is available. Means

At this time, the values of the management area of the working router and the management area of the protection router of the dual router management table DRTT are completely exchanged in order to make the protection router the working router (S20-11).

In this case, since the faulty router is the spare router, the spare router fault indication bit DRTT of the dual router management table DRTT.
35 is set to 1 which means that a failure has occurred (S20
-12).

Since the working router and the spare router have been replaced, the CPU 12 is notified of this fact and the CPU 1
It is necessary to rewrite the management information on the second side.

Therefore, an ARP response message is assembled from the IP address of the duplicated router and the MAC address of the router that has become the new working router (S20).
-13), and transmits an ARP response message to the CPU 12 by interruption (S20-14).

As a result, the CPU 12 can rewrite the management information on the duplicated router and, if necessary, take necessary measures according to the network management program.

Finally, after this, the standby state for waiting for the reception of the ARP response of the standby router and the active router (S20-5, S20-6) is the same as above.

(II) Evaluation of Operation of Network System of this Embodiment This embodiment is intended to enhance the security of the network system by duplicating the router. According to the present embodiment, even if the working router is down, switching to the backup router allows the network system as a whole to operate normally. Eventually, the network system as a whole fails to operate when both the active router and the standby router are down, and such a case would be extremely rare.

Also, a so-called hot swap is possible in which a faulty router is replaced while the system is operating, and the management information of the active router and the standby router is automatically switched. There is no need to stop and no new configuration from the console is required.

As described above, the number of routers according to this embodiment is two.
In the duplicated network system, both availability and reliability of the network system can be improved.

(III) Operation of Network System Based on Specific Configuration and Data of this Embodiment In (I), the operation of the network system according to this embodiment has been described based on the flowchart. Here, description will be made with reference to FIGS. 5 and 17 to 19 based on a specific network system configuration and data. As described above, FIG. 5 is a configuration diagram of the network system according to the embodiment of the present invention. However, also in the present embodiment, the same configuration as that of the first embodiment is used, and the same IP address and MAC address are used. Is used,
I will refer to this as appropriate. FIG. 17 is a schematic diagram (part 1) when specific data is set in the duplicated router management table DRTT. FIG. 18 is a schematic diagram when specific data is set in the duplicated router management table DRTT (No. 2), and particularly shows a case where the management information of the working router and the protection router is replaced and set. . FIG. 19 is a schematic diagram when specific data is set in the duplicated router management table DRTT (part 3).

In the following description, a model case will be taken when a failure occurs in the working router α shown in FIG. 5, the router is replaced with a new router, and the newly replaced router is used as a new backup router. Lets. The MAC address of this new router will of course have a different MAC address than the router that caused the failure.

As shown in Tables 1 and 2 of the first embodiment, IP addresses and MAC addresses are assigned to each computer and router connected to the network system.

Now, assuming that the computers a2 to am and b1 to bn and the routers α and β have already been started, the computer a
Suppose you try to launch 1.

First, when the computer a1 is started, FIG.
The CPU 12 shown in FIG. 11 is a start-up process and the routing table RTT in the memory 13 shown in FIG.
From the system information preset by the user of the computer, "the network address is 160.161.
The message addressed to 0.0 is the IP address 160.16
The information "send to 0.1.1" is read and set in the routing table RTT. CPU 12
In the startup process, the IP address and MAC address of the ARP table are left unset.

Next, the CPU 12 activates the communication control device 11. When the communication control device 11 is activated, the CPU 12
Is read from system information set in advance by the system administrator and stored in the IP address DRTT1 which is the IP address registration area of the duplicated router management table DRTT of the duplicated router management table in the internal memory 118 of the communication control device 11. Set. Specifically, this value is the IP address 1 assigned to the interface on the common transmission path A side of the router α and the router β as shown in FIG.
It is 60.160.1.1.

The working router MAC address DRTT21 and the spare router M in the duplicated router management table DRTT.
The AC address DRTT31 is unregistered, and the active router address registration bit DRTT23 and the spare MAC router address registration bit DRTT33 are set to the respective MAs.
It is 0 indicating that the C address is not registered.

The working router ARP response reception waiting bit DRTT24 and the protection router ARP response reception waiting bit DRTT34 are 0, which indicates the reception waiting release state, and the working router failure indication bit DRTT25 and the protection router failure indication bit DRTT35 are It is 0 indicating that no fault is detected.

When the communication control unit 11 is started up, FIG.
The crystal oscillator 113 shown in FIG.
-2 outputs a count-up signal. Upon receiving the count-up signal, the ARP request transmission timer 1121 continues counting up, and when it reaches Tq defined as the ARP request transmission period, it interrupts the MPU 111 from the signal line S7-1. After inputting the interrupt, the count value is reset to 0 and counting is started again up to Tq.

The ARP request transmission cycle Tq can be determined in consideration of factors such as reliability and performance required by the entire network system. In this embodiment, for example, Tq is set to 10 seconds.

Now, the MP which has received the interrupt from S8-1
U111 uses the IP address 160.160.1.
The ARP request message inquiring the MAC address corresponding to 1 is transmitted on the network.

FIG. 17 immediately after the communication control device 11 is activated.
In the duplicated router management table of the value shown in, the MAC address is not registered in both the working router and the spare router.
The initial ARP request message will be sent in an Ethernet broadcast.

The ARP request message transmitted by the broadcast is the computer a connected to the common transmission path A.
2-am, the router α, and the router β receive them. Then, according to the principle of the ARP protocol as described above, only the router α and the router β having the inquired IP address 160.160.1.1 have their respective MAC addresses m.
Create an Ethernet frame of ARP response message with + n + 1 and m + n + 3 as replies, and
Send to 1.

When the communication control device 11 receives the ARP response message, the MPU 111 follows the previously described flow chart of FIG.
With the source router of the RP response message as the active router, the MAC address is registered and received later
The transmission source router of the RP response message is a backup router (S19-4, S19-5, S1 in FIG. 15).
9-9, S19-11, S19-18 etc.).

In this embodiment, it is assumed that the ARP response message sent from the router α is received first, and the ARP response message of the router β is received later.

When the transmission circuit 115 receives the Ethernet frame of the ARP response message from the router α via the common transmission path A, the transmission circuit 115 writes the received frame in the internal memory 118 and then signals the MPU 111 of a frame reception interrupt. It will be entered by line S6.

MPU1 interrupted from the signal line S6
11 retrieves the received frame from the internal memory 118, and according to the flowchart of FIG. 15, as already described, sets the MAC address m + n + 1 of the router α written in the ARP response message to the working router MAC address of the duplicated router management table DRTT. Write to DRTT21, working router MAC address registration bit DRT
T23 is set to 1 indicating registered.

Similarly, when the ARP response message from the router β is received, the spare router MAC address registration area DRT is followed according to the flowchart of FIG.
At T31, the MAC address m + n + 3 of the router β written in the ARP response message is written, and the spare router MAC address registration bit DRTT33 is set to 1 indicating registered.

In this way, values are set in the duplicate router management table as shown in FIG. 18 (a).

Thereafter, the MPU 111 follows the flowchart of FIG. 14 every time the ARP request transmission timer times out, and the working MAC address m + n + 1 and the spare MAC registered in the duplicated router management table.
AR with address m + n + 3 as destination MAC address
The P request message is transmitted to the active router α and the standby router β, respectively. In this way, by checking whether the ARP response message is returned from each router, it is possible to confirm whether each router is operating normally.

Now, the communication control when the computer a1 sends a message to the computer connected to the common transmission line B via the active router α, for example, the computer b1 whose IP address is 160.161.0.1. Let us explain the operation of the device.

After the start-up process is completed, the CPU 12 creates an IP datagram addressed to the IP address of the destination computer when the user requests a message communication.

Then, the CPU 12 makes the network address 160.16 of the segment A to which the computer a1 belongs.
If the IP address having the network ID different from 0.0.0 is the destination IP address, it is necessary to refer to the routing table RTT to fetch the relay destination IP address.

In this case, the computer a1 creates an IP datagram whose destination IP address is 160.161.0.1. In addition, the destination IP address 160.16
Since 1.0.1 is different from the network address 160.160.0.0 of the segment of the computer a1, the IP address 160.160.1.1 of the relay destination router is referred to by referring to the routing table RTT. To earn.

Next, the CPU 512 sends the ARP table A
Referring to the RPT, the IP address 160.160.1.
Attempts to acquire the MAC address corresponding to 1. However, immediately after the start-up processing of the computer a1, the ARP
Since it is not registered in the table ARPT, the computer b1
No MAC address is registered, of course.

Therefore, the CPU 12 creates an ARP request message for inquiring the MAC address corresponding to the IP address 160.160.1.1, puts a transmission interrupt to the MPU 111 through the signal line S11, and transmits the MPU 111.

IP address 160.16 of the duplex router
The ARP message to 0.1.1 is transmitted to the dual router management table DRTT as shown in FIG.
If the address is already registered, the MPU 111 determines the area active MAC address DRTT of the duplicated router management table.
An ARP response message in which the MAC address stored in 21 is set as a reply is created, and is notified to the CPU 12 through the signal line S10 (S21-5, S21-).
6). Source IP of this ARP response message
Address duplication Router IP address 160.16
It is 0.0.

Upon receiving the ARP response message, the CPU 112 writes the IP address written in the message.
The address and the MAC address are registered in the ARP table ARPT.

Here, the MAC address of the IP datagram waiting to be transmitted can be obtained again by referring to the ARP table to obtain the MAC address corresponding to the relay destination IP address 160.160.1.1. .

Then, the CPU 12 prepares the IP datagram and the destination MAC address obtained as described above in the memory 14 and puts the transmission interrupt in the MPU 111.

The MPU 111 which has received the transmission interruption
The AC address is the working router α, and the Ethernet frame in which the IP datagram is written in the data part is transmitted to the working router α.

Next, assume that a failure occurs in the working router α.

When a failure occurs in the working router α, an ARP response message is sent from the working router α to the ARP request message sent from the communication control device 11 to the working router and the spare router every cycle Tq (10 seconds). It will not be sent.

As a result, the ARP response message reception monitoring timer 1122 times out, and the MPU1
11 receives the interrupt through the signal line S7-2 and starts the time-out process according to the flowchart of FIG.

As shown in FIG. 16, when a failure occurs in the working router α, as long as there is no failure in the protection router β, the protection router β is registered as a new working router in the dual router management table DRTT. This is what is going to be done (S20-11).

As shown in FIG. 18B, when the standby router is the working router, the router β having the MAC address m + n + 3 is registered as the new working router.

As shown in the flow chart of FIG. 16, the spare router MAC address registration area DRT
The MAC address m + n + 1 of the old working router α is written in T31, and the spare router failure indication bit DRT
1 indicating that a failure has occurred is registered in T33.

When the MAC addresses of the spare router and the working router are exchanged, the MPU 111 sets the MAC address of the new working router β to the source address m + n + 3.
An ARP response message in which the IP address 160.160.1.1 of the duplicated router is written in the transmission source address is created, and a reception interrupt is input to the CPU 12 from the signal line S10. CPU 12 receiving a reception interrupt
Extracts the source MAC address and the source IP address from the ARP response message received from the MPU 111, and rewrites the MAC address corresponding to the IP address 160.160.1.1 registered in the ARP table ARPT from m + n + 1 to m + n + 3. . In this way, the IP address 160.160.1.
The IP datagram sent to 1 will be sent to the router β having the MAC address m + n + 3.

Further, the spare router times failure indication bit is
It is assumed to be 1 indicating that a failure has occurred.

Here, the ARP request message transmission timer 1122 has timed out, and the MPU 111 sends an A to the backup router α according to the flowchart of FIG.
An RP request message shall be sent (S1
8-6 etc.).

In response to this ARP request message, when the ARP response message is received again from the router α which has become the protection router from the working router due to the failure, the MPU 111 receives the message shown in the flowchart of FIG. According to the processing, the failure occurrence state of the spare router failure indication bit is canceled (S19-19), and 0 is set.
To Thus, after that, the router α stands by on the network as a backup router.

By registering the router α as the active router and the router β as the standby router in the duplication management table DRTT as described above, the active router α fails and the standby router β becomes the new active router. Let's assume that it becomes a router and replaces the old working router α with a new router γ.

It is assumed that the MAC address of the interface on the common transmission path A side of the router γ is m + n + 5, and the MAC address of the interface on the common transmission path B side is m + n + 6.

Further, the IP address assigned to the router γ is set to be the same as the IP address assigned to the router α, and the IP of the interface on the common transmission path A side is set.
The address is 160.160.1.1, and the IP address of the interface on the common transmission line B side is 160.161.1.
Let it be 1.

The duplicated router management table DRTT of the communication control device 11 immediately after the failure of the router α is as shown in FIG. 18 (b).

When the router α is removed and the router γ is newly connected to the common transmission line A and the common transmission line B, in response to the ARP request message transmitted by the communication control device 11,
The router β and the router γ send the ARP response message to the transmission control device 11. When the communication control device 11 receives the ARP response message from the router γ, M
The PU 111 performs reception processing according to the flowchart of FIG. 15 and registers the MAC address of the router γ in the spare router MAC address registration area DRTT31 (S19-12, S19-18). As a result, the values of the duplicated router management table DRTT are as shown in FIG.

After that, the router γ can be used as a standby router and waits on the network until a failure occurs in the active router.

As described above, by using the communication control device of this embodiment, even if one of the routers has a failure, the network system can be operated without stopping and the system information is reset. There is no need.

[0287]

According to the present invention, in a network system with dual routers, when switching a route from an active router to a backup router, the switching time is shortened compared to when RIP is used, and the availability of the network is improved. It is possible to provide a network system in which a router is duplicated and a failure countermeasure method for a network system in which a router is duplicated without increasing the load on the CPU of the computer.

[Brief description of drawings]

FIG. 1 is an enlarged view of a main part of a network system according to an embodiment of the present invention.

FIG. 2 is a diagram schematically showing the form of each class of IP address.

FIG. 3 is a schematic diagram showing a relationship between an OSI reference model and a TCP / IP communication protocol.

FIG. 4 is a schematic diagram illustrating the mechanism of ARP.

FIG. 5 is a schematic diagram showing a configuration of a network system according to an embodiment of the present invention.

FIG. 6 is a schematic diagram when an ARP request message is issued from the computer a1 in the network system according to the embodiment of the present invention.

FIG. 7 is a schematic diagram when an ARP response message is issued from the computer a2 in the network system according to the embodiment of the present invention.

FIG. 8 is a schematic diagram showing an example of a routing table.

FIG. 9 is a schematic diagram showing an example of an ARP table.

FIG. 10 is a schematic diagram showing a configuration of a duplicated router management table DRTT.

FIG. 11 is a block diagram showing a configuration of a communication control device according to an embodiment of the present invention.

FIG. 12 is a communication control device 11 according to an embodiment of the present invention.
FIG. 6 is a schematic diagram showing an outline of the operation of FIG.

FIG. 13 is a flowchart showing an operation when the MPU 111 in the communication control device 11 transmits an Ethernet frame.

FIG. 14 is a flowchart showing a process when an ARP request message is sent to the duplex router.

FIG. 15 is a flowchart showing an operation when the MPU 111 in the communication control device 11 receives an Ethernet frame.

FIG. 16 is a flowchart showing processing when an ARP response reception monitoring timer 1122 in the communication control device 11 times out.

FIG. 17 is a schematic diagram when specific data is set in the duplicated router management table DRTT (No. 1).

FIG. 18 is a schematic diagram when specific data is set in the duplicated router management table DRTT (part 2), and particularly shows a case where the management information of the working router and the protection router is exchanged and set. .

FIG. 19 is a schematic diagram when specific data is set in the duplicated router management table DRTT (part 3).

[Explanation of symbols]

a1 to am, b1 to bn ... Computer, A, B ... Common transmission path, α, β, γ ... Router, 11 ... Communication control device, 12 ...
CPU, 13 ... Memory, 14 ... System bus, DRTT
... duplicate router management table, IPAT ... IP address storage table, 11MAT ... MAC address management table, 111 ... MPU, 1121 ... ARP request message transmission timer, 1122 ... ARP response message reception timer, 113 ... crystal oscillator, 114 ... RO
M, 115 ... Transmission circuit, 116 ... Bus scheduler, 1
17 ... Bus interface, 118 ... Internal memory, 15
… Transceiver, DRTT2… Current router management area,
DRTT3 ... Spare router management area, DRTT21 ...
Working router MAC address, DRTT23 ... Working router MAC address registration bit, DRTT24 ... Working router reception wait bit, DRTT25 ... Working router failure display bit, DRTT31 ... Spare router MAC address registration area, DRTT33 ... Spare router MAC address registration bit, DRTT34 ... Spare router reception wait bit, DRTT35 ... Spare router failure indication bit, RT
T ... Routing table, RT1 ... Destination network address, RT2 ... Relay destination IP address, ARPT ...
ARP table, ARPT1 to ARPTk ... Each entry of ARP table, ARPT11 to ARPTk1 ... IP
Address, ARPT12 to ARPTK2 ... MAC address.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Masayuki Tanji 5-2-1 Omika-cho, Hitachi City, Ibaraki Prefecture Hitachi Ltd. Omika Plant, Ltd. (72) Inventor Hisayuki Maruyama 5-chome, Omika-cho, Hitachi City, Ibaraki Prefecture No. 1 Incorporated company Hitachi Ltd. Omika factory (72) Inventor Hirofumi Inoue 5-2-1 Omika-cho, Hitachi city, Ibaraki Prefecture Incorporated company Hitachi Ltd. Omika factory

Claims (10)

[Claims] 1. A network system in which one or more networks are connected via a router and the routers are duplicated, wherein the computer connected to the network has a communication control device, Communicate with each other by a communication control device, and one of the duplicated routers is an active router,
The other one is used as a spare router, and the same logical address and different physical addresses are assigned to each other, and the communication control device has means for holding the logical address and physical address of the duplicated router. When this communication control device communicates with another computer via the router, it communicates using the physical address of the working router, and periodically communicates with the duplicated router from the logical address to the physical address. A request message for an address is sent, and whether or not a response message that is a response to the sent request message is sent from the router is monitored for each duplicated router, and the response from the router is sent. If no message is sent, it means that the router has failed, and if the working router fails. When it is determined that an error has occurred, the network system in which the router is duplicated is characterized in that when communicating with another computer via the router, the communication is switched to use the physical address of the backup router. . 2. The communication control device receives a logical address and a physical address assigned to the duplicated router from a computer, and receives them from a logical address and a physical address of the duplicated router. The network system in which the router is duplicated according to claim 1, wherein the network means is set as a holding means. 3. The communication control device obtains a logical address and a physical address assigned to the duplicated router, and obtains a physical address from the logical address transmitted to each duplicated router. 2. The router according to claim 1, wherein the router is duplicated, which is obtained from a response message which is a response to the request message and which is set as a means for holding a logical address and a physical address of the duplicated router. Network system. 4. When a failure occurs in the working router, the backup router is used as a new working router, and the working router is replaced so that the communication control device is duplicated. 7. A means for holding a logical address and a physical address of an established router, sets a logical address and a physical address assigned to the new working router and the new backup router, respectively. A network system in which the router according to any one of claims 1 to 3 is duplicated. 5. The physical address is MAC (Media Ac
cess control) address, the logical address is an IP (Internet Protocol) address, and a request message for a physical address from the logical address is an ARP (Address Resolution Protocol).
l) A network system in which any of the routers according to any one of claims 1 to 4 is duplicated, wherein a response message which is a response to the request message is an ARP response message. 6. A failure countermeasure method for a network system in which one or more networks are connected via a router and the router is duplicated, wherein the computer connected to the network has a communication control device. However, this computer communicates by means of a communication control device that each has, and one of the duplicated routers is a working router,
The other one is used as a spare router, and the same logical address and different physical addresses are assigned to each other, and the communication control device has means for holding the logical address and physical address of the duplicated router. When this communication control device communicates with another computer via the router, it communicates using the physical address of the working router, and periodically communicates with the duplicated router from the logical address to the physical address. A request message for an address is sent, and whether or not a response message that is a response to the sent request message is sent from the router is monitored for each duplicated router, and the response from the router is sent. If no message is sent, it means that the router has failed, and if the working router fails. When it is determined that an error has occurred, the network system in which the router is duplicated is characterized in that when communicating with another computer via the router, the communication is switched to use the physical address of the backup router. How to deal with failures. 7. The communication control device receives a logical address and a physical address assigned to the duplicated router from a computer, and receives them from a logical address and a physical address of the duplicated router. 7. A method for coping with a fault in a network system having dual routers according to claim 6, characterized in that the setting means is set. 8. The communication control device obtains a logical address and a physical address assigned to the duplicated router, and obtains a physical address from the logical address transmitted to each duplicated router. 7. The router according to claim 6, wherein the router is duplexed, which is obtained from a response message which is a response to the request message and which is set as a means for holding a logical address and a physical address of the duplexed router. Network system failure countermeasure method. 9. When a failure occurs in the working router, the protection router is used as a new working router, and the working router is replaced so that the communication control device is duplicated as a new protection router. 7. A means for holding a logical address and a physical address of an established router, sets a logical address and a physical address assigned to the new working router and the new backup router, respectively. A fault countermeasure method for a network system in which the router according to claim 6 is duplicated. 10. The physical address is MAC (Media
Access control) address, the logical address is an IP (Internet Protocol) address, and a request message for a physical address from the logical address is ARP (Address Resolution Protocol).
l) A failure countermeasure method for a network system in which any one of the routers is duplicated, wherein the response message which is a response to the request message is an ARP response message.