JPH0823352A - Datagram repeat system - Google Patents

Abstract

PURPOSE:To suppress a useless repeat by finding out a path early in which a fault takes place through polling so as to avoid the use of information of the faulty path. CONSTITUTION:A timer processing section 134 checks whether or not 1st path information is updated for 30sec. When not updated, the processing section 134 makes a request of transmission of an ICPM echo request addressed to a gateway of the entry to a fault detection processing section 135. The fault detection processing section 135 sends the ICPM request to a designated gateway destination upon the transmission request of the ICPM echo request by the timer processing section 134 and starts the reply wait timer. When the reply wait timer expires, it is discriminated that a fault takes place in the 1st relay path to the network and a count of path information in the routing table 133 is set to infinite to form the state of learning 2nd path information.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a datagram relay system, and more particularly to a datagram relay system in a data communication system using a LAN.

[0002]

2. Description of the Related Art FIG. 11 is a block diagram showing a configuration of a datagram relay processing device applied to a conventional datagram relay system. As a conventional datagram relay method,
For example, there is one described in Takeshi Nishida, "-LAN Connection-TCP / IP Internetworking" (Soft Research Center Co., Ltd., P303, 1993). In the relay processing device 200 shown in FIG. 11, reference numeral 210 is RFC (Request For Commn-ts) 79.
1 is an IP processing unit that processes an Internet protocol (IP) defined by 1; 220 is a U that processes a user datagram protocol (UDP) defined by RFC768.
The DP processing unit 230 is a RIP processing unit that processes a routing information protocol (RIP) defined by RFC1058.

The RIP processor 230 has an input processor 231 for processing the received route information, an output processor 232 for transmitting the route information to another device, a routing table 233 for recording the route information, and a routing table 233.
Is provided and a timer processing unit 234 for periodically transmitting route information is provided.

In the conventional datagram relay system, the datagram is relayed as follows. First, IP
The processing unit 210 discriminates whether the received data is addressed to the own device 200 or another device, and if it is addressed to the own device 200, the data is sent to the upper layer and is addressed to the other device. If so, the routing table 233 is searched.
As a result of this search, if the destination network is in the routing table 233 and the metric is within the range of 1-15, the data is transmitted to the gateway which is the first terminal to pass through. The metric here indicates the number of terminals through which the target network is reached. Further, if the received data is from the upper layer, the IP processing unit 210 similarly searches using the routing table 233 and transmits the data to the gateway which is the first terminal to pass through.

In the RIP processing section 230, the RIP request message or the RI which is a route information control message is sent.
The P response message is processed and the routing table 233 is managed. The input processing unit 231 scrutinizes the data received by the IP processing unit 210, and if the data is normal, performs processing according to the content. That is, if the received data is, for example, a RIP response message that advertises route information, the route information is extracted from the message and compared with the route information in the routing table 233, and the received route information is the routing table 233.
If the route information and the destination network are the same and the metric is smaller, the conventional route information is discarded and the received route information is registered in the routing table 233. When the route information in the routing table 233 and the received route information have the same destination network, metric, and gateway address, the route information in the routing table 233 is updated.

On the other hand, the output processing section 232 receives an instruction from the timer processing section 233 to periodically advertise the route information and an instruction from the input processing section 231 to respond to an inquiry from another device. A part or all of the information in the routing table 233 is transmitted in the RIP response message.

The timer processing unit 234 gives an instruction to the output processing unit 232 to transmit the information in the routing table 233 in order to advertise the route information every 30 seconds. Also, by monitoring the routing table 233,
If the route information in the routing table 233 is not updated for 180 seconds, the metric to the destination network is set to "16", which is a value indicating unreachability, and
After 20 seconds, the route information for the destination network is deleted from the routing table 233.

[0008]

Since the conventional datagram relay system of this type is configured as described above, when a failure occurs in the relay route and the relay route becomes unusable. , From the time it became unusable 1
It is not possible to detect that the route has become unavailable until 80 seconds have passed. Therefore, in the meantime, there is no choice but to relay data using the route with the fault.
There is a problem that the relay operation is useless.

Further, even if the destination network is the same and the metrics are equal to or larger than the route information, that is, the alternative circuit information is sent from the other device until the lapse of 180 seconds, learning is already performed during this period. Since the route information being provided is still valid, the newly sent alternative route information is not learned. Therefore, even if there is a bypass route for sending data to the destination network, there is a problem that it cannot be used.

The present invention has been made in order to solve such a problem, and it is possible to detect early that a failure has occurred in a relay path, and to relay data to the failed relay path. It is an object of the present invention to provide a datagram relay system in which useless data relay is suppressed as much as possible and further, if there is an alternative route, that route can be used immediately.

[0011]

A datagram relay system according to the present invention comprises a LAN each having a unique network address, a router connecting these LANs directly or via a WAN, a LAN or It is composed of one or more terminals connected to WAN, uses Internet Protocol (IP) as a network layer protocol, and Routing Information Protocol (R) as a routing protocol.
In the datagram relay method used in a communication network that uses IP (IP), if the entry indicating the first relay route to a certain network in the routing table is not updated for a predetermined time, the relay to the first route to that network A means for transmitting an Internet Control Message Protocol (ICMP) echo request to a terminal address indicating the destination, an Internet Control Message Protocol (ICM) within a predetermined time.
P) A means for judging that a failure has occurred in the first relay path to the network and not relaying the datagram to the network when the echo reply cannot be received Is characterized by.

Further, when a failure occurs in the first relay route to the network, when there is a second relay route having a different terminal address from the terminal address indicating the relay destination to be passed through first, It is characterized by further comprising means for registering the second relay route information in the routing table.

Further, when the second relay route information is registered, a means for relaying the datagram by using the second relay route information with priority over the first relay route is provided. Is characterized by.

Further, when the first relay route information is updated by the route information control message whose source is the first relay destination described in the entry indicating the first relay route to the network, this It is characterized in that a means for relaying the datagram is provided via the first relay route.

Further, when the first relay route information is not updated for a predetermined time, the first relay route information is deleted from the routing table, and the second relay route information is newly added to the first relay route information. Is provided with means for learning the above-mentioned routing table.

[0016]

In the datagram relay system of the present invention,
When the relay processing device does not receive the route information to the destination network for a certain period of time, the gateway address to the destination network address is used as the destination address, and R
Polling is performed using the ICMP echo specified in FC792, and if there is no normal response to this polling from the gateway address, the routing information to the destination network address in the routing table is infinitely measured. To do. By performing the polling, it becomes possible to detect the route in which the failure has occurred at an early stage, and by avoiding using the failed route information, it is possible to suppress useless relay operation.

Further, when a failure occurs in the first relay route to a certain network and the first route information is not used, it has the same destination network as the first route information, When the second route information with a different gateway address is received, this second route information is also entered in the routing table. Therefore, the alternative route can be grasped quickly.

In addition, the datagram is relayed by using the second route information newly described at that time, and if there is an alternative route, it is quickly caught and used immediately. Can communicate quickly. At this point, the first relay route information is left on the routing table.

Even when the datagram is relayed using the second route information, the first relay route is re-established from the device registered as the gateway of the route information. When the route information is sent, the second route information is deleted from the routing table, and thereafter, the relay route is determined based on the first route information. Therefore, when the failure that occurred in the first relay route, which is the route with the least amount of measurement, is recovered, it is possible to immediately return to this relay route and relay the datagram, further improving the relay efficiency. Be improved.

Further, when the datagram is relayed using the second route information, if the first route information is not updated for a certain period of time, the first route information is stored in the routing table. Delete it so that the second route, that is, the alternative route, is newly determined as the first route in the routing table, and even if the failure occurring in the first route information does not recover, Made it possible to perform relay efficiently.

[0021]

【Example】

Example 1. Embodiment 1 of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of a relay processing device that executes the datagram relay method of the present invention.
In FIG. 1, reference numeral 100 is a relay processing device, and 110 is R.
Internet Protocol (I
P) processing unit 120, a UDP processing unit 120 processing a user datagram protocol (UDP) defined by RFC768, and a routing information protocol 130 defined by RFC1058 (R).
It is a RIP processing unit for processing IP).

The RIP processing unit 130 processes an input processing unit 131 that processes the received route information, an output processing unit 132 that transmits the route information to another device, and a routing table 13.
3. A timer processing unit 134 for monitoring the routing table 133 and periodically transmitting route information,
A failure detection processing unit 135 for detecting a failure on the relay route.

2 is a flow chart showing the operation of the timer processing unit 134, FIG. 3 is a flow chart showing the operation of the failure detection processing unit 135, and FIG. 4 is a flow chart showing the operation of updating the routing table 133 in the input processing unit 131. 5 is a flowchart showing the relay processing operation in the IP processing unit 110. The operation of the relay processing device 100 will be described below based on these flowcharts.

First, the timer processing unit 134 will be described with reference to FIG.
Will be described. The timer processing unit 134 is activated periodically, and first updates the timer count indicating the time since the route was last updated for all the route information entries in the routing table 133 (step S201). . Here, it is checked whether or not a certain entry has learned only the first route information and the first route information has not been updated for 30 seconds (step S202). If it has been updated (N
O) proceeds to step 204 described later, and if not updated (YES), the fault detection processing unit 135 is addressed to the gateway of the entry by ICMP (Internet Control).
l Message Protocol) Request to send an echo request (step S203).

Then, it is checked whether or not the entry has the second route information (step S2).
04), if it has the second route information, it is 1
It is confirmed whether it has been updated for 80 seconds (step S205). If the entry has not been updated for 180 seconds, the first route information is deleted from the routing table 133 and the second route information is updated as the first route information (step S206), and the timer processing is performed. Ends.

Next, based on FIG. 3, the fault detection processing unit 135
Will be described. ICP in the timer processing unit 134
Request to send M echo request (FIG. 2: step S20
I) to the specified gateway destination according to 3).
Send a CPM echo request (step S30
1) and then start a response waiting timer (step S3)
02). Here, the ICPM echo reply is received (step S303), or the process waits until the response waiting timer times out (step S304), and the process ends when the IPCM echo reply is received. On the other hand, when the response waiting timer times out, the metric of the corresponding route information in the routing table 133 is set to infinity so that the second route information can be learned (step S30).
5), the process ends.

Next, the operation of the input processing unit 131 will be described with reference to FIG. The input processing unit 131 receives R from another device.
When the route information is received by the IP response (step S401), the following process is performed for each route information entry. First, the metric of the route information sent is checked (step S402), and if it is infinite, the information is ignored (step S411), and the process ends. On the other hand, if it is not infinity, the gateway of the route information sent is used as the address of the device that sent the route information, and 1 is further added to the metric to process (step S403).

Next, it is checked whether the received route information already exists in the routing table 133 (step S404), and if it does not exist, the received route information is learned in the routing table as the first relay route. (Step S409) The process ends. On the other hand, if it exists, the received route information is compared with the route information existing in the routing table (step S405). Here, when the metric of the received route information is smaller, the route information existing in the routing table is discarded and the received route information is learned in the routing table 133 (step S40).
9) End the process.

In other cases, that is, when the metric of the received route information is equal to or larger than the metric of the route information existing in the routing table, the gateway address of the received route information and the routing table The gateway address of the route information existing in 133 is compared (step S406), and if they are the same, the route information existing in the routing table 133 is updated (step S410) and the process ends. . If the gateway addresses are different, it is checked whether the second route information can be learned (step S407). If the second route information can be learned, the received route information is used as the second relay route in the routing table 13.
3 to learn (step S408), and if the learning is not possible, ignore the received route information (step S41).
1) The process ends.

Next, based on FIG. 5, the IP processing unit 110
The operation when the data relayed by is received will be described. First, when receiving the relay data (step S501), the IP processing unit 110 checks whether or not the first route information to the destination IP address of the relay data exists in the routing table (step S502). If the first route information does not exist, the data is discarded (step S507) and the process ends, and if it exists, it is checked whether the metric is infinite (step S50).
3).

If the metric is not infinite, the data is relayed according to the first route information (step 504), but if the metric of the first route information is infinite, the second route information is used. It is checked whether or not the route information is present (step S505), and if the second route information is present, the data is relayed accordingly (step S50).
6). If the second route information does not exist, the data is discarded (step S507) and the process ends.

FIG. 6 is a diagram showing a configuration example of a network to which the datagram relay system of the present invention is applied. Here, reference numerals 1, 2, and 3 are relay devices using the datagram relay system according to the present invention, and reference numerals Na and N are used.
b, Nc, Nd, and Ne are networks, and Na
1, Na2, Nb2, Nb3, Nc1, Nc3, Nd1
Indicates the IP address of the interface that connects each device to the network.

FIG. 7 is a diagram showing the contents of the routing table when all the network systems shown in FIG. 6 are operating normally. The routing table 133 of the relay device 1 has information having the contents shown in FIG. 7 when all the network systems are operating normally. Here, if some failure occurs in the network Na and communication between the relay apparatuses 1 and 2 is disabled, the route information from the gateway Na2 is included in the route information of the routing table shown in FIG. It will not be sent. Therefore, the route to the destination network Ne is not updated in the device 1.

When 30 seconds have passed in this state, the relay device 1
Sends an ICPM echo request to Na, which is the gateway of the destination network Ne. However, since the network Na has a failure, this echo request also does not reach the relay device 2, and therefore the relay device 1 cannot receive the echo reply. Here, the relay device 1 determines that a failure has occurred in the first relay route destined for the network Ne and sets the metric of this route to infinity “16” so as not to perform the relay, and the routing table 133. To learn the second relay route.
FIG. 8 is a diagram showing the contents of the routing table at this time.

After determining that a failure has occurred in the first relay route to the network Ne, the relay device 1 switches to the relay device 3
When receiving the relay route information having the destination network as Ne from, the device 1 causes the routing table 133 to learn this information as the second relay route. FIG. 9 is a diagram showing the contents of the routing table 133 at this time. Here, the relay device 1 uses the second route having the gateway Nc3 as the relay route to the network Ne.

Further, when the first route information in which the destination network is Ne and the gateway is Na2 in the relay device 1 is not updated for 180 seconds, the first route is deleted from the routing table 133 of the device 1, The second route having Ne as the destination network and Nc3 as the gateway is updated as the first relay route. FIG.
0 indicates the contents of the routing table at this time.

[0037]

As described above, according to the datagram relay system of the present invention, when a failure occurs in the data relay path, the failure can be detected early and the failure has occurred. It is possible to reduce unnecessary transmission of data to the relay route.

Also, by writing the second route information also in the routing table, the alternative route can be grasped quickly.

Further, since the second route information can be learned, if there is an alternative route, it can be promptly used to relay data.

When the faulty relay route is restored, the relay route can be smoothly returned to the original route.

Further, when the failure of the relay route is not recovered for a predetermined time or longer, there is an effect that the alternate route is registered as the first relay route so that the relay can be efficiently performed.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration example of a datagram relay processing device that realizes a datagram relay system of the present invention.

FIG. 2 is a flowchart showing an operation of a timer processing unit in the relay processing device shown in FIG.

FIG. 3 is a flowchart showing an operation of a failure detection processing unit in the relay processing device shown in FIG.

FIG. 4 is a flowchart showing an operation of an input processing unit in the relay processing device shown in FIG.

5 is a flowchart showing an operation of an IP processing unit in the relay processing device shown in FIG.

FIG. 6 is a diagram showing a configuration example of a communication network to which the datagram relay system of the present invention is applied.

FIG. 7 is a diagram showing the contents of a routing table when the network shown in FIG. 6 is operating normally.

8 is a diagram showing the contents of a routing table when a failure occurs in the network shown in FIG.

FIG. 9 is a diagram showing the contents of a routing table when a second relay route is learned.

FIG. 10 is a diagram showing the contents of a routing table when a second relay route is registered.

FIG. 11 is a block diagram showing an example of the configuration of a datagram relay processing device that realizes a conventional datagram relay method.

[Explanation of symbols]

100 datagram relay processing device, 110 IP processing unit, 120 UDP processing unit, 130 RIP processing unit, 1
31 input processing unit, 132 output processing unit, 133 routing table, 134 timer processing unit, 135 failure detection processing unit.

Claims (5)

[Claims] 1. LANs each having a unique network address, and a direct or WA connection between these LANs.
It is composed of a router connected to each other via N and one or a plurality of terminals connected to the LAN or WAN, using Internet Protocol (IP) as a protocol of a network layer, and Routing Information Protocol ( R
In the datagram relay method used in a communication network using (IP), if the entry indicating the first relay route to a certain network in the routing table is not updated for a predetermined time, the relay to the first route to that network A means for transmitting an Internet Control Message Protocol (ICMP) echo request to a terminal address indicating the destination, a first means for transmitting to the network when the Internet Control Message Protocol (ICMP) echo reply cannot be received within a predetermined time. A datagram relay method comprising means for judging that a failure has occurred in a relay route and not relaying the datagram to the network. 2. When a failure has occurred in the first relay route to the network, and when there is a second relay route having a different terminal address from the terminal address indicating the relay destination to be passed through first. The datagram relay system according to claim 1, further comprising means for registering the second relay route information in the routing table. 3. When the second relay route information is registered, a means for relaying a datagram using the second relay route information with priority over the first relay route is provided. The datagram relaying method according to claim 2, wherein 4. When the first relay route information is updated by a route information control message whose source is the first relay destination described in the entry indicating the first relay route to the network, 4. The datagram relay system according to claim 3, further comprising: means for relaying a datagram via the first relay path. 5. When the first relay route information is not updated for a predetermined time, the first relay route information is deleted from the routing table, and the second relay route information is newly added to the first relay route information. The datagram relay system according to claim 4, further comprising means for learning the routing table as relay route information.