JP2001217869A - Ip packet transfer control system and method and recording medium with recorded processing program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve a problem that traffic loads in an IP network cannot efficiently be uniformized and decentralized in a conventional technology. SOLUTION: In order to level an extreme traffic unbalance in links in an IP network through decentralization of loads, when traffic is concentrated on links between optional nodes 2 (n), 6 (n), 7 (D) in the IP network, one node (a node 4 (B) or a node 8 (C)) in a node group in the vicinity of the substantial termination node 7 (D) is decided as a tentative termination point, a flow is transferred thereto and the flow is transferred from the tentative termination point to the termination point node 7 (D) to decentralize the traffic.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an IP (Internet P
network (I) that transfers data by rotocol
The present invention relates to a load distribution technique for distributing traffic in a P network), and particularly to an I / O network suitable for efficiently reducing congestion.
The present invention relates to a P-packet transfer control system and method, and a recording medium on which a processing program is recorded.

[0002]

2. Description of the Related Art Conventionally, in an IP network, a load distribution technique to be adopted when a certain link becomes congested has not been particularly considered. No particular countermeasure technology has been considered as a congestion control technology, and a technology for eliminating a congestion state by discarding a packet has been adopted in a node which has received an excessive load.

Further, for example, IPv6 (Internet Proto
col Version 6), defined in the addressing system, is the “anycast” technology of IETF (Internet Engineering Task Force) RF
C (Request For Comment) 1546. This technology is called "anycast address."
A packet that identifies an interface group is defined, and a packet sent to that address is sent to one of the interfaces in the group indicated by that address (this is determined by the The closest one in the group is selected).

However, in this anycast address technology, a packet is delivered to an interface that is uniquely determined in the entire network, and it is not possible to perform load distribution and congestion control related to transfer of an IP packet based on a normal destination. Can not.

[0005]

The problem to be solved is that the conventional technology cannot efficiently equalize and distribute the traffic load in the IP network.

SUMMARY OF THE INVENTION An object of the present invention is to solve the problems of the prior art and to record an IP packet transfer control system and method capable of alleviating congestion by eliminating uneven traffic load in an IP network, and recording a processing program therefor. The purpose of the present invention is to provide a recording medium.

[0007]

In order to achieve the above object, in an IP packet transfer control system and method according to the present invention, at an arbitrary node A serving as a source address,
For example, if the de facto link to the destination node D that accommodates the destination terminal of the received IP packet is congested, the header of the IP packet is newly connected to a link other than the congested de facto link. In addition, the address of the relay node B close to the node D is assigned as the “temporary end point address”, and the data is transferred to the node B (temporary end point node). Each node between the node A and the node B transfers the IP packet based on the “temporary destination address” of the received IP packet, and deletes the “temporary destination address” at the node B that has received the IP packet, From this node B, the packet is transferred to the node D and the destination terminal according to the normal destination address (IP address). In addition, from node A (source node) to node B
The node B (temporary end node) is selected so that the link to the (temporary end node) does not overlap with the link from the node B (temporary end node) to the node D (end node).

Further, a node near the local congestion area where the network is congested in a certain area is selected as an evacuation node Z, and each IP packet destined for the destination within the local congestion area is assigned with a “temporary destination address”. The data is transferred to the evacuation node Z, accumulated in the evacuation node Z, and after the local congestion has subsided, the transfer is performed based on the normal destination address while shaping from the evacuation node Z.

[0009]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing a configuration example according to the present invention of the IP packet transfer control system of the present invention.

In FIG. 1, nodes 1 to 8 are routers or gateways for controlling transfer of IP packets (routing), and 10 and 20 are LAN (Local).
AreaNetwork), 11 is a computer terminal installed in the LAN 10 (described as “host α” in the figure), 21 is LA
A computer terminal (described as “host β” in the figure) installed in N20, and 30 to 35 are IP packets transmitted through the link between the nodes 1 to 8.

In the LAN 10, "10.1.1.
Each computer terminal to which an IP address of “0/24” is set is accommodated, and the IP address of the computer terminal 11 is “10.1.1.3”. Node 1
(A) The IP address of the (router) on the LAN 10 side is “1”.
0.1.1.1 ".

In the LAN 20, "10.1.2.
Each computer terminal to which an IP address of “0/24” is set is accommodated, and the IP address of the computer terminal 21 is “10.1.2.10.”. Further, the IP address of the node 7 (D) (router) on the LAN 20 side is “10.1.1.2.1”.

The address of node 1 (described as “A” in the figure) is “10.10.0.1”, and the address of node 4 (“B” in the figure)
Is described as "10.1.0.3" and the node 7
The address of (described as “D” in the figure) is “10.1.0.
4 ", and the address of node 8 (described as" C "in the figure) is" 1 ".
0.1.0.2 ".

The operation of the IP packet transfer control system having such a configuration will be described with reference to the transfer of an IP packet from the computer terminal 11 to the computer terminal 21 as an example. In this case, the node 1 (A) accommodating the computer terminal 11 becomes the source node, and the computer terminal 21
Is the end node (destination node)
Becomes

A series of packets, that is, a certain information flow (when sending certain application information from a sender to a receiver) is transmitted from the node 1 (A) to the computer terminal 21 connected to the node 7 (D). Packet flow) is transmitted. In this case, the original route is the route indicated by the dotted line in the figure, and the destination “1”
IP packet 30 having the address information of “0.1.1.20” and the source “10.1.3.1.3” in the header is transmitted from node 1 (A) to node 2 (n) and node 6 (n). Is transferred to the node 7 (D) via

However, in this example, node 1 (A)
First, a flow based on a certain condition, for example, a case where the amount of information to be transmitted satisfies a condition that the amount of information to be transmitted is equal to or more than a reference value, is determined to be normal (hereinafter, a high order in a congestion state). In contrast to "abnormal transfer processing" in which layer processing is not performed, normal processing for performing processing up to the higher layer processing is called "normal" processing), and is switched to "temporary end point transfer control" according to the present invention. Decide that.

Then, a timer (T
n) (Tn = O). The “temporary end point transfer control” relating to this flow is based on the assumption that this timer (Tn) has a value (Tz:
Until Tn = Tz, the process returns to the normal transfer control.

In the “temporary end point transfer control” in the node 1 (A), first, based on the routing table,
One node is randomly selected as a temporary end node from the nodes around the node 7 which is the end node. In this selection, when the IP packet is transferred from the temporary end node to the end node (node 7) which is the original end point based on the normal destination address, the node 1 (A) of the source node
A node that satisfies the condition that the IP packet can be reached without passing through the transfer route of the IP packet from the to the temporary end node is selected and specified as the temporary end node.

Here, node 4 (B) or node 8 (C) is selected as a temporary end point node. Then, the address "10.1.0.3" of the node 4 (B) and the address "10.1.0.2" of the node 8 (C) are determined as the temporary end point addresses.

Then, the node 1 (A) places the determined temporary end address (“10.1.0.3”) at a predetermined location in the header of the packet group (packet group on the session) from the computer terminal 11. , “10.1.0.2”).

It is known at each node that a predetermined location in the header is used for the "temporary end point transfer control". The value of a predetermined location in this header is “empty” or a temporary end address is described,
If the temporary destination address is described (not empty), the IP packet is sent to the node 7 (D), which is the destination node.
Priority is given to this temporary destination address.

That is, as shown in the IP packet 31, the address of the node 4 (B) or the node 8 (C) (“10.
1.0.3 ”and“ 10.1.0.2 ”), and as a result, all IP packets from the computer terminal 11 addressed to the computer terminal 21 are the node 4 (B) or the node as the temporary end node. 8 (C) via the default route.

In the case of the IP packet 32 addressed to the node 4 (B), the IP packet 32 is transferred to the node 4 (B) via the node 2 (n) and the node 3 (n), and is transferred to the node 8 (C). In the case of 34, the node 8 via the node 2 (n)
(C).

In the node 4 (B) as the temporary end node, the IP address assigned the same temporary end address as its own address
When the packet 32 is received, the temporary end address “10.1.0.3” is deleted from the IP packet 32 and “empty”
Then, the data is transferred to the destination computer terminal 21 under normal transfer control, that is, according to the IP address. In each of the subsequent nodes 5 (n) to 7 (D), the IP
Since the “temporary destination address” of the packet 33 is empty,
Normal transfer according to the P address is performed. This route is the route indicated by the arrow in the figure.

The node 8 as a temporary end point node
In (C), when receiving the IP packet 34 to which the same temporary end address as the own address is added, the temporary end address “10.1.0.2” is deleted from the IP packet 34 to be “empty”, The data is transferred to the computer terminal 21 under normal transfer control according to the address. In each of the subsequent nodes 7 (D), since the “temporary destination address” of the IP packet 35 is empty, normal transfer according to the IP address is executed. This route is the route indicated by the arrow in the figure.

The configuration of each of the nodes 1 to 8 will be described below with reference to FIGS. FIG. 9 is a block diagram showing an example of the internal configuration of each node in FIG. 1, and FIG. 10 is a block diagram showing an example of the hardware configuration of each node in FIG.

In FIG. 10, reference numeral 101 denotes a CRT (Cathod).
e. a display device such as an e Ray Tube) or an LCD (Liquid Crystal Display); 102, an input device such as a keyboard and a mouse; 103, an external storage device such as a hard disk drive; 104, a CPU (Central Processing).
An information processing device having a storage unit 104a, a main memory (main storage device) 104b, and the like and performing computer processing according to a storage program method; 105, an optical disk on which a processing program and data according to the present invention are recorded; A drive device 107 for performing an operation is a communication device for controlling communication with another node via a communication line.

By installing the processing programs and data recorded on the optical disk 105 and reading them into the main memory 104b and processing them by the CPU 104a, the processing functions shown in FIG.

The configuration of the node 9 shown in FIG. 9 is common to each of the nodes 1 to 8 in FIG.
A routing table 9a used to determine the transfer destination of the P packet; a provisional destination address determination unit 9b for determining whether a provisional destination address is set at a predetermined location in the header of the received IP packet; A judgment / switching unit 9c for judging, based on predetermined conditions, whether the packet transfer should be normal transfer control or temporary end point transfer control, and a normal transfer processing unit 9d for performing normal transfer control.
And a temporary end point transfer processing unit 9e that performs transfer control in the temporary end point transfer control.

When the temporary end address determining unit 9b determines that the temporary end address is not set at a predetermined position in the header of the received IP packet, the node 9 performs normal transfer control by the determining / switching unit 9c. And provisional end point transfer control. For example, if it is determined that the amount of information to be sent is equal to or larger than a certain amount, the switching to the temporary end point transfer control is determined.

If normal transfer control is determined by the judgment / switching unit 9c, the normal transfer processing unit 9d receives the received I
The P packet is transferred based on the routing table 9a according to the normal transfer procedure. When the switching to the temporary end point transfer control is determined, the transfer is performed by the temporary end point transfer processing unit 9e.

Further, the temporary end point address judging section 9b
If it is determined that the temporary end address is set at a predetermined position in the header of the received IP packet, the node 9
Performs the transfer by the temporary end point transfer processing unit 9e without performing the processing of the determination / switching unit 9c.

The temporary end point transfer processing section 9e is provided with a temporary end point address comparing section 9f, a temporary end point address deletion transfer section 9g, a temporary end point node determining section 9h, a temporary end point address setting section 9i, and a sending section 9j. The temporary destination address comparison unit 9f searches the routing table 9a to confirm whether the temporary destination address set in a predetermined location of the header of the received IP packet is the address of the own node.

If the temporary end address comparison unit 9f confirms that the temporary end address is the address of the own node, the temporary end address deletion transfer unit 9g receives the I
The temporary end address in the header of the P packet is deleted, an output destination corresponding to a normal destination address is searched for in the routing table 9a, and specified and output. If the temporary end address of the received IP packet is not the address of the own node in the determination result of the temporary end address comparing unit 9f, the sending unit 9j searches the routing table 9a for the output destination corresponding to the temporary end address as it is. Specify and output.

Further, the temporary end address determining section 9b
If it is determined that the temporary end address is not set at a predetermined position in the header of the received IP packet, and if the determination / switch unit 9c determines to switch to the temporary end point transfer control, the temporary end point transfer processing unit 9e First, the temporary end node determining unit 9h searches the routing table 9a to extract a node located near the end node, and further performs the above-described predetermined condition (from the node concerned, the node which is the original end point). In the case where an IP packet is forwarded based on a normal destination address, a node that satisfies the condition that the IP packet can be reached without passing through a transfer route of the IP packet from the source node to the node) is determined as a temporary end node.

Using the address of the temporary end node determined by the temporary end node determination unit 9h, the temporary end address setting unit 9i sets a temporary end address at a predetermined location in the header of the received packet. The IP packet in which the address is set is routed by the transmitting unit 9j.
With reference to the table 9a, the data is sent to the temporary end node.

FIG. 2 is a flowchart showing an operation example of the source node in FIG. 1, and FIG. 3 is a flowchart showing an operation example of the temporary end point node in FIG.

The processing example in FIG. 2 corresponds to the node 1 in FIG.
FIG. 3A shows an example of processing as a source node (source node), and each time an IP packet arrives (step 2).
01), it is determined whether or not this IP packet is a packet of a flow that has already been switched to the temporary end point (step 2).
02).

If not, it is determined whether or not the amount of information addressed to the end node of this flow exceeds a reference value (step 203), and if not, normal processing is performed (step 204). If it is exceeded, it is determined that this flow is to be switched to the temporary end point transfer control (temporary end point transfer method) (step 205), and a timer is set (Tn = 0) for this flow (step 20).
6).

Then, with respect to this flow, a candidate node satisfying the provisional end point condition is selected (steps 207 and 207).
8), the address of the node is described in the header in this packet (step 209), and the packet is transferred to the temporary end node (step 210).

According to the result of the determination processing in step 202, if
If the arriving packet is a packet of a flow that has already been switched to the temporary end point, it is determined whether or not the timer value Tn of this flow exceeds the duration value (Tz) of the temporary end point transfer method (step 211). ). If not, the packet is transferred to the temporary end point (step 21).
0). If so, the temporary end point transfer method is released for this flow (step 212).

Next, the processing example in FIG. 3 shows a processing example of the node 1 (A) in FIG. 1 as a temporary end point node, and every time a packet arrives (step 301), the header in the packet is It is determined whether or not the temporary end point address is “empty” (step 302). If "empty",
Transfer according to the original IP address (step 30)
6). If it is not "empty", it is determined whether the temporary end address is the same as the own address (step 30).
3). If it is the same as the own address, the temporary end address is deleted (step 305), and the transfer is performed according to the original IP address (step 306). If it is not the same as the own address, the transfer is performed according to the temporary end address (step 304).

As described above, in the IP packet transfer control system according to the present embodiment, when traffic concentrates on a certain node in the IP network, the IP packet transfer control system performs an operation in a group of nodes near the original destination node. By defining one of the nodes as a temporary end point node and transferring the flow to that node, traffic can be distributed.
As a result, extreme imbalance of the links in the IP network can be equalized by load distribution.

Further, the control by such an IP packet transfer control system can be applied to control of a congestion state. Hereinafter, another example of the embodiment in the case where control by the IP packet transfer control system of the present invention is applied to control of a congestion state will be described.

FIG. 4 is a block diagram showing another example of the configuration of the IP packet transfer control system according to the present invention. In FIG. 4, reference numeral 40 denotes a local congestion range indicating a range in which congestion occurs. Nodes 41 to 49 perform transfer control (routing) of the IP packet, and the nodes 43 (B) to 45 (C) are in the local congestion range 40 and are in a local congestion state at the present time.

The node 41 (Z1) and the node 47
(Z2) is an evacuation node near the outside of the local congestion range 40 and provided with large-capacity buffers 41a and 47a, respectively. With such a configuration, in the IP packet transfer control system according to the present embodiment, the congestion in the local congestion range 40 can be calmed down efficiently.

Hereinafter, a node 44 indicated by an arrow in FIG.
The operation will be described taking an IP packet going to (A) as an example. In a normal state, the IP packet destined for the node 44 (A) is transmitted to the nodes 42 (D) and 43.
Although the node 44 (A) arrives at the node 44 (A) through (B), the node 44 (A) which is currently in a local congestion state transmits itself a “control packet” to a source node (not shown), thereby causing itself to become congested. Notify that it is in status.

The source node (source node), having learned that the node 44 (A) is in a congestion state, sets "1" in a certain "congestion bit" to perform congestion control, and Evacuation node 4 closest to 44 (A)
1 (Z1) is specified, the address of the evacuation node 41 (Z1) is selected as a "temporary destination address", and an IP packet is transmitted toward it.

FIG. 5 is a block diagram showing a result of congestion control in the IP packet transfer control system in FIG.
As shown in FIG. 5, the IP packet sent from the source node (originating node) is sent to the node 42 (D) based on the “temporary destination address” and not to the node 43 (B) that is the original route. It is transferred to the evacuation node 41 (Z1).

In this way, the evacuation node 41 (Z1) that has received the IP packet sent from the source node from the node 42 (D) has the “temporary end point” as its own node and
If “1” is set in the “congestion bit”, this I
The P packet is regarded as an evacuated packet, and is stored in a huge capacity buffer 41a in the evacuation node until the congestion state subsides.

When the evacuation node 41 (Z1) knows that the congestion in the local congestion area 40 has subsided, the evacuation node 41 (Z1) shapes the group of IP packets stored in the buffer 41a, and returns to the node 44 which is the original end node. Transfer to (A).

As described above, by limiting the amount of packets transferred to the congested node 44 (A), the congestion state in the local congestion range 40 can be effectively alleviated.
Further, in the conventional technique, it is possible to rescue a packet that would be lost in a node in a congested state.

The IP packets of the flow indicated by the arrows in FIGS. 4 and 5 are also temporarily evacuated to the evacuation node 47 (Z2) having the large-capacity buffer 47a in the same manner as described for the IP packets of the arrows. By doing so, the congestion in the local congestion range 40 can be reduced.

FIG. 6 is a flowchart showing an operation example of the source node of the IP packet transfer control system in FIG. 4, and FIG. 7 is a flowchart showing an operation example of the evacuation node in FIG.

In FIG. 6, if a source node (source node) of the IP packet transfer control system in FIG. 4 receives a control packet (step 601), it identifies a congested node from the information of the control packet, and Add to the list (step 602). Thereafter, it is determined whether or not a packet has arrived (step 603), and if so, it is determined whether or not there is a congested node and whether or not the congestion is continuing (step 604).

If the congestion is continuing, it is determined whether this packet is destined for a congested node (step 60).
7) If so, select / designate an evacuation node (step 608); otherwise, perform normal processing (step 606). Also, as a result of the determination in step 604, if there is no congested node, or if the congestion has subsided, the congestion list is updated (step 60).
5) The process proceeds to normal processing (step 606).

When the evacuation node is selected / designated in the process of step 608, the address of the evacuation node is written in the header of the packet specifying the evacuation node, and "1" is set to the congestion bit (step 609), and transfer to the temporary end point (step 610).

Next, as shown in FIG. 7, the evacuation node first determines whether a packet has arrived (step 701). If there is no arrival, it is determined whether or not packets are accumulated in the buffer and congestion has been alleviated (step 702). If the congestion is reduced, the packets stored in the large-capacity buffer are read (step 711), the temporary end address is deleted (step 707), and the original IP address is deleted while shaping (step 708). (Step 709).

When the packet arrives (step 7)
01), it is determined whether or not the temporary end address of this packet is “empty” (step 702). If it is “empty”, transfer is performed according to the original IP address (step 709).
If it is not "empty", it is determined whether the temporary end address is the same as its own address (step 703). If it is the same, it is determined whether the congestion bit is "1" (step 705).

If the congestion bit is "1", it is stored in a large capacity buffer (step 706). If the congestion bit is not "1", the temporary end point address is deleted (step 707), and shaping is performed (step 70).
8) Transfer according to the original IP address (step 709). If it is determined in step 703 that the temporary end address is not the same as the own address, the packet is transferred according to the temporary end address (step 704).

FIG. 8 is a block diagram showing a configuration example of the evacuation node in FIGS. 4 and 5. The evacuation node 80 in this example includes a clock 81, a timer 82, a routing table 83, a packet header analysis / calculation unit 84, a header read measurement unit 85, a packet header write / control unit 86, and a large capacity buffer for a temporary end point. 87, buffer write / read control unit 8
8 and a scheduling unit 89.

The clock 81 indicates the current time synchronized with the network, and the timer 82 measures the time set based on the clock 81. The routing table 83 is a path control table indicating a packet transfer path, and the contents are updated based on update information with another node.

The packet header analysis / calculation section 84 obtains destination information of the packet by looking up the routing table 83 by computer processing, and performs analysis for writing destination and congestion information as packet header information. The calculation related to the congestion control based on the measurement result of the timer 82 is performed.

The header reading / measuring section 85 reads information from the packet header, measures the destination, each class, congestion information, and the like, and transfers the information to the packet header analyzing / calculating section 84. The packet header writing / control unit 86 performs congestion control and writes destination information to a packet header based on the analysis / calculation result of the packet header analysis / calculation unit 84.

The buffer write / read control unit 88, based on the congestion control instruction from the packet header write / control unit 86, provides a temporary end point large capacity buffer 8 for evacuated packets.
7 or a temporary end point large capacity buffer 87
The evacuation packet stored in the evacuation packet is read.

The scheduling unit 89 performs congestion information processing and layer instructions.
1, a packet processing unit 92, and a shaper 93.
Is temporarily stored, the header information is read by the header reading / measuring unit 85, and the header information is read by the packet processing unit 92 or the buffer writing / reading control unit 88
The data is written into the temporary end point large capacity buffer 87.

In the packet processing section 92, based on the instruction from the packet header writing / control section 86, the IP buffer stored in the FIFO buffer 91 or the temporary end point large capacity buffer 87 is stored.
It performs packet processing such as adding header information to the packet. The shaper 93 outputs the packets 94 at intervals of a certain value or more.

As described above with reference to FIGS. 1 to 10, in the IP packet transfer control system of this embodiment, when traffic is concentrated on a certain node in the IP network, the original By deciding one node in a group of nodes near the end point node as a temporary end point and transferring a flow toward the temporary end point, traffic can be distributed. As a result, extreme imbalance of links in the network can be equalized by load distribution.

The control by the IP packet transfer control system is applied to the control of the congestion state, and the packet for the node in the congestion state is temporarily stored in the evacuation node and evacuated. By returning the packet to the original destination after returning to the normal state, it is possible to save a packet that would have been lost in a congested state in the related art.

The present invention is not limited to the examples described with reference to FIGS. 1 to 10, and can be variously modified without departing from the gist thereof. For example, in this example, the node 7 (D) is set as the end node. However, in the present invention, a node that is recognized as being closest to the destination computer terminal is determined as the end node. In this example, the program is installed using an optical disk as a recording medium. However, an FD (Flexible Disk) may be used as a recording medium, or the program may be installed from another computer via the communication device 107. May be downloaded and installed.

[0071]

According to the present invention, when traffic concentrates on a certain node in an IP network, a temporary end node is determined and an IP packet is transferred toward the temporary end node. It is possible to efficiently equalize and distribute the traffic load in the network, and change the node provided with a large-capacity storage device to a temporary end node (evacuation node).
If a certain node is in a congested state, a packet for the node is stored in an evacuation node and evacuated, and when the node returns to a normal state, a packet is transferred from the evacuation node to the node. It is possible to easily mitigate congestion in the IP network, save a packet that should be lost in a congested state, and improve the performance of the IP network.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration example according to the present invention of an IP packet transfer control system of the present invention.

FIG. 2 is a flowchart illustrating an operation example of a source node in FIG. 1;

FIG. 3 is a flowchart illustrating an operation example of a temporary end point node in FIG. 1;

FIG. 4 is a block diagram showing another example of the configuration of the IP packet transfer control system according to the present invention;

5 is a block diagram showing a result of congestion control in the IP packet transfer control system in FIG.

FIG. 6 is a flowchart showing an operation example of a source node of the IP packet transfer control system in FIG.

FIG. 7 is a flowchart illustrating an operation example of an evacuation node in FIG. 4;

FIG. 8 is a block diagram showing a configuration example of an evacuation node in FIGS. 4 and 5;

FIG. 9 is a block diagram illustrating an example of the internal configuration of each node in FIG. 1;

FIG. 10 is a block diagram illustrating a hardware configuration example of each node in FIG. 1;

[Explanation of symbols]

1 to 9: node, 9a: routing table, 9
b: provisional end point address discriminating section, 9c: judgment / switching section, 9
d: normal transfer processing section, 9e: temporary end point transfer processing section, 9f:
Temporary end point address comparing section, 9g: Temporary end point address deletion transfer section, 9h: Temporary end point node determining section, 9i: Temporary end point address setting section, 9j: Sending section, 10, 20: LAN, 11: Computer terminal ("host α" ), 21: computer terminal (“host β”), 30-35: IP packet, 4
0: local congestion range, 41: node ("evacuation node"), 42 to 46, 48, 49: node, 47: node ("evacuation node"), 41a, 47a: buffer, 8
0: evacuation node, 81: clock, 82: timer, 83: routing table, 84: packet header analysis / calculation section, 85: header reading measurement section, 86: packet header writing / control section, 87: provisional Large capacity buffer for end point, 88:
Buffer write / read control unit, 89: scheduling unit, 90: packet, 91: FIFO buffer, 92:
Packet processing unit, 93: shaper, 94: packet,
101: display device, 102: input device, 103: external storage device, 104: information processing device, 104a: CPU, 1
04b: main memory (main storage device), 105: optical disk, 106: drive device, 107: communication device.

Claims (9)

[Claims] 1. An IP packet transfer control system for transferring a packet in an IP network between a plurality of nodes according to an IP address, the system comprising: a node other than an end node to which a received packet is transferred according to a destination address of the packet; Identify one as the temporary end node,
A source node that assigns the address of the temporary end node (temporary end address) to the received packet and transfers the packet in accordance with the temporary end address, and receives the packet from the source node and assigns the temporary end point to the packet. A relay node that forwards the packet according to the address, and a temporary destination node that forwards the received packet according to the destination address of the packet, when receiving the packet having the same temporary destination address as the own address and deleting the temporary destination address. Characteristic IP packet transfer control system. 2. The IP packet transfer control system according to claim 1, wherein the source node transmits the packet transferred from the temporary end node to the destination node according to a destination address of the packet. A temporary end node determining means for specifying, as the temporary end node, a node that satisfies a condition that can be reached without passing through a route that is transferred when transferring from the own node to the temporary end node according to the end address. IP packet transfer control system. 3. The IP packet transfer control system according to claim 1, wherein each of the source node, the relay node, the temporary end node, and the end node has a predetermined location in a header of a received packet. A temporary end point address determining means for determining whether or not the temporary end point address is set, and if the temporary end point address is not set as a result of the determination by the temporary end point address determining means, Determining / switching means for determining whether or not the provisional end point address needs to be added based on a predetermined condition; and determining the addition of the temporary end point address to the received packet by the determination / switching means. The packet transferred from the temporary end node to the destination node according to the destination address of the packet is Temporary end node determining means for specifying, as the temporary end node, a node that satisfies a condition that can be reached without passing through a route that is transferred from the own node to the temporary end node according to the temporary end address according to the temporary end node; The temporary end address setting / sending means for adding the address of the temporary end point node specified by the determining means to the received packet and transferring the received packet according to the address, and the temporary end address determining means, If the destination address is set,
A temporary end point address comparing means for confirming whether or not the temporary end point address is the address of the own node; and a result of the confirmation by the temporary end point address comparing means, if the temporary end point address is not the address of the own node, the temporary end address As a result of checking by the sending means for transferring the received packet according to the end address and the temporary end address comparing means, if the temporary end address is the address of the own node, the temporary end address is deleted from the received packet. And a temporary end address deletion transfer means for transferring the packet in accordance with the destination address of the packet. 4. The IP packet transfer control system according to claim 1, wherein said source node monitors reception of a control packet notifying congestion, and said control node receives said control packet. Means for identifying a node of a packet transmission source, and determining the temporary end node for the node, transferring a packet to a congested node of a transmission source of the control packet to the temporary end node,
The temporary end node includes a storage unit that stores a packet whose own address is assigned as the temporary end address, a unit that monitors a congested state in a node that is a source of the control packet, and a result of the monitoring, When the congestion state at the transmission source node of the control packet has subsided, means for reading the packet stored in the storage means, and deleting the temporary end point address from the read packet and shaping according to the destination address of the packet. And a means for transferring the packet. 5. An IP packet transfer control method for a network system, comprising: transferring a packet in an IP network from a source node to a destination node via a plurality of relay nodes according to an IP address. Identifying the one of the relay nodes as the temporary end node, assigning the address of the relay node specified as the temporary end node (temporary end address) to the received packet, and transferring the packet according to the temporary end address; At the relay node, the temporary end address assigned to the packet received from the source node is compared with its own address, and if they are different, the received packet is transferred according to the assigned temporary end address, The temporary end address is deleted from the received packet, and the Transferring in accordance with the destination address of the packet. 6. The IP packet transfer control method according to claim 5, wherein said source node specifies said temporary end node from said temporary end node to said end node according to a destination address of said packet. An IP having a step of identifying a node satisfying a condition that a packet to be transferred can be reached without passing through a route when the packet to be transferred is transferred from the own node to the temporary end node according to the temporary end address. Packet transfer control method. 7. The IP packet transfer control method according to claim 5, wherein each of the source node, the relay node, and the temporary end node and the end node has a predetermined location in a header of a received packet. A first step of determining whether or not the temporary end address is set, and if the result of the determination in the first step is that the temporary end address is not set, the temporary A second step of determining whether or not the addition of the end point address is necessary based on a predetermined condition; and determining the addition of the temporary end point address to the received packet by the second step. The packet transferred from the end point node to the end point node according to the destination address of the packet is the temporary end point address. Therefore, a third step of specifying a node that satisfies the condition that the node can be reached without passing through a route that is transferred from the own node to the temporary end node as the temporary end node, and a third step of specifying the node in the third step A fourth step of adding the address of the temporary end node to the received packet and transferring the received packet according to the address; and if the result of the determination in the first step is that the temporary end address is set, A fifth step of confirming whether or not the temporary end address is the address of the own node; and, as a result of the confirmation in the fifth step, if the temporary end address is not the address of the own node, the temporary A sixth step of transferring the received packet according to the destination address, and as a result of the confirmation in the fifth step, the temporary destination address is If the address, and deletes the temporary destination address from the packet thus received, IP packet transfer control method characterized in that it comprises a seventh step of transferring according to the destination address of the packet. 8. The IP packet transfer control method according to claim 5, wherein the source node monitors reception of a control packet notifying of congestion, and monitors the received control packet. Identifying a source node, determining the temporary end node for the node, and forwarding a packet to the congested node of the source of the control packet to the temporary end node; And storing in the storage device a packet having its own address assigned as the temporary end address, monitoring the quenching of the congestion state in the source node of the control packet, and as a result of the monitoring, determining the source of the control packet. If the congestion state at the node of the node has subsided, the packet stored in the buffer means is read out, and the temporary end address is deleted. Transferring according to the destination address of the IP packet. 9. A processing procedure program of an IP packet transfer control method for a network system in which a packet is transferred in an IP network according to an IP address from a source node to a destination node via a plurality of relay nodes by respective computer processes. A recording medium for recording, wherein a program for causing a computer to execute each step of the IP packet transfer control method according to any one of claims 5 to 8 is recorded. .