JP2003289320A - Address determination method and node in network - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To determine an address with a simple scheme without making a processing scheme remarkably different by the presence/absence of a server when determining the address required for initial setting of a network in accordance with the scale of the network and the allowable volume of traffic. <P>SOLUTION: In a predetermined designated mode, a novel node multicasts an initialized packet requesting responses from all nodes or requesting a response only from the server into a subnet and when there is a response packet from the server as an initialized response packet, the novel node determines a given low-order layer address contained therein as a low-order layer address of the present node but when there is no response from the server, error stop occurs according to the designated mode or based upon the information of existing low-order layer addresses contained in the initialized response packets returned from one or a plurality of started nodes, unused one value is selected and determined as a low-order layer address of the present node. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to an address determination method and node in a network.

[0002]

2. Description of the Related Art Generally, when deciding an address necessary for initial setting of a network, an address deciding method is selected and decided according to the size of the network and an allowable traffic amount. Conventionally, it is described by RFC in an IP network. As shown by the DHCP server and AutoIP, etc., there is a problem in that the processing method of address determination differs greatly depending on the presence or absence of a server, and the implementation is complicated. Also, if there is a server in the subnet, if the server itself is abnormal or communication with the server is temporarily disconnected, the existence of the server is ignored and the address is distributed when operating in the distributed mode. In addition, there is a problem that it takes time to determine an address.

[0003]

SUMMARY OF THE INVENTION According to the present invention, the address required for initial setting of the network is set to the network scale,
It is an object of the present invention to provide an address determination technique capable of determining an address by a simple method, in which the processing method does not greatly differ depending on the presence or absence of a server when determining according to the traffic allowance.

Further, according to the present invention, when an abnormality occurs in the server itself existing in the subnet or when communication with the server is temporarily interrupted, the existence of the server is ignored and the address is distributed. I solved the conventional problem that
It is an object of the present invention to provide an address determination technique capable of shortening the time until the address determination.

[0005]

According to a first aspect of the present invention, there is provided an address determination method in a network, wherein a new node is a response mode request for all nodes or a server according to a designation mode predetermined by a hardware switch or a software setting method. Only one of the initialization packets of the response request is multicast in the subnet, and if there is a response packet from the server as the initialization response packet, the new node assigns the assigned lower layer address as the lower layer address of the own node. Determined, if there is no response from the server, error stop according to the specified mode, or use based on the information of the existing lower layer address contained in the initialization response packet returned from each of one or more activated nodes Select one value that has not been set and determine it as the lower layer address of the local node It is an feature.

According to a second aspect of the present invention, in the address determining method in the network according to the first aspect, the initialization packet includes at least a destination address, a self address, a packet type, a self hardware address, and a temporary lower layer address. A multicast address within a subnet is designated as a destination address, and the designated mode is identified by information of a packet type or additional bits.

According to a third aspect of the present invention, in the address determination method in the network according to the first or second aspect, the initialization response packet includes a destination address, a self address, a packet type, a self hardware address, an assigned lower layer address or a self address. It is characterized in that it includes a lower layer address and a code indicating whether or not it is a response from the server, and the source address of the initialization packet is designated as the destination address.

According to a fourth aspect of the present invention, in the address determination method in the network according to the first to third aspects, the network is a Master / Slave system network, and the activated node has received the initialization packet from the new node. However, the initialization response packet for the initialization packet is transmitted when a fixed predetermined time has elapsed after the initialization packet was received.

According to a fifth aspect of the present invention, in the address determination method in the network according to the first to third aspects, the network is a CSMA / CD system network, and the activated node has received the initialization packet from the new node. However, the initialization response packet for the initialization packet is transmitted when a predetermined time proportional to the lower layer address value held by the node has elapsed after the initialization packet was received.

According to a sixth aspect of the invention, in the address determination method in the network of the fifth aspect, the minimum value of the settable range of the address value is set as the reserved address in the lower layer address value of the server node. It is a feature.

According to a seventh aspect of the present invention, in the address determination method for a network according to the fourth aspect, a reserved value of a lower layer address is determined for a slave node to be a server in a subnet, and the Master node acquires the address. Before operation, a database including the hardware address, lower layer address, and slave management number of each slave node to be managed is generated, and a new slave is created.
When the initialization packet is received from the node, the slave corresponding to the lower layer address defined as the reserved value
Transfer to the slave node with the management number, and
The initialization response packet from the e-node is returned to the slave node that is the transmission source of the initialization packet.

According to the invention of claim 8, in the address determination method in the network of claims 1 to 3, each node is a combination element of a life and death Bit and a hardware address of each of the other nodes obtained from the contents of the received packet. It holds a management table whose value is cleared and recreated when the self-node is restarted, the life-and-death Bit is added to the initialization response packet, and it is addressed to the new node as lower layer address information of the started node. It is characterized by transmitting to.

According to a ninth aspect of the present invention, when a node newly enters the network, an initialization packet including identification data of either all-node response request or server-only response request is subnetted based on the operation mode of the network. When an initialization response packet is received from another node, when the initialization response is sent from another node, the initializing packet transmitting means for multicasting within the assigned lower layer address included in the response response from the server in the subnet. If it is determined to be the lower layer address of the own node, and there is no initialization response packet from the server, and only the ones returned from each of the one or more activated general nodes, each initialization response Select one unused value based on the information of the existing lower layer address contained in the packet Self-lower layer address determining means for determining the lower layer address of the above, and when receiving an initialization packet from another node, returns an initialization response packet including the self lower layer address to the node that is the source of the initialization packet. And an initialization response packet transmitting means for performing the same.

That is, according to the node of the invention of claim 9,
If the node is a new node, an initialization packet for either all-node response request or server-only response request is multicast in the subnet according to a predetermined designated mode, and a response packet from the server is sent as an initialization response packet. If there is, the assigned lower layer address included here is determined as the lower layer address of the own node, and if there is no response from the server, error stop according to the designated mode, or 1
Or, based on the information of the existing lower layer address included in the initialization response packet returned from each of the plurality of activated nodes, one unused value is selected and determined as the lower layer address of the own node. . If the node of the present invention is an activated node in the subnet, it receives an initialization packet multicast from a new node and returns an initialization response packet including its own lower layer address.

As a result, when the node of the present invention newly enters a subnet, if a network exists in the subnet, the lower layer address is directly acquired from the server, and in the network in which the server does not exist in the subnet. If so, an address value that is not included in the lower layer address returned from another node can be autonomously determined as its own lower layer address.

[0016]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail below with reference to the drawings.

In a network system, there is a method (server method) in which a server uniquely existing in a subnet distributes and manages the addresses of all nodes in the subnet, and a serverless method in which each node autonomously manages its own node in the subnet. There is a method of determining an address (distributed method).

The address determination method in the network of the present embodiment is divided into a method in which a new node determines its own lower layer address in each of these server method and distributed method, and a method in which the time until the lower layer address is determined is made as small as possible. It has characteristics. The lower layer address means the lower layer L3 in the protocol stack shown in FIG.
It is an address that should be held, and is a globally unique address in the subnet that is assigned to all the nodes that have been started in the subnet. Further, the upper address is uniquely determined from the lower layer address, but this does not matter here. Further, before the operation of acquiring the lower layer address, the acquisition of the unique IP address in the subnet is completed, and the acquisition operation of the lower layer address is triggered by the completion of the acquisition. Furthermore, the hardware address is a value that is set in advance to a unique value for each transmission medium at the time of manufacture.

Each node 100 including the server has a hardware configuration shown in FIG. 4, and includes a transmission medium for digital signal transmission and a modulation / demodulation circuit 1, and address distribution, management, and other predetermined operations by executing a program. A micro computer 2 that executes control, an EPROM 3 as a non-volatile memory, and a ROM 4 and a RAM 5
Composed of. The address determination described below is executed by the software program.

Packets carrying data between nodes in the network are of the structure shown in FIGS. Here, “self” means a node (including a server node) that transmits each packet, and the element of each packet is 1
It is expressed in bits in bytes.

When the new node rises, it outputs the initialization packet A from the lower layer L3 shown in FIG. The initialization packet A is shown in FIG. 5, and includes a destination address A1, a self address A2, a packet type A3, a response type A4, a self hardware address A5, and a temporary lower layer address A6. Of these, the destination address A1 is
IP multicast destination within subnet, self address A
For 2, the self IP address is written. Then, depending on whether the packet type A3 is written and the response type A4 is not written, or the packet type A3 and the response type A4 are both written, it is possible to identify whether the request is an all-node response request or a server only response request. ing. Whether the new node makes a response request to all nodes or only the server makes a response request is preset in the new node by a hardware switch, a software setting method, or the like. In other words, each node can be selected by the user or the installer when requesting the response of all nodes or only the server when entering the network.

The node receiving the initialization packet A
Depending on whether or not the response type of the initialization packet A is written, it is identified whether the response is only the server or the all-node response. If the response type is server-only response, FIG.
The initialization response packet C shown in is returned. On the other hand, if there is no write in the response type of the received initialization packet A and therefore it is an all-node response, the nodes other than the server return the initialization response packet B shown in FIG. The encrypted response packet C is returned.

The initialization response packet B shown in FIG. 6 includes a destination address B1, a self address B2, and a packet type B.
3, response type B4, own hardware address B5,
It is composed of the self lower layer address B6. The initialization response packet C shown in FIG. 7 includes a destination address C1, a self address C2, a packet type C3, a response type C4, a self hardware address C5, a self lower layer address C6, and an assigned lower layer address C7.

In these initialization response packets B and C,
The destination addresses B1 and C1 are the IP addresses of the transmission source nodes of the initialization packet A, and the self addresses B2 and C2 are the self I
The P address, the packet type B3, C3 or the packet type and the response type B4, C4 identify whether the packet received by the code written therein is returned from the server or the general node. Further, it indicates that the response is the initialization packet A. These initialization response packets B and C are their own hardware addresses B5 and C5 and their own lower layer address B.
6, including C6. Furthermore, if the initialization response packet is from the server, the initialization response packet C
Includes the lower layer address C7 given to the new node.

The packet D shown in FIG. 8 is an acceptance response returned from the server to the lower layer address from the server, and includes a destination address D1, a self address D2, a packet type D3, a response type D4, and a self hardware address. D5 and self lower layer address D6. The destination address D1 of this packet D is the IP address of the server.

The packet E in FIG. 9 is multicast in the subnet after the provisional lower layer address is determined. The destination address E1 of the multicast, the self address E2 indicating the sender, the packet type E3, and the self hardware. The address E4 and the temporary lower layer address E5 are included. If the node receiving the provisional lower layer address E5 by this packet E is equal to its own lower layer address,
The duplicate notification packet F shown in FIG.
The address is transmitted with the destination address F1. This packet F is composed of a destination address F1, a self address F2, a packet type F3, a self hardware address F4, and a self lower layer address F5.

The specific operation of the present invention will be described below for each of the method of determining some lower layer address (A mode) and the method of reliably determining the lower layer address by the server (SR mode) regardless of the presence or absence of a server. explain. The designation of the A mode and the SR mode can be switched by a hardware switch, a software setting method, or the like, but it is desirable to determine these based on the network operation rule. That is, it is determined based on the network operation rules such as the network scale, the traffic occupancy allowance, and the presence / absence of an administrator. Generally, it is recommended to select the SR mode when the scale is large, there is no traffic occupancy allowance, and there is an administrator. Under other conditions, the A mode is recommended.

FIG. 11 is a sequence diagram of a lower layer address determination procedure in the SR mode executed by each node. S
In R mode, the lower layer address is always distributed by the activated server. If there is no communication between the started server and the new node, or if the server is down, there is no response from the server and this mode ends with an error. In this case, administrator intervention is required to recover from the error.

The new node multicasts the initialization packet A shown in FIG. 5 (S1). On the other hand, the initialization response packet C shown in FIG. 7 is returned only from the activated server (S2). The new node that has received this initialization response packet C receives the packet D shown in FIG.
And the lower layer address is determined.

FIG. 12 is a sequence diagram of a lower layer address determination procedure in A mode executed by each node. A
In the mode, if the new node transmits the initialization packet A by multicast (S11), all the activated nodes (including the server node) in the subnet return the initialization response packet B or C (in the case of the server). Comes (S12).

Therefore, even if there is no response from the server (in this case, there is no initialization response packet C), the new node is not the self lower layer address B6 included in the received initialization response packet B from each node. The temporary lower layer address is determined for the new node from the lower layer addresses of the above, and the packet E having the destination address of each activated node that has returned the initialization response packet B including this value is transmitted by multicast (S13). .

When it is determined that one of the activated nodes overlaps the self-lower layer address with respect to the packet E by the multicast, the address duplication is notified by the packet F (S14). When the new node receives this address duplication notification, it excludes the above-mentioned reply address group and the address for which the address duplication notification has been made, redetermines the temporary lower layer address, performs multicast in S13, and returns the duplication notification. wait.

If the duplicate notification is not returned even after a certain time has passed from the multicast in S13, the temporary lower layer address is determined as the main address (S16).

On the other hand, when the new node transmits the initialization packet A by multicast in S11, the initialization response packet C is sent from the activated server in the subnet.
If there is a response (S12 ′), the new node gives the assigned lower layer address C7 included in this initialization response packet C.
Is determined as the self lower layer address (S17). In this case, the lower layer address can be immediately determined without transmitting / receiving the packets B, E, F. This allows
Even in the A mode, it is possible to quickly distribute the lower layer address.

In the network address determination method of this embodiment, a method of determining some lower layer address in any case (A mode) and a method of surely determining the lower layer address by the server (SR mode) are used.
The new node can be used properly by preparing two types of packet types for multicasting within the subnet (changing some code numbers of the initialization packet A). Then, the activated node only responds according to the initialization packet from this new node, and the new node selects the S
The lower layer address can be determined by the R mode or the A mode.

The address determination method in the SR mode is a stable method with a small amount of traffic because only one new node needs to exchange data with the server, and there is no failure even in a network composed of many nodes. On the other hand, the address determination method in the A mode has a high traffic occupancy rate because responses from all the nodes are returned for one new node, and is not suitable for a network composed of many nodes. However, according to the present embodiment, it is possible to selectively and easily change the address determination method by only partially changing the code number of the initialization packet A according to the scale of the network.

When the server exists, it is natural that the new node wants to receive the lower layer address distribution as soon as possible. In order to solve this, Ma shown in FIGS. 1 and 13 is used.
The case of communication in the star / Slave system and the communication in the CSMA / CD system node shown in FIGS. 2 and 14 will be described.

Master / Sla shown in FIGS. 1 and 13.
In the ve method, a plurality of slaves and one master communicate with each other in a time slot, and the master is each Sl.
The ave is sequentially accessed to exchange packets with a specific Slave. However, the Master that received the packet from the Slave sends the packet to which Sl
If it is known in advance whether to transfer to ve, the packet can be passed to the target Slave in the shortest time.

Since only one node should be a server in the subnet, one reserved value is set as the lower layer address for the server node. And the new node is S
In the case of a love node, the Master node that receives the packet A transmitted by this is immediately transferred to the slave node having the lower layer address reservation value (or the Master node).
(If the r-node itself has a reserved value, transfer is unnecessary), and as a response, the initialization response packet C can be immediately returned from the server node having the lower-layer address reserved value to the new node as the transmission source.

The Master node is a Slav under management.
Before acquiring the lower layer address, the management tables 10A and 10B including at least the IP address, the hardware address, the lower layer address, and the slave management number of the e-node are shown in FIG.
5, it can be stored in the RAM 5 or the EPROM 3 and distributed to the slave node corresponding to the reserved value of the lower layer address. Here, a reserved value is written in advance as a lower layer address value in the server node, and the Master reads this together with the IP address and the hardware address, and the management table 10 in FIG.
Reflect on B. In the management table 10B of FIG. 15, the lower layer address is written as 0x00. As a result, even if the Master itself is a server node, the initialization response packet C can be immediately returned to the transmission source.

Next, the CSMA / CD shown in FIG. 2 and FIG.
A case of communication at the method node will be described. In this method, there is no particular concept of Master / Slave, and the node that wants to transmit performs carrier sense immediately before transmission,
In this method, if the carrier is not on the transmission medium, it is sent to any node. In such a network, the time from the reception of the initialization packet A to the transmission of the initialization response packet B or C is set to the lower layer address value * fixed time, so that a plurality of nodes send out packets at the same time. Sometimes packet collisions are avoided. At the same time, the lower layer address value of the server node is set to a minimum value smaller than other normal nodes as a reserved value, so that the server node can initialize the initialization response packet C earliest.
To be able to send. That is, the new node can obtain the lower layer address value in the shortest time.

Therefore, the server determines the lower layer address value as the reserved value to the minimum value of the lower layer address settable range, and particularly in the Master / Slave system, the Master immediately accesses the reserved value of the lower layer address value. By transmitting the initialization packet A and receiving the initialization response packet C, and in the CSMA / CD system node, each node has a communication rule in which a packet response is made by the product of the lower layer address value and the fixed time. When the server exists, the node can receive the lower layer address distribution in a short time.

Further, the initialization response packet B shown in FIG.
Can notify more reliable used address by adding the following information. The other node life and death management table 11 shown in FIG. 16 is a data table held by each node. This data table is
t, each node hardware address, and timer element. Life and death Bit is RAM5 in the order of lower layer address number
For example, the first memory address indicates the lower layer address = 0, the next address indicates the lower layer address = 1, and if the corresponding memory address exists, the corresponding address Bit = 1, for example. , And if it does not exist, Bit = 0 is defined. The hardware address of the node corresponding to each lower layer address is also held together. Also, the corresponding live / dead bit of the source hardware address of the packet that each node received last has a countdown timer, and the countdown timer setting value is reset to the initial value each time the packet corresponding to the source hardware address is received. As a result of the timer being decremented, when the countdown timer value becomes equal to or less than a predetermined value, it is determined that the bit does not exist, and Bit = 0 is set.

Data table 1 managed in this way
It is possible to give more reliable information of the used lower layer address value by adding the life / death Bit of 1 to the initialization response packet B as shown in FIG. 17 and notifying the new node of the used lower layer address value. it can. The data in the data table 11 is cleared and recreated when the node is restarted.

[0045]

As described above, according to the present invention, the method of determining the lower layer address required for the initial setting of the network can be changed according to the scale of the network by changing the partial code number of the initialization packet to the designated mode. Only by setting, it is possible to deal with both the server system and the distributed system selectively and easily, and the time required for address determination can be shortened.

[Brief description of drawings]

FIG. 1 is a topology of a Master / Slave type network.

FIG. 2 is a topology of the CSMA / CD system.

FIG. 3 is a protocol stack diagram of a node used in one embodiment of the present invention.

FIG. 4 is a block diagram showing a configuration of a node used in one embodiment of the present invention.

FIG. 5 is a configuration diagram of a packet A in the above embodiment.

FIG. 6 is a configuration diagram of a packet B in the above embodiment.

FIG. 7 is a configuration diagram of a packet C in the above embodiment.

FIG. 8 is a configuration diagram of a packet D in the above embodiment.

FIG. 9 is a configuration diagram of a packet E in the above embodiment.

FIG. 10 is a configuration diagram of a packet F in the above embodiment.

FIG. 11 is a sequence diagram of lower layer address determination in the server system (SR mode) according to the above embodiment.

FIG. 12 is a sequence diagram of lower layer address determination in the distributed mode (A mode) according to the above embodiment.

FIG. 13 is a time chart of packet transmission / reception in the Master / Slave system.

FIG. 14 is a time chart of CSMA / CD system packet transmission / reception.

FIG. 15 is an explanatory diagram of a slave management table held by the Master node.

FIG. 16 is an explanatory diagram of another-node life-and-death management table held by each node used in the third embodiment of the invention.

FIG. 17 is a configuration diagram of a packet B used in the third embodiment of the invention.

[Explanation of symbols]

1 microcomputer 2 Transmission media / modulation / demodulation circuit 3 EPROM 4 ROM 5 RAM 10A, 10B Slave management table 11 Other node life and death management table 100 nodes L1 media processing layer L2 UDP / IP layer L3 lower layer L4 upper layer

Claims (9)

[Claims] 1. A new node multicasts an initialization packet of either an all-node response request or a server-only response request into a subnet in a predetermined designated mode, and the new node is sent from the server as an initialization response packet. If there is a response packet of the above, the assigned lower layer address included here is determined as the lower layer address of the own node, and if there is no response from the server, error stop according to the specified mode or return from one or more activated nodes respectively Address determination in a network characterized by selecting one unused value and determining it as the lower layer address of its own node based on the information of the existing lower layer address contained in the received initialization response packet Method. 2. The initialization packet includes at least a destination address, a self address, a packet type, a self hardware address, and a temporary lower layer address, and the destination address specifies a multicast destination within a subnet,
The address determination method in a network according to claim 1, wherein the designated mode is identified by information of a packet type or additional bits. 3. The initialization response packet includes a destination address, a self address, a packet type, a self hardware address, an assigned lower layer address or a self lower layer address, and a code indicating whether or not a response from the server, and the destination address. The address determination method in the network according to claim 1 or 2, wherein a source address of the initialization packet is designated as the address. 4. The network is Master / Sl.
In the ave system network, when the activated node that has received the initialization packet from the new node sends an initialization response packet for the initialization packet to a fixed predetermined time after the initialization packet is received. The address determining method in the network according to claim 1, wherein the address is transmitted. 5. The network is a CSMA / CD system network, and an activated node which has received the initialization packet from the new node receives an initialization response packet for the initialization packet and receives the initialization packet. 4. The address determination method in the network according to claim 1, further comprising the step of transmitting when a predetermined time proportional to the lower layer address value held by the node has elapsed. 6. The lower layer address value of the server node includes:
The address determination method in a network according to claim 5, wherein the minimum value of the settable range of the address value is set as the reserved address. 7. Sla to be a server in a subnet
The reserved value of the lower layer address is determined in advance for the ve node, and the Master node, before the address acquisition operation, the hardware address of each Slave node to be managed,
A database including a lower layer address and a slave management number is generated, and when the initialization packet is received from a new slave node, the database is transferred to the slave node having the slave management number corresponding to the lower layer address defined as a reserved value, The address determination method in a network according to claim 4, wherein the initialization response packet from the slave node is returned to the slave node that is the transmission source of the initialization packet. 8. A management in which each node is composed of a combination of a life and death Bit and a hardware address of each of the other nodes obtained from the contents of the received packet, and the value is cleared and recreated when the self node is restarted. The table is held, the life-and-death Bit is added to the initialization response packet, and the lower-layer address information of the activated node is transmitted to the new node. Method for address determination in the Internet. 9. An initialization packet for multicasting, within a subnet, an initialization packet containing identification data of either all-node response request or server-only response request based on the operation mode of the network when newly entering the network. When receiving the initialization response packet from the transmitting means and the other node,
If the initialization response packet includes a response packet from the server in the subnet, the assigned lower layer address included in the response packet is determined as the lower layer address of the own node, and the initialization response packet does not include the one from the server, or 1 or If only the general nodes that have been started have been returned, select one unused value based on the existing lower layer address information contained in each initialization response packet. A self lower layer address determining means for determining the lower layer address of the own node, and an initialization response packet including the self lower layer address when an initialization packet is received from another node, the node which is the source of the initialization packet. An initialization response packet transmitting means for returning to the node.