JPH1084385A - Communication connection device and communication connection method with multiple servers - Google Patents

Abstract

(57) [Summary] [Problem] To provide a system in which a copy of provided information is simultaneously stored in a plurality of different servers and transferred to a requesting client via an IP communication network without being aware of the plurality of servers. The side distributes the communication load so that information can be provided efficiently. SOLUTION: When a client 13 sends a request for access to information to a GW 14, the GW 14 transmits an I
The DA / SA field of the P header is manipulated and encapsulated, and an information access request is transferred to one server selected from the server group 12. The server decapsulates the transmitted IP packet and transfers the stored provision information to the client 13.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a communication system and method for transferring information stored in a server to a client requesting information acquisition via an Internet Protocol (hereinafter, referred to as IP) communication network. In particular, the present invention relates to a communication connection device and a communication connection method with a plurality of servers, in which copies of information are simultaneously stored in a plurality of different servers and a communication load on a client is distributed.

[0002]

2. Description of the Related Art In a service that provides a large amount of information from a server that stores information to a client that browses the information, when a plurality of clients access the server at the same time and retrieve the stored information, the server must The response to the client may be reduced due to overload. In order to prevent this, a method of accumulating copies of information in a plurality of servers and distributing the load is adopted.

[0003]

In the above-mentioned prior art, in order to access information, a client obtains all different IP addresses of a plurality of servers that store a copy of the information, and the client obtains all of them according to the judgment of the client. You need to select one of them to access. Therefore, the server selected by the client may be in an overloaded state, and in that case, the transfer speed of information from the server is reduced. SUMMARY OF THE INVENTION An object of the present invention is to improve such a problem, to enable a client to access stored information without being aware of the existence of a plurality of servers, and to suppress overload of a server or congestion of a communication path. It is an object of the present invention to provide a communication connection device and a communication connection method with a plurality of servers, which are suitable for maintaining a transfer rate of information from a server. Another object of the present invention is to suppress overload of a gateway that performs communication between a server and a client or congestion of a communication path. Another object of the present invention is to realize a configuration in which it is not necessary to add a function to another node in an IP communication network including a client, except for an encapsulation function of a gateway and a decapsulation function of a server.

[0004]

In order to achieve the above object, according to the present invention, a correspondence table (14-3 in FIG. 2) and a connection log file (14 in FIG. 2) are provided in a gateway device (GW14 in FIG. 1). -5) and an IP packet encapsulation function (14-4 in FIG. 2), and the server device (servers 12-1,..., 12-n in FIG. 1) has an encapsulation release function (FIG. 3).
It is characterized in that the communication connection process is performed according to the following procedures (1) to (6). (1) There are a plurality of servers storing copies of information, and a client requesting access to information sends an access request to the GW. (2) The GW executes the connection destination server selection program based on a preset server selection algorithm so that the access load does not concentrate on a specific server. Select (3) The GW converts the IP packet of the access request to the information sent from the client into an IP packet addressed to the selected server.
The packet is encapsulated in the header and transmitted to the IP communication network. (4) The server that has received the encapsulated IP packet of the access request releases the encapsulation of the IP packet and sets the Source Address field (hereinafter, S) of the IP header.
A field), the IP address of the client requesting access to the information is determined. (5) The server transfers information to the client. At that time, the IP field of the server is stored in the SA field of the IP packet.
Write the IP address of the GW instead of the address. (6) Since the IP address of the GW is written in the SA field of the IP packet sent from the server to the client, the IP packet sent by the client to communicate with the server is sent to the GW. The GW stores the server with which the client is communicating in the storage device, and transfers the IP packet sent from the client to the server to the server by referring to the server.

[0005]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In the present invention, a gateway (GW) distributes access from a client to a plurality of servers that store a copy of information based on a preset server selection algorithm. In the GW, since the communication from the client to the server is relayed by encapsulation of the IP packet, the client only needs to know the IP address of the GW in order to access the information, and accumulates a copy of the information. I for multiple servers
The P address does not need to be considered. In the communication from the server to the client, the server uses the SA of the IP packet.
Since the IP address of the GW is written in the field, the IP address of the server is completely hidden from the client.

An embodiment of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a diagram schematically showing a network configuration and a communication connection method according to an embodiment of the present invention. Reference numeral 11 denotes an IP configured by a plurality of IP networks and routers.
A communication network 12 includes n physically different (n ≧ 1) servers 12-1 for storing a copy of certain information X.
, 12-n, 13 is a client that requests the server to access information, 14 is a client that receives an access request to information X from the client 13, and selects one server from the server group 12. And a gateway (hereinafter referred to as GW) for transferring an access request to the server.

FIG. 2 is a block diagram of the internal functions of the GW 14, and 14-6 is a network interface for transmitting and receiving IP packets to and from the IP communication network 11. 14-1 is I
A P packet reading program which reads an access request for information sent from the client 13 and identifies information requested by the client 13 or an IP for confirming a response from the client 13 to information transfer from the server. It reads a packet or reads a failure notification message sent from a router on the route when a failure occurs on the route during the information transfer from the server to the client 13. A connection destination selection program 14-2 is a server list that stores the server group 12 storing the information X requested by the client 13 and the correspondence between the information and the server storing the information. One server is selected from the server group 12 with reference to the file 14-3 and based on a preset connection destination selection algorithm. 14-4 is I
A P encapsulation device, which is a Destinat of an IP packet for an access request to information X sent from a client 13
ion Address field (hereinafter referred to as DA field)
Then, an IP header in which the IP address of the server selected by the connection destination selection program 14-2 is written is added, and the IP address is encapsulated. A connection log file 14-5 stores the correspondence between the client and the server communicating through the GW.

FIG. 3 is a block diagram showing the internal functions of an arbitrary server 12-i included in the server group 12. All the servers included in the server group 12 have the same configuration. 12-i-4 is a network interface for transmitting and receiving IP packets to and from the IP communication network 11, and 12-i-1 is a device for decapsulating the encapsulated IP packets sent from the GW 14. 12-i-2 is an IP packet reading program which reads the decapsulated IP packet and recognizes the requesting client and the requested information, or a response from the client which sent the information. When a confirmation is received, or when a failure occurs on the route during the information transfer from the server to the client 13, a failure notification message sent from a router on the route is read. 12-i-3 is an information storage device for storing information.

Next, a communication connection method according to this embodiment will be described. The IP address of each network node is
Client 13 is IPc, GW 14 is IPg, server 1
,..., 12-n are IPs ₁ ,. First, the client 13 sends an IP packet for an access request to the information X to the GW 14. That is, IPc is written in the SA field and IPg is written in the DA field and transmitted. The router in the IP communication network 11 is DA
Routing is performed based on the IPg in the field, and the IP packet arrives at the GW 14. When the IP packet reading program 14-1 in the GW 14 reads the IP packet of the access request from the client 13 and determines that the client 13 is requesting the information X, the connection destination selection program 14-2 is started, Reference is made to the server list file 14-3 storing the correspondence between the information and the server storing the information, and the server 12-
12-i (1 ≦ i ≦ n) is selected from 1,..., 12-n. Then, the IP encapsulation device 14-4 adds SA = IP to the IP packet received from the client 13.
c, DA = I of IPsi (IP address of server 12-i)
An IP packet with SA = IPc and DA = IPg is encapsulated with a P header and sent to the IP communication network 11. I
The router in the P communication network 11 performs routing based on the IPsi of the DA field in the IP header of the encapsulated IP packet, and the encapsulated IP packet reaches the server 12-i. The IP decapsulation device 12-i-1 in the server 12-i releases the encapsulation, and the IP packet reading program 12-i-2 converts SA = IPc and the contents of the datagram in the IP header of the IP packet. By reading the information, it recognizes that the client 13 has requested access to the information X, and transfers the information X stored in the information storage device 12-i-3 to the client 13.
At this time, SA = IPg and DA = IPc are written and transmitted in the IP header of the IP packet storing the information X, so that the client 13 receiving the IP packet is as if the information X was transferred from the GW 14. Looks like.

An IP packet for confirming a response from the client 13 and an IP packet storing a failure notification message sent from the router when there is any failure in the IP communication network are transferred to the GW 14. The GW 14 that has received the IP encapsulation device 14 refers to the connection log file 14-5, determines that the server communicating with the client 13 (IP address IPc) is the server 12-i (IP address IPsi), and In -4, the IP packets of SA = IPg and DA = IPsi
The P-header is attached and transferred to the server 12-i. When another client sends an access request to the information X to the GW 14, the connection destination selection program 14-2 sets the server 1
2-i, a server 12-j (1 ≦ j ≦ n) is selected, and the IP encapsulation device writes the IP address IPsj of the server 12-j in the DA field using the IP header of the access request. Is encapsulated and transferred to the server 12-j, thereby decentralizing access to the information X from the client. The following examples are given as algorithms for the connection destination selection program 14-2 to change the server to be selected. (1) The GW 14 monitors the operation status of the server 12-i, and when a failure occurs in the server 12-i, or when the access amount to the server 12-i exceeds a preset threshold. In this case, another server 12-j that is operating normally and whose access amount does not exceed the threshold is selected as a connection destination instead of the server 12-i. (2) When transferring the access of a specific number of clients set in advance, for example, 12-1, 12-2,...
12-n, 12-1, and so on. (3) At a specific time interval set in advance, for example, 1
2-1, 12-2,..., 12-n, 12-1,.

[0011]

According to the present invention, the client does not need to be aware that there are a plurality of servers that store the copy of the information, and sends a request for access to the information to the GW.
W transfers the access request to the information to one server selected from the plurality of servers. The communication performed by the GW is only reception of an access request from the client, transfer of the access request to the server, and relay of a response confirmation from the client in information transfer communication from the server to the client. The load of these communications is light. The possibility that GW overload or congestion in the communication path occurs due to these communications is low. On the other hand, the communication for transferring a large amount of information from the server to the client is heavy, but since this communication is distributed to a plurality of servers that store the copy of the information, the server is overloaded or the communication path is congested Is less likely than when all clients access a single server. In addition, in order to use the above-described means, a newly required function is an I
Only the P packet encapsulation function and the decapsulation function in the server need not be added to other nodes in the IP communication network including the client.

[Brief description of the drawings]

FIG. 1 is a diagram showing a connection procedure when a client issues an access request to a gateway and receives information transfer from a server in one embodiment of the present invention.

FIG. 2 shows a gateway (G) according to an embodiment of the present invention.
It is a block diagram of W).

FIG. 3 is a configuration diagram of a server according to an embodiment of the present invention.

[Explanation of symbols]

11: IP communication network, 12: server group, 12-1 to 12-
n: server, 13: client, 14: gateway (GW).

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Hideyo Murakami, Inventor 3-19-2 Nishi Shinjuku, Shinjuku-ku, Tokyo Japan Telegraph and Telephone Corporation (72) Isami Nakajima 3-2-1, Nishi-Shinjuku, Shinjuku-ku, Tokyo No. Japan Telegraph and Telephone Corporation

Claims (3)

[Claims] 1. An information processing apparatus, which is installed in an Internet protocol communication network composed of a plurality of networks using the Internet protocol and stores information stored by a plurality of physically different server devices connected via the communication network. A gateway device for receiving an access request from a client to a server device and connecting to the server device, a correspondence table between information requested by the client and the server device storing the information, and Storage means for storing a connection log file which records the correspondence with the server device in communication; and a capsule for encapsulating the internet protocol packet by adding a predetermined internet protocol header to the received internet protocol packet Means and client to server When an Internet Protocol packet indicating a request for access to the device is received, one server device is selected from the server device group with reference to the correspondence table, and the address of the server device is set as the destination by the encapsulating means. Means for encapsulating the internet protocol packet with the internet protocol header written in the address field and sending the encapsulated internet protocol packet to the server, and confirming a response from the client to information transfer communication from the server to the client Receiving the Internet protocol packet, the server refers to the connection log file, recognizes the server that is communicating with the client, and uses the encapsulating means to set the Internet protocol address of the server to the destination.・
Control means for encapsulating the Internet Protocol packet with an Internet Protocol header written in an address field and transmitting the encapsulated packet to a server device. 2. A client installed in an Internet protocol communication network constituted by a plurality of networks using the Internet protocol, having a data storage means, and transmitted via a gateway device connected to the communication network. Server device for performing transfer communication in response to a request for access to self-storage information from a gateway device, a release unit for releasing encapsulation of an encapsulated Internet Protocol packet sent from the gateway device, When the received Internet Protocol packet is received from the gateway device, the release means decapsulates the Internet Protocol packet, reads the Internet Protocol address of the client, and sends the Internet protocol packet to the client. Server apparatus characterized by comprising means for controlling to write the Internet protocol address of the gateway device to the source address field in the header of the Col-packet. 3. A plurality of physically different server devices each comprising a plurality of networks using an Internet protocol and storing copies of information are distributed and arranged in each network, and information is stored between the server device and the client. The gateway device has a correspondence table between information requested by the client and a server device storing the information, and a correspondence between the client and the server device communicating with the client. Storage means for storing the recorded connection log file; and encapsulation means for adding a predetermined internet protocol header to the received internet protocol packet to encapsulate the internet protocol packet. Is encapsulated from the gateway device What is claimed is: 1. A communication connection method for an Internet protocol communication network system comprising a release means for releasing encapsulation of an Internet protocol packet, comprising: transmitting an Internet protocol packet of an access request from a client to a gateway device; One of the plurality of server devices that stores a copy of the server device is selected, and the Internet protocol address of the selected server device is written in the destination address field. Internet protocol of access request
The access request from the client is transferred to the selected server device by encapsulating the packet and sending it out to the Internet protocol communication network. When the server device receives the encapsulated Internet protocol packet from the gateway device, The de-encapsulation means de-encapsulates the Internet protocol packet, reads the Internet protocol address of the client, and stores the Internet protocol packet in the source address field in the header of the Internet protocol packet to be transmitted to the client. A communication connection method comprising writing an Internet protocol address and performing transfer communication to a client which has issued an access request.