KR20020016198A - Distribution Performance System constructed by fabricat ing Several User terminals into single system and Construction method thereof - Google Patents

Abstract

PURPOSE: A distributed processing system and a building method thereof are provided to reduce the processing capacity of a server and to relieve the bottleneck of an exclusive line by making multi clients distributively process data. CONSTITUTION: The system comprises a main server(200) and multi individual clients composing a client system(210). The client system forms a tree structure consisting of a first stage client(10) directly connecting to the main server, a second stage client connecting to the first stage client as a server, and a third stage client connecting to the second stage client as a server. An N stage client(30) is connected to one of n-1 stage clients(20) as a server and n+1 clients(40) as subordinating systems. If no subordinating client is registered or all subordinating clients are disconnected, respective clients are placed in the waiting state. The data is distributively processed by the clients through a network.

Description

Distribution Performance System constructed by fabricating Several User terminals into single system and Construction method

The present invention relates to a distributed processing system through sharing among a plurality of user terminal devices connected through a network, a method of forming the system, and a use of the system. More specifically, the present invention relates to a plurality of clients connected to a main server on a network under a specific form. It is to make the processing capacity of large capacity by organizing organically.

Recently, a variety of information has been provided or obtained through virtual cyberspace such as the Internet or communication. The Internet is a collection of computer networks using various protocols such as the TCP / IP protocol, and is generally operated by a client / server structure using the HTML (Hypertext) language. The client and server communicate with each other using the functionality provided by HTTP (Hyperte xt transfer protocol), and the client can access the server via a URL.

Many technologies have been registered at home and abroad regarding devices and methods for providing information through the Internet. Among them, Korean Patent No. 201259 (registered on March 13, 1999) is an apparatus that automatically delivers newspapers using the Internet, and is composed of newspaper reservation means, newspaper delivery means, and newspaper output means, and Korean Patent No. 217378 ( 1999.6.4) relates to providing breaking news by using e-mail.

The service basically consists of a client / server system connected to the network. The client / server structure is a model for constructing a service system by linking computers distributed in a network. A server managing various resources receives a request from a client, processes resources, and sends the result to the client. Do the work of sending. For example, the WWW server manages documents written in HTML, and retrieves and sends them according to requests from clients (browsers such as Explorer and Netscape).

Recently, information delivery service using a client / server system using the Internet from a network is prevalent. One example is the webmail service (for example, the trade names 'Netien,' 'Hanmail', 'Dae', etc.), and the service provider has a large number of members and gives them a free (or paid) email address. E-mail delivery service. Another example is a web service (trade name 'c hanceit.co.kr') that acts as a substitute for various prizes.

These are basically implemented in a client / server structure. When a server performing an information delivery service receives a request from a client (e-mail transmission, prize entry, and various message transmission), the server grasps the recipient address of the corresponding information and receives each recipient. To send to.

Referring to Figure 1 will be briefly described a conventional information delivery service system. In FIG. 1, the server 100 is a server of an operator that performs a webmail service or an application agent service. The user 10 connected to the server 100 may be one of the clients registered as a member in the operator server, but it does not necessarily need to be a member, and the subject 20 according to the work request (for example, in case of mail transmission). The mail recipient may or may not be registered as a member of the server.

For example, if a client, among the numerous clients registered as a member of the server 100 running the webmail service, makes a request to the server to send an e-mail to 1,000 other clients B or a non-member client C, the client A Is the user client 10, B client or C is the subject (20-1, 20-2, ... 20-1000). When the user client 10 requests a specific job to the server 100 in this way, the server 100 processes the job internally and delivers the job to each target person 20.

In the case of the conventional information delivery service, if the number of users or work subjects increases, the amount of work to be processed in the server may be very large. Therefore, the processing capacity of the server must be increased and the load on the server increases. As a result, the cost of operating the web information delivery service increases and the failure rate of the server increases.

In addition, when the number of subjects is very large, as in the case of FIG. 1, the bottleneck on the transmission line may become very severe because the server must transmit information to all subjects at once, thereby greatly reducing the processing speed of the network.

Accordingly, an object of the present invention is to provide a distributed processing system for reducing the processing capacity of a server and alleviating bottlenecks of transmission lines by allowing a server to distribute a plurality of clients and to distribute the information to a connected client without transferring all information. To provide a method.

1 is a diagram of a conventional information transmission system.

2 is a distributed processing system diagram of the present invention.

Figure 3 is a block diagram of a distributed processing module constituting the system of the present invention.

4 is a flowchart showing a method for forming a distributed processing system of the present invention.

5 is one embodiment of a distributed processing system of the present invention.

6 is a flowchart showing an e-mail transmission method as a first embodiment of the present invention.

7 is a flowchart of a search method using a search engine on the web as a second embodiment of the present invention.

8 is a flowchart of a direct reverse auction method as a third embodiment of the present invention without passing through a broker at all;

9 is a flowchart of a prize information delivery method implemented on the web as a fourth embodiment of the present invention;

The present invention provides a distributed processing system through sharing between a plurality of user terminal devices connected through a network, and includes a main server and a client system connected to the main server through a network and configured as m sub-systems of the main server.

The client system is configured such that any client included in the nth stage constituting the client system is any one of the clients included in the upper n-1 stage through the network as a server, and is included in the nth stage. Any client that is configured to be a server of 0 or at least one client included in the lower level n + 1 through the network is configured to enable distributed processing of the main server and the amount of data that each client must process. Characterized in that it is a distributed processing system between user terminal devices through a network having a.

However, m is an integer variable that depends on the total number of clients currently connected, and infinite expansion is possible. When m≥2, n is any integer satisfying 1≤n≤m-1 (where n = 1 If the step 0 refers to the main server).

For example, when m = 10, n has a value of 1 to 9, and if n is 9, the lower client n + 1 becomes 10, which becomes a final client.

In addition, the network is generally used to mean the Internet.

According to the above system, the user system (hereinafter referred to as a client system which accesses the main server and flows into the distributed processing system) is referred to as a user system. In this case, the server is included in the system using any one of the clients connected to the main server and the lower client having the server as the server.

The main server is a main server that operates and manages a distributed processing system and provides an arbitrary service. The first stage client means a client directly connected to the main server. A client using the first client as a server is referred to as a second client, and when the relationship is generalized, an n-th client is referred to as an n-th client, which is the above step. It can be defined to mean that the n + 1 level client is a subsystem. In this case, the n-th level client may be in a state in which at least one sub-client is connected at a specific point in time of the system, or may not have a sub system yet in a connection standby state (when the lower n + 1th level client is 0). However, it is a potential server in terms of having a subsystem in the future.

As described above, any one client belonging to the n-th step can have its own subsystem through the network, and the subsystem includes at least one client. However, as described above, the client corresponding to the last stage of the tree at a specific point in time does not have a sub-system yet, so it is natural to be in a waiting state.

Therefore, according to the above definition, the client constituting the distributed processing system of the present invention functions as both a client of an upper server and a server of a lower client.

The dispersion processing system of the present invention will be described in more detail with reference to FIG. 2. The system of the present invention is composed of a main server 200 and a plurality of individual clients constituting the client system 210 (arrows are connected to a communication network and constitute only a part for convenience of description). The client system 210 includes a first stage client 10 that is directly connected to a main server, a second stage client (not shown) as a server, and a second stage client. It can be seen that the tree structure extended to the n + 1 stage is taken as a method of forming a third stage client (not shown) serving as a server. In the figure, the n-th stage client 30 shows that one of the n-th stage clients 20 is a server and the lower n + 1th stage client 40 is a subsystem. As described above, each client constituting the distributed processing system of the present invention is required to be connected to any one server of a higher level, and is placed in any one of a standby state which does not have a lower system. The standby state may occur when the lower client is not yet registered or when all registered lower clients disconnect.

In FIG. 2, it is preferable that each step client reduces the server load by limiting the number of lower clients that can be accommodated. If it does not have a certain limit as described above, it is possible to have an infinite number of sub-clients, bringing a limit to the processing capacity as a server, such as paralyzed the system is not desirable.

However, the maximum number of lower clients per main server and client may be selected and determined according to the data processing capability held by each terminal device and the required characteristics of the main server operator, and the present invention does not need to limit it to a specific value.

The client of each step constituting the client system in the system of the present invention has IP information of all servers that are its higher level in the tree structure (for example, if the client is K in FIG. 5, G, F, A and IP information of the main server), and at least IP information and information on the number of lower clients are required for the lower clients managed by the main server. In particular, in order to minimize the load on the server, it is preferable that the main server and each client store only the information on the client immediately below the minimum level with respect to the lower client.

In addition, the distributed processing system is not fixed in the form of a tree taking only a specific structure. In other words, the structure of the tree changes over time due to the inflow or outflow of the user system into the system. That is, it is not preferable to leave the leaked out of the system of a specific client constituting the client system because it blocks the connection with the upper server of the client subsystem. In this case, the subsystem in which the connection is blocked is required to be connected to the upper server in some way, and for this purpose, the lower client is required to have IP information of the upper server. Therefore, when the server of the previous stage leaks from the system, the client realizes a new tree type by the procedure of accessing the new upper server and reassigning its location using the upper IP information.

In practice, the distributed processing system may be configured through a distributed processing module in a specific program type. The distributed processing module is commonly installed in the client constituting the system, and the installation method is a method of automatically downloading and installing a new user system at any time when connecting to the main server, and manually downloading and accessing the main server. All of them are suitable for implementation of this invention.

The distributed processing module is provided with a function required to configure the system by entering the distributed processing system of the tree structure after the user system is connected to the main server and has the function to automatically update the data update in the main server. It is necessary to do it.

3 is a block diagram of the distributed processing module 300. The distributed processing module includes automatic updating means 201 for automatically updating when data held by an upper server is updated, and associated with positioning of a server in a distributed system. And a system holding means (203) for holding a distributed processing system in a tree structure.

Therefore, when the main server has a data update including a module which is configured to give a specific function to the distributed processing module such as an e-mail sending / receiving function, add a specific event, or process a specific task, Through the automatic update means, it is possible to download the data from the upper level server of the previous stage and update its system with the upper level server.

In addition, the system holding means 303 for maintaining the distributed processing system of the tree structure is a processing capacity check means for periodically checking the work processing capacity of the lower client 10 and the new user system introduced into the system from the upper server in advance System disposition means 20 for arranging in the sub system by the predetermined distribution rule, and system repositioning means for relocating the sub system of the specific client to any one sub system of the upper server according to a predetermined distribution rule when the specific client is leaked. 30, valid information management means 40 for managing information valid for system maintenance, including IP information of upper and lower clients, and information about the number of lower systems, and a distributed processing system for predetermined business processes. It is configured to include a business processing means (50).

The processing capacity checking means 10 is a means for checking the work processing capacity that can be processed by the lower client. The processing capacity checking is performed by quantifying information including CPU speed, memory capacity, current workload, etc. according to a predetermined evaluation criteria. do. The evaluation result is provided as information for forming a distributed processing system by determining the load and work capacity of the system.

The system arranging means 20 is a means for arranging a new user system flowing into the system from the upper server in the lower system by a predetermined distribution rule. This is the case when a registration of a new user system is assigned to a client whose system has the best processing capability.

The system relocating means 30 is a means for redistributing the sub-system of the specific client to any one sub-system of the upper server according to a predetermined distribution rule when the specific client is leaked. If the processing system leaks, that is, terminates the connection with its parent server, a new server / client structure is established using the IP of the parent server that is known to maintain the system. Preferably, the sub-system of the client that terminates the connection as described above is connected to the upper server of the client that terminated the connection and relocated. In this case, as described above, the distribution rule includes a method of allocating a client having the best job processing capability to the server as described above.

If the host server is leaked, it is desirable to allow the subsystem to send certain signals and data to the host system upstream to periodically verify its existence in order to relocate the system. The specific signal is for checking whether the server of the above step exists in the current system, which can be confirmed through a response from the upper server. If there is no response from the server even after a predetermined allocation time elapses after the signal transmission, the upper server is regarded as leaked from the system and the lower client is relocated in the system.

Preferably, the data transmitted from the lower client to the upper server includes its own IP, the number of subclients managed by the client, the total number of subsystems including the self, and the current workload to processing capacity ratio of the total subsystems including the self. Contains information about The information on the work processing capability of the subsystem is data necessary for evenly distributed arrangement of clients constituting the system when the new client enters the system.

The job processing means 50 is a means for executing a job to be actually processed, and in the case of mail transmission, it includes a program related to a job processing object such as a mail transmission / reception program, a prize site application program, a reverse auction operation program, a search program, and the like. .

In the above, each means constituting the distributed processing module may be implemented by a program. Since the hardware configuration of the central processing unit, the data communication structure, the storage device including the RAM, etc. for executing the module is well known to those skilled in the art in a known configuration, a detailed description thereof will not be provided.

The present invention also includes a method for forming a distributed processing system between user terminal devices through a network.

The distributed processing system forming method of the present invention

(1) the user system connecting to the main server,

(2) registering the user system as a client for any one selected from a main server or a client constituting the client system;

(3) wherein the registered client is provided to a server of another user system.

4 shows a method for forming a distributed treatment system of the present invention.

Step 401 is a step in which the user system accesses the main server through the network. In the case of a user newly connecting to the main server, the user may access the home page of the main server using a normal web browser such as Internet Explorer, Netscape, etc. In the case of an existing user (a user of a system in which a distributed processing module is installed), the user can directly access the main server by clicking a predetermined icon separately provided on the monitor without using a web browser. The predetermined icon is to be automatically connected to the server by executing the distributed processing module by a click action, which can be realized through a program using a predetermined language.

Step 403 relates to the positioning process of the user system into the distributed processing system after the connection to the main server, wherein the user system configures a subclient of the main server or a subclient system connected to the main server. One of each client is a lower client which is the parent server.

The process may be performed by actually connecting a user system to the main server and then accessing the IP address of the first level client, or by allowing the main server to pass the IP address of the user system to one of the first level clients. It is feasible. The first step client connected in this way registers itself if it can register itself. Otherwise, the location designation is made by passing the connection to the next step client.

The location method in the system of the client requesting the connection does not need to be a specific method. Basically, the client system is maintained based on the main server and the tree structure of the client system is maintained, that is, if it is possible to take a configuration having a server at a higher level and at least one lower client at a lower level based on a specific client. It is suitable for forming the system of the invention.

As described above, a user system located as a lower client is provided to be a server of another user system to which the client subsequently connects, thereby enabling infinite expansion of the tree structure.

In the case of the new user who has not yet installed the distributed processing module in step 403, the distributed processing module which is programmed to distribute the data between the server and the client from the main server at any time after access by the user system is downloaded. Step 407 is required. The distributed processing module is installed in the user system by step 409 and simultaneously executed.

On the other hand, if the user system has already downloaded and installed the distributed processing module through the connection with the main server in step 405, in the subsequent connection, the connection icon displayed in the window window is clicked without executing a separate web browser. You can connect to the main server just by doing it. Such a connection method is obvious to those skilled in the art depending on how to configure a program, and thus will not be described in detail because it does not constitute the gist of the present invention.

Step 411 is a process of performing a necessary task in a distributed processing system of a new user system or an existing registered user system, which will be described in detail with reference to the following embodiments.

The method of forming the distributed processing system of the present invention as described above will be described in detail through an embodiment of the present invention.

In a preferred embodiment of the above-mentioned methods, in positioning a user system, a method of registering the client with a server having a minimum number of lower clients or a server having a highest processing capability by checking a work processing capability of each lower client. The method of registering to this etc. is mentioned. The above methods have the great advantage of distributing the load of the server evenly into the system.

Fig. 5 is a system configuration diagram showing the above embodiment, and is a configuration example in which the method for registering the user system is applied to a server having a minimum number of lower clients.

In the above embodiment, the maximum number of subordinate clients accessible to each server is preliminarily designated as three. Therefore, if three lower clients are saturated in a specific server, no further connection can be made. Therefore, as shown in FIG. 5, the clients A, B, C, D, E, F, and G are already saturated and can no longer connect to the lower client. Client H may have three subclients since there is currently no subsystem, and Client I may have two clients.

After a request for access from a user's system, the main server first checks to see if its system is saturated. If it is not saturated, it registers itself as its subordinate client first, and if it is already saturated at the time of connection request, it includes transferring the status of the server to the subordinate client system.

In FIG. 5, the first stage client of the main server is saturated with A, B, and C. The number displayed on the right side of the letter indicates the total number of lower clients. The number information of the lower clients is checked by the distributed processing module as described above. In the drawings, however, sub-systems of B, C, D, and E are omitted for convenience of description.

A now has 30 clients, including itself, which is B, and B is 35, C is 37. The main server is notified to the parent server, and the main server has the minimum number of subsystems based on this information. Subsequently, the user system that enters is located as a subsystem of A, and through the same process, A is located as a subsystem of F having the least of the subsystems, and F is again the least subsystem (subclient). Number: 0 is positioned first with H constituting.

In the above process, the location designation generally includes a method of transmitting an IP of a user system requesting a connection to a client (H in the above example), which becomes a server among clients of a lower system.

The positioning method provides the advantage of stabilizing the system evenly distributed without any client bias to any one system. However, such a configuration is not limited to the scope of the present invention as an embodiment, and those skilled in the art can easily change the degree of changing the arrangement of the distributed processing system through the above configuration, so the arrangement of the present invention. Is not limited to the above cases.

In the distributed processing system according to the present invention, it is possible to secure a large-scale system by only one act of accessing a main server on a network, and in the case of a job requiring a large amount of data processing, When dealing with the server, it is possible to solve the weighting problem of the server load.

Hereinafter, a work processing process using the distributed processing system of the present invention will be described with reference to specific embodiments.

6 shows an e-mail transmission method as a first embodiment of the present invention.

E-mail transmission method using a network as a first embodiment of the present invention

(1) requesting mail transmission to the main server (601);

(2) Determining whether the main server can directly process (602) direct processing if the entire amount of processing is possible (603), otherwise the server constitutes a distributed processing system except for the amount of transfer that can be directly processed Transmitting the mail contents and the recipient list to the lower client to request transmission (604);

(3) a step 605 of transmitting the result to the user system after each client requesting the transmission transmits the mail. The processing result is output through the screen display unit and provided to the user (606).

In the mail transmission method using the distributed processing system, when a mail transmission is requested to any one client in the client system, the mail transmission method is preferentially transmitted through its own and its own sub-systems, and the rest of the mail contents are required to be transmitted. And the recipient list to the parent client of the corresponding client to be distributed with the lower client.

For example, in the case of sending 10,000 mails through the F client in FIG. 5, assuming that 1,000 packets can be sent by the F client, the first 1,000 mails are processed by the user, and the remaining throughput is a subsystem. In the case of calculating the processing capacity of G, H, and I, assuming that 1,000 processing can be done per client, G can handle 4,000 copies including the subsystems J, K, and L, and H is 1,000, and G is With a sub-system, K, 2,000 transactions are possible, allowing the F-client to send a total of 7,000 e-mails to the subsystem.

In addition, 2,000 unprocessed packets are delivered to A, the upper server of F, and A receives 1,000 packets for itself, and the rest is distributed to D and C, the lower clients of A, to request processing.

In addition, in the mail transmission method using the present invention, it is preferable to include a script means so as to enable a predetermined editing function in a message requiring transmission. The script means may be implemented programmatically through the above-described distributed processing module, and is mainly necessary for diversification of a message according to a name or a member of a receiver.

7 shows a search method using a search engine on the web as a second embodiment of the present invention.

The retrieval method as the second embodiment of the present invention

A step 701 in which the user system is located in the distributed processing system by connecting to the main server;

Inputting a search term to search through the search box (703);

Requesting the search to each client constituting the distributed processing system using different search engines allowed by the web (705);

And transmitting and outputting the search result information of the lower client to the user system (707).

In addition, when the user system is any one of the clients in the client system, it is preferable to preferentially search through its own and its sub-systems, and deliver the rest to the upper server of the user system to distribute the processing.

For example, in FIG. 5, when requesting a search execution for 'distributed processing system' scattered on the web through the F client, G, H, and I clients constituting the F and F subsystems are different search engines that are allowed on the web. Search for the search word. The search engines currently include Yahoo, Alta Vista, Simmani, Hanmir, Naver, Lycos (trade name), etc. All of them are designated by one client for each search.

However, when the number of subsystems is insufficient, a process of requesting a higher level server to perform a distributed processing is included as in the first embodiment of the electronic mail transmission. After the search is completed, all the results are delivered to the F client who requested the search, and the duplicated results are processed as one and displayed on the monitor under the familiar GUI environment.

FIG. 8 shows a direct reverse auction method as a third embodiment of the present invention without passing through a broker at all.

The reverse auction method as the third embodiment of the present invention

A step 801 in which the user system is connected to the main server and positioned in the distributed processing system,

Step 803 of a user inputting a condition of an item desired to purchase through a product search window and requesting a search;

In step 805, the user system individually requests a different company registered in the server for each client constituting the distributed processing system and requests product information meeting the conditions.

The lower client transmits a plurality of product information obtained from the registered companies to the user system and outputs the outputs (807, 809).

For example, if the e-commerce system access method of the previously negotiated companies is already delivered to each client constituting the distributed processing system, the user enters the search conditions through the product search window and clicks the search button. The search will be assigned to one subclient per vendor. For example, if there are 10 registrars for an item, the subsystem would require 10 clients, and each client would connect to 10 vendors 1: 1 to perform a search. For each of the above designated rooms, the client receives the quotation from the corresponding company, arranges it in a predetermined form, and delivers it to the requesting client. Then, the client requesting the work consolidates the quotations received from the lower client into one and displays the check box and the order button to the user.

When a user checks a desired product and presses a purchase button, a purchase request is transmitted to each corresponding product company, and when the user pays the purchase amount, each company sends the product.

Therefore, according to the reverse auction method to which the system of the present invention is applied, it can be very convenient in that it can be performed without having to go through the existing mediation means at all.

9 shows a prize information delivery method implemented on the web as a fourth embodiment of the present invention.

Prize information delivery method as the fourth embodiment of the present invention

A main server generating input data including user information and request information required by various prize events currently underway on the web and storing the input data in a storage device (901);

The input data is automatically transmitted when each client constituting the distributed processing system connects to an upper server (902);

Each client comprises a step (903) of processing the user information transmitted to the giveaway event management server.

In the above, the user information is information for confirming the identity of each user in the prize event management server, and the system-specific items such as age, name, etc., are processed by the script function in each client receiving a variable when the server creates them. To be added.

Given that an average of the day's sweepstakes is carried out at 20 sites, an operator of the main server operating the distributed processing system of the present invention can operate the sweepstakes information delivery service as one of the special benefits to the members. .

The main server examines the event of the day of the day in advance and programs the input data required by the server, that is, the user information and the request information (for example, the answer to the problem and the response to the questionnaire). The lower clients receive the input data from the upper server through the automatic update means to complete the variable-processed user information and transmit the same to the server. The user can receive 20 prizes just by being connected to the distributed processing system. You will get the effect of participating in the event.

The result of the prize winning shall be notified to the visitor of the main server in advance on the predetermined screen before the scheduled release date of each prize management server so that the user can check the result of the prize directly on the relevant site on the date of the prize. Do.

In addition, the present invention provides a method for a user to directly post his own advertisement using the distributed processing system.

The advertisement method includes assigning a predetermined advertisement window in an execution step of the distributed processing module to display an advertisement desired by the advertisement window. Accordingly, since the lower client using the user system as the server downloads the distributed processing module including the advertisement from the upper server, it is possible to transmit its own advertisement to the lower system. If your subsystem consists of 10,000 clients, you have the opportunity to advertise to 10,000 people.

The advertisement window constitutes a distributed processing module and may be programmed to be installed at any point on the screen when the distributed processing module is being executed. In addition, the advertisement window may be executed only when a specific program included in the distributed processing module is executed (for example, when the distributed processing module includes an entertainment function). The advertisement window may be implemented by a program through a predetermined language, and the present invention also includes using an existing banner advertisement format.

According to the present invention, the cost saving effect of the server expansion is very large for the operator who operates the server, the bottleneck occurring in the server-side transmission line can be alleviated, and for the user who requires a large amount of work processing capacity. It is possible to easily perform work by allowing distributed processing through the network.

Claims (18)

In the distributed processing system by sharing between a plurality of user terminal devices connected via a network, The distributed system includes a main server, a client system connected to the main server through a network and configured as m sub-systems of the main server. The client system is configured such that any client included in the nth stage constituting the client system is any one of the clients included in the upper n-1 stage through the network as a server, and is included in the nth stage. Any client that is configured to be a server of 0 or at least one client included in the lower level n + 1 through the network is configured to enable distributed processing of the main server and the amount of data that each client must process. Distributed processing system between user terminal devices through a network having a. However, m is an integer variable that depends on the total number of clients currently connected, and infinite expansion is possible. When m≥2, n is any integer satisfying 1≤n≤m-1 (where n = 1 Step 0 refers to the main server. The client of claim 1, wherein the client of each step of configuring the client system includes IP information of all servers that are higher levels of the client system and at least IP information and information about the number of lower clients for the lower client managed by the client. Distributed processing system between user terminal devices through a network, characterized in that. The distributed processing system of claim 1, wherein the network is the Internet. In the method of forming a distributed processing system by sharing between a plurality of user terminal devices connected via a network, Connecting the user system to the main server, Registering the user system as a client for any one selected from a client constituting the main server or the client system; And providing the registered client to a server of another user system. 5. The method of claim 4, wherein when the client is a new client connecting to the main server, the distributed processing module is downloaded and installed so as to distribute data between the server and the client from the main server at any time after the connection. Forming a distributed processing system between the user terminal device over a network, characterized in that it further comprises a step. The method according to claim 5, wherein the distributed processing module includes automatic updating means for automatically updating when the data held by the upper server is updated; And a system holding means for maintaining a distributed processing system of a tree structure, wherein the system holding means comprises: processing capacity checking means for periodically checking a work processing capability of a lower client; System disposing means for disposing a new client flowing into the system from an upper server in a lower system by a predetermined distribution rule; System relocating means for relocating the sub-system of the specific client to any one sub-system of the upper server according to a predetermined distribution rule when the specific client is leaked; Valid information management means for managing information valid for system maintenance, including IP information of upper and lower clients, and information on the number of subsystems; A method of forming a distributed processing system among user terminal devices through a network, comprising: a processing unit for enabling a predetermined processing of a business by using a distributed processing system. 5. The method of claim 4, wherein when any one of the client systems constituting the distributed processing system has a lower client connection with the upper server, the lower client connects to the upper server of the upper server which terminated the connection. Forming a distributed processing system between user terminal devices through a network, the method comprising the steps of: reassigning its own location. 5. The method of claim 4, wherein when the number of clients that can be registered in each server is determined, the registration of the new client is preferentially assigned to the server having the minimum number of lower clients, and the distributed processing among the user terminal devices through the network. System formation method. 5. The method of claim 4, wherein when the number of clients that can be registered in each server is determined, the registration of the new client is preferentially assigned to the server having the maximum processing capacity of the lower client, and distributed among the user terminal devices through the network. Process system formation method. In the e-mail transmission method using a network, Requesting mail transmission to the main server; The transfer amount that can be directly processed by the main server is performed by the server. The rest of the method comprises the steps of requesting transmission by delivering a mail content and a list of recipients to be sent to a lower client constituting the distributed processing system of claim 1; And transmitting the result to the user system after each client requesting the transmission transmits the mail. In the mail transmission method using the distributed processing system of claim 1, when a mail transmission is requested to any one client in the client system, the mail transmission method is preferentially transmitted through its own and its own sub-systems, and the rest requires transmission. And transmitting the mail contents and the recipient list to the upper client of the client to distribute the mail contents and the recipient list together with the lower client. 12. The method of claim 10 or 11, wherein each client includes scripting means for enabling a predetermined editing function on a message requiring transmission. In a search method of a specific website using a network, A user system connected to the main server and positioned in the distributed processing system of claim 1, Enter the search terms you want to search through the search box, Requesting the search to each client constituting the distributed processing system of claim 1 using different search engines allowed by the web; And transmitting the search result information of the lower client to a user system and outputting the search result information. 15. The method of claim 13, wherein when the user system is any client in the client system, the user system preferentially searches through its own and its sub-systems, and the remainder is distributed to the upper server of the user system for distributed processing. Search term search method using a network. In the reverse auction method using a network, A user system connected to the main server and positioned in the distributed processing system of claim 1, The user enters the conditions of the item to be purchased through the product search box and requests a search; The user system for each client constituting the distributed processing system of claim 1 to specify different companies registered in the server to request product information that meets the conditions, And transmitting, by the lower client, the plurality of product information obtained from the registered companies to the user system and outputting the product information to the user system. 16. The method of claim 15, wherein when the user system is any client in the client system, the user system preferentially searches through its own and its sub-systems, and the rest is distributed to the upper server of the user system for distributed processing. Direct auction method using a network using a network. In the prize information delivery method using a network, A main server creating input data including user information and request information required by various prize events currently underway on the web and storing them in a storage device; The input data is automatically transmitted when each client constituting the distributed processing system of claim 1 is connected to an upper server; And each of the clients transmits the input data to the prize event management server by completing the user information. In the advertising method using a network, Allocating a predetermined advertisement window in the execution step of the distributed processing module to display the desired advertisement on the advertisement window; Accessing a lower client using the user system as a server; Downloading and installing the distributed processing module including the advertisement from the lower client by a user system; And outputting the advertisement screen through a predetermined advertisement window while the lower client executes the distributed processing module, and the distributed processing module is delivered to a lower client to be connected thereafter.