JPH10187524A - Information management method using computer network and information management system using computer network - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To provide means for quickly and surely performing processing for making the information owned by each computer the same in consideration of the state of a communication line connecting the computers. A lower-level computer for which registration of new information has been applied selects one higher-level computer, and transfers all information owned by the higher-level computer. Ask. The lower calculator is
Updates the own information with the transferred information. If there is no overlap between the updated information and the new information, the information is newly registered as own information, and the registered information is registered. To the selected host computer. On the other hand, the selected host computer responds to the information transfer request
All the information owned by itself is transferred to the lower-level computer, and the registered information transferred from the lower-level computer is newly set as information owned by itself.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to information management means using a computer network having a plurality of computers, and more particularly, to providing a plurality of higher-level computers and taking into account the state of a communication line connecting the computers. The present invention relates to a technology for quickly registering new information or the like, thereby quickly performing a process of making information owned by each computer the same.

[0002]

2. Description of the Related Art Conventionally, in a software construction environment,
Attention is being paid to design methods and development methods using object-oriented techniques, and agent-oriented development methods and the like are also proposed ("Object-oriented analysis and design", Information Processing Society of Japan Vol. 35, PP392-401). , May 1994 "
etc). Among them, the present inventors have developed an integrated object-oriented technology that implements a complete object-oriented approach at all stages from the system concept to coding (refer to the document "Realization of open database application construction support system by object-oriented", Etc.) and apply the technology to the development of database applications such as database search and statistical processing. In this environment, the user can freely create necessary part objects without any coding, and can easily build an application by combining the part objects.

[0003] In this application construction method using a combination of objects (software parts), it is difficult to share a software part with another construction environment because the construction environment is mostly on a single computer.

In response to this problem, the present inventors have proposed a distributed software development environment in which software components are developed and registered on a plurality of computers connected to a network, and applications are developed by combining the distributed components. (Reference: "Realization of software construction technology using multi-agents", IPSJ Object-Oriented '96 Symposium, Jul, 1996). In this environment, distributed object technology is used as a basic technology, and by utilizing the position transparency and scale transparency of objects by this technology, software components on multiple computers can be developed by all developments connected to the network. It can be used uniformly from the environment. As a result, sharing of software components over the entire network is realized.

However, in order to realize this development environment on the basis of a wide area network such as the Internet, there are some problems to be solved with respect to this distributed object technology.

Hereinafter, the problems of the current wide area network will be clarified, and the problems of the existing distributed object technology resulting therefrom will be described as follows. First,
"Issues of wide area network represented by the Internet" are compared with regional networks, and then "existing distributed object technology"
The problems that occur when used in a wide area network are described.

At present, a wide area network (WideArea Network: W) represented by the Internet and an intranet
AN) are spread all over the world, and tens of millions of computers are connected. For this reason, it is possible to use and exchange information from computers all over the world. Although it is a wide area network with such characteristics,
There are the following problems when compared with a local area network (Local Area Network: LAN).

First, there is a problem of "ununiformity of communication speed due to unification of network connection methods". Normal, L
In the case of AN, the connection method is unified, but in the case of WAN, Ethernet (Ethernet, 10BaseT, 100BaseTX, et
From c.), there are various connection methods such as FDDI and ISDN, and an appropriate connection method is used depending on the situation. Therefore, the communication speed varies from a number (Gbit / sec) to a number (Kbit / sec) and is not uniform.

Next, there is a problem of "instability of the effective communication speed due to network sharing". For LAN,
Network use is limited. However,
In a WAN, since the same network is shared by a plurality of people for various purposes, its effective communication speed dynamically changes according to the usage status of other users. Further, depending on the situation, the communication speed may be zero, that is, the connection may be disconnected.

As described above, when constructing a distributed software construction environment for a WAN, a distributed object technology taking the above problems into consideration is indispensable. Next, problems of the existing distributed object technology when used in a wide area network will be described.

At present, various distributed object technologies have been proposed and implemented in various products (refer to the document "Open Software Fundation" DCE RPC Internals
and Date Structures Revision 1.0 "Aug, 1993: Open Software Foundation DECA
RPC Internals and Data Structure Revision, "Sun Microsysytems Comp
uter Company "The Java Language Environment," A Whit
e Paper, Oct, 1995: Sun Microsystems Computer Company Java Language Environment "). In addition, as a unified standard of these technologies, CORBA (document "Object Managemen
t Groupe "The Common Object Request Broker: Archite
cture andSpecification Revision 2.0, "OMG, July, 1995
: Object Management Group, The Common Object Request Broker: Architecture and Specification Revision, etc.) has been established and has become the de facto standard. In distributed object technology represented by CORBA,
"Name service", "Life cycle service",
Functions such as "event service" are defined.
This realizes functions related to the distributed environment (location transparency, access transparency, failure transparency, and the like).

[0012] Among them, the name service is a service for registering and retrieving reference information and the like for an object in order to realize location transparency of the object on the network. In order to realize a software construction environment for a wide area network, it is necessary to use this name service function to manage software components (objects) distributed on the network. Therefore, there is a need for the name service to support wide area networks.

Normally, this name service has a hierarchical structure similar to the directory structure of a file system in order to consider the correspondence to a network. For example, C
ORBA2.0 adopts a method of constructing a naming graph in which a plurality of name servers are hierarchically arranged as shown in FIG. In this case, the uniqueness of the name of the object is guaranteed for each namespace (Space), and as a whole, "space name +
The combined name "name" guarantees the uniqueness of the entire network. Information about objects in the entire network, the name server NS ₀ of the uppermost hierarchy will be integrated management, for each object name unique for each space, does not perform the management of uniqueness here. With this hierarchical structure,
Manage information about objects across the network.

Now, a conventional name service will be specifically described with reference to FIGS. The Figure 2 shows the configuration of a conventional system, the system includes a single upper side of the computer X (corresponding to the name server NS _0), a plurality of lower side of the machine (name in the namespace server NS _1a ,
Corresponding to NS _1b, etc.) 1a, 1b, ..., and a 1n, lower computer 1a, 1b, ..., 1n, respectively, communication lines 2a, 2b, ..., can communicate with the computer X through 2n It is connected to the.

Each of the lower computers 1a, 1b,
, 1n are management units 10a, 10b that perform various processes,
, 10n and memory units 20a, 20b,..., 20n for storing information with identifiers.
The computer X has a management unit 15 that performs various types of processing, and a memory unit 25 that stores information with an identifier.

The operation of this system will be described with reference to FIG. Now, each computer has information of identifiers a and b, and the identifier c is transmitted from another computer (not shown) to the computer 1a.
It is assumed that new information marked with is transmitted and this new information (corresponding to the information of the newly created object) is registered together with the identifier c.

First, in step S300, the management unit 10a of the computer 1a sends a request from another computer (not shown).
It is assumed that an application for registering new information with the identifier c has been received. At this time, in step S305, the management unit 10a suspends reception of another process.

Next, in step S310, the management unit 10a performs a process of detecting whether or not any of the identifiers in the memory unit 20a overlaps with the identifier of the new information. If so, in step S335, the application for registration is rejected, a message to the effect that a duplicate is present is sent to the other computer, and the process is terminated (END). Proceeding to S315, new information with the identifier c is registered in the memory unit 20a.

Further, in step S320, the management unit 10a stores all information in the memory unit 20a in the host computer X.
The management unit 15 stores only information (in this case, information with the identifier c) other than the information already registered in the memory unit 25 among the transferred information. It is newly stored in the unit 25. In step S325, the management unit 15 of the upper computer X transfers all the information of the memory unit 25 to the lower computers 1b,.
1n. At this time, the other lower computers 1b,..., 1n other than the computer 1a, among the transferred information, information other than the information already registered in the memory unit (in this case, the identifier c is added). Information) is newly stored in the memory unit.

Then, in step S330, the management unit 10a resumes the interrupted reception of another process and ends the process (END). By such processing, the newly registered information is stored in the memory units of all the computers 1a, 1b,..., 1n.

On the other hand, when referring to information stored in the memory unit, a command specifying an identifier is transmitted from another computer (not shown) to the management unit of an arbitrary lower-level computer. This is performed by obtaining information stored in a memory unit of a lower computer.

As described above, the conventional information management method using a computer network has been performed by constructing a hierarchical system configuration having one upper computer and a plurality of lower computers.

[0023]

In the hierarchical structure shown in FIG. 1, information on objects in the entire network can be managed.
Because it is targeted at AN, when targeting WAM,
There were the following problems.

First, there is a problem of processing delay due to concentration of processing load. Top level server NS constituting the naming graph ₀ is present only one, but to manage the entire network information, if number of computers that are connected to a WAN is large, even the number of objects to be registered on the computer very More. Therefore, the information processing amount to focus the server NS ₀ becomes enormous, the delay of processing occurs.
Therefore, management and reference of information are hindered.

Second, there is a problem that responsiveness of information management is deteriorated. This is the traditional name service, W
Uneven communication speed between servers, which is a problem of AN,
It does not consider the instability of the effective communication speed. In other words, depending on the communication state between the server that manages each space and the higher-level name server, the responsiveness of information exchange is deteriorated, and there is a problem in referring to object information. The problem described here exists in the conventional distributed object technology. for that reason,
It is difficult to apply this technology to a distributed software construction environment. Therefore, a new distributed object technology that solves these problems is needed.

More specifically, in the system shown in FIG. 2, all the update information is transferred from the lower computer to the management unit of the upper computer X, and the information is updated. Is configured to be transferred from the computer X to all lower-level computers, whereby all computers can share the same information. However, the lower-level computer and the computer X always have the same information. It is essential that information can be transferred. However, in general, when the amount of management information increases due to the expansion of the network or the performance of the communication line is insufficient, it is not always necessary to transfer sufficient information between the upper computer and the lower computer. Was not guaranteed.

In other words, depending on the state of the communication line, delays in information transfer and deterioration of responsiveness are caused. Unless necessary information transfer is performed, information owned by each computer is different. As a result, a situation has arisen in which the reliability of computer information management has deteriorated.

Therefore, the present invention has been made to solve such a conventional unsolved problem, and its purpose is to increase the amount of management information due to the expansion of the network and to have insufficient communication line performance. In such a case, a plurality of high-level computers are provided, and it is possible to quickly register new information or the like using any of the high-level computers, and all the computers constituting the computer network share the same information. The point is to provide a means for quickly performing the sharing process.

Another object of the present invention is to register new information and the like in consideration of the network state, to make it possible to carry out these processes more quickly, and to make the information owned by each computer consistent. In other words, the point is to provide a means for improving the reliability of information. Still another object of the present invention is to provide means for quickly and reliably performing synchronization processing between higher-level computers.

[0030]

In order to achieve the above object, according to the first aspect of the present invention, information owned by each of a plurality of name servers communicably connected to each other via a network is made identical. In the method of performing information management in such a manner, a hierarchical structure is configured by a plurality of lower-level name servers and a plurality of upper-level name servers, and the lower-level name server to which registration application (application) of new object information is performed is After confirming the uniqueness of the registration information in the name space in which it exists (check), registering the information of the new object (register), and when the registration is completed, selecting one of the upper name servers (select), The information of the registered new object is transferred (transmitted) to the upper-level name server, and the upper-level server to which the information of the new object is transferred is transmitted. Musaba performs sharing of other upper name servers and information (synchronizing) information management method using a computer network is provided.

According to a second aspect of the present invention, there is provided a method for performing information management so that information owned by a plurality of name servers connected to each other via a network so as to have the same information. , A hierarchical structure is constructed with the lower-level name servers and multiple higher-level name servers,
The lower-level name server that has made the inquiry (request) for the object information confirms its own registration information (check), and if the information related to the request is unknown information that does not exist in its own registration information, the upper-level name server Server 1
Requesting the upper-level name server to transfer the information of the entire network (request), and the requested higher-level name server transfers the information registered by itself to the lower-level name server. An information management method used is provided.

According to a third aspect of the present invention, in any one of the first and second aspects, when the apply or the requested lower-level name server selects one of the higher-level name servers, There is provided an information management method using a computer network, wherein a full message is sent to a name server, and an upper name server which responds to the response message earliest is selected as an upper name server to be selected. .

Here, a message for information update and reference request can be cited as a message to be fully notified to each higher-level name server. According to the fourth aspect of the invention,
2. The high-level name server according to claim 1, wherein the high-level name server shares information with another high-level name server (synchronization).
At this time, the communication status with other higher-level name servers is determined to be one of the normal mode, the unstable mode, and the disconnection mode. The transfer is performed, and when the communication state is in the unstable mode, the synchronous message transfer is performed. When the communication state is in the disconnected mode, the other higher-level name servers are registered in a queue list. An information management method using a computer network is provided.

According to a fifth aspect of the present invention, in the fourth aspect, the upper-level name server further sets the communication state of the upper-level name server registered in the queue list to a mode other than the disconnection mode. When this happens, the sharing of information with this higher-level name server is referred to as “extended ihave /
The present invention provides an information management method using a computer network, which is performed by a "sendme" protocol.

It is also conceivable to set a registration upper limit time at the time of registration in the queue list, and to delete the one having passed the registration upper limit time from the queue list.

Next, according to the invention according to claims 6 to 10, a system according to another aspect of the present invention is proposed. First, according to the invention according to claim 6, it is possible to communicate with each other via a network. In a system that manages information so that information owned by each of a plurality of name servers connected to each other is the same, a hierarchical structure is configured by a plurality of lower name servers and a plurality of upper name servers, and information on a new object is provided. The lower-level name server to which the registration application (application) has been made confirms the uniqueness of the registration information in the namespace in which it exists (check), and registers (registers) the information of the new object. Upon completion,
Means for selecting (selecting) one of the high-order name servers and transferring (transmitting) information of the registered new object to the high-order name server;
An information management system using a computer network, which includes means for sharing (synchronizing) information with another higher-level name server that has received the transfer of information on a new object, is provided.

According to the seventh aspect of the present invention, there is provided a system for managing information so that information owned by each of a plurality of name servers connected to each other via a network is the same. Form a hierarchical structure with lower-level name servers and multiple higher-level name servers,
The lower-level name server that has made the inquiry (request) for the object information confirms its own registration information (check), and if the information related to the request is unknown information that does not exist in its own registration information, the upper-level name server Server 1
And a means for requesting the upper-level name server to transfer information of the entire network.
The requested upper-level name server is provided with an information management system using a computer network, which includes means for transferring information registered by itself to the lower-level name server.

According to the invention of claim 8,
8. The method according to claim 6, wherein the means for selecting one of the apply name servers provided in the apply or the requested subordinate name server broadcasts a message to each of the superordinate name servers, and outputs a response message. There is provided an information management system using a computer network, wherein the information management system is a means for setting an upper-level name server that has given a response the earliest as a higher-level name server to select.

According to a ninth aspect of the present invention, in the sixth aspect, the means for sharing (synchronizing) information provided in the upper-level name server determines a communication state with another upper-level name server. Normal mode, unstable mode, and determines whether to correspond to the disconnect mode, if the communication state is normal mode, perform asynchronous message transfer, if the communication state is unstable mode, An information management system using a computer network, wherein a means for performing synchronous message transfer and registering another higher-level name server in a queue list when a communication state is a disconnection mode is provided. You.

Further, according to the tenth aspect of the present invention,
In claim 9, the higher-order name server further comprises:
When the communication state of the higher-level name server registered in the queue list changes to a mode other than the disconnection mode, the information sharing with this higher-level name server is referred to as “extended ihave”.
/ Sendme (Ihab Sendomi) protocol. An information management system using a computer network is provided.

According to the eleventh aspect of the present invention, in the method for managing information so that each of a plurality of computers communicably connected to each other via a network has the same information, A hierarchical structure is composed of a lower computer and a plurality of upper computers, and the lower computer for which registration application for new information has been made selects one of the upper computers and transfers all information owned by the upper computer. And updates the information owned by itself with the transferred information. If there is no overlap between the updated information and the new information, the new information and the updated information are replaced by the information newly owned by the own. The registered information is transferred to the selected higher-level computer, and the selected higher-level computer is
In response to a request for transfer of information from a lower computer, all information owned by the lower computer is transferred to the lower computer, and the registered information transferred from the lower computer is newly added to information owned by the own computer. An information management method using a computer network is provided.

According to the twelfth aspect of the present invention, in the method for managing information so that information owned by each of a plurality of computers communicably connected to each other via a network is the same, a plurality of A lower-level computer and a plurality of higher-level computers form a hierarchical structure, and the lower-level computer for which information has been applied for reference selects one of the higher-level computers and transfers all information owned by this higher-level computer. Request, update the information owned by itself using the transferred information, refer to the updated information, perform a reference process, and the selected higher-level computer performs the transfer of the information from the lower-level computer. An information management method using a computer network is provided which transfers all information owned by itself to the lower-level computer in response to a request. Here, as reference processing, if there is information to be referred to, this information is transmitted to the other computer that has made the reference application, while if there is no information to be referred to, it indicates that there is no information. A process of transmitting the message to another computer that has made the reference application is included.

According to the thirteenth aspect,
13. The computer according to claim 11, wherein the lower-level computer for which registration or reference has been applied is one of the upper-level computers.
In the case of selecting one of the above, information management using a computer network characterized in that a message is fully broadcast to each higher-level computer, and the higher-level computer that responded to the response message earliest is selected as the higher-level computer to be selected. A method is provided.

For example, the lower-level computer transmits a packet of a predetermined size to each higher-level computer, obtains a time from this transmission until a response message indicating that the packet has been received, and determines which of the higher-level computers responds. The host computer is selected by detecting whether the message was replied the earliest.

According to the fourteenth aspect, in any one of the eleventh and twelfth aspects, the selected higher-level computer further performs information sharing (synchronization processing) with another higher-level computer. When performing this synchronization, the communication state with the other host computer is changed to the normal mode, the unstable mode,
In addition, it determines which of the disconnection modes is applicable, and performs normal information transfer when the communication state is the normal mode, and performs information transfer by error correction processing when the communication state is the unstable mode. And when the communication state is the disconnection mode, another higher-level computer is registered in a queue list to provide an information management method using a computer network.

Further, according to the invention of claim 15,
15. The selected higher-level computer (selected-side computer) according to claim 14, wherein when the communication state of the higher-level computer (registered-side computer) registered in the queue list changes to a mode other than the disconnection mode, the selected higher-level computer (selected-side computer) becomes its own. Transfers all the information owned to the registered computer, uses the information transferred from the registered computer as new own information, and refers to the information owned by itself and the received information. In addition, an information management method using a computer network is provided, which transfers information owned only by itself to the selected computer.

Next, according to the invention according to claims 16 to 20, a system according to another aspect of the present invention is proposed. According to the invention according to claim 16, the systems are communicably connected to each other via a network. In a system that manages information so that the information owned by each of the plurality of computers becomes the same, a hierarchical structure is configured with a plurality of lower computers and a plurality of upper computers, and an application for registration of new information is made. The lower computer selects one of the upper computers, requests the transfer of all the information owned by the upper computer, updates the information owned by itself with the transferred information, and updates the updated information with the new information. A means for registering new information and update information as new information owned by itself, and a means for transferring the registered information to the selected host computer when there is no overlap with the information The selected higher-level computer has means for transferring all the information owned by itself to the lower-level computer in response to a request for information transfer from the lower-level computer, and the registration transferred from the lower-level computer. Information management system using a computer network, the information management system comprising:

According to the seventeenth aspect of the present invention, there is provided a system for managing information so that each of a plurality of computers communicably connected to each other via a network has the same information. The lower-level computer and a plurality of higher-level computers form a hierarchical structure, and the lower-level computer to which the information reference application has been made selects one of the higher-level computers and requests transfer of all information owned by this higher-level computer. Means, using the transferred information, updating information owned by itself, referring to such new information, and performing means for performing reference processing, wherein the selected higher-level computer is An information management system using a computer network is provided which includes means for transferring all information owned by itself to the lower-level computer in response to the information transfer request.

According to the eighteenth aspect,
18. The method according to claim 16, wherein the means for selecting one of the high-order computers included in the low-order computer for which registration or reference has been made is for notifying all the high-level computers of a message and for replying a response message. The information management system using a computer network is characterized in that it is means for setting the higher-level computer that performed the first time as the higher-level computer to be selected.

According to the nineteenth aspect, in any one of the sixteenth and seventeenth aspects, the selected higher-level computer further performs information sharing (synchronization processing) with another higher-level computer. Synchronizing means, comprising:
The communication state with another host computer is determined to be one of the normal mode, the unstable mode, and the disconnection mode.If the communication state is the normal mode, normal information transfer is performed. When the communication state is in the unstable mode, the information is transferred by the error correction processing, and when the communication state is in the disconnected mode, the other upper computers are registered in the queue list. An information management system using a computer network is provided.

Further, according to the twentieth aspect,
20. The communication device according to claim 19, wherein the selected higher-level computer (selected-side computer) further includes: when a communication state of the higher-level computer (registered-side computer) registered in the queue list is set to a mode other than the disconnection mode. A means for transferring all information owned by the registered computer to the registering computer; and a means for converting information transferred from the registered computer to new own information. An information management system using a computer network, comprising: means for referring to information and received information to transfer information owned only by the own side to the selected computer.

When the synchronization process for a certain registered computer is completed, the selected computer deletes the name of the registered computer from the queue list. It should be noted that the time registered in the queue list may be managed, and those whose time has passed a predetermined time or more may be deleted from the queue list.

[0053]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the drawings, and then specific examples will be described.

Here, a new distributed object technology which realizes an efficient and highly reliable name service function in consideration of the problems of the existing distributed object technology will be described.

First, the structure of a new name service will be described, and an object information management method (registration and reference of information) using the function will be described. Next, a method of sharing information by a plurality of top-level name servers will be described. (1) An object information management method using a name service. (1) -1 Naming graph structure.

The proposed naming service forms a naming graph as shown in FIG. The difference between this graph and the conventional naming graph shown in FIG. 1 is that there are a plurality of top-level name servers, and the plurality of servers share object information of the entire network. (1) -2 Object information management method.

A method of managing (registering and referencing) object information using the name service will be described. First, the information registration procedure is as follows (FIG. 5). FIG. 6 is a flowchart showing the registration procedure.

(1) Apply for registration of the information of the newly created object to the name server NS _1a in the name space (Appl)
y (apply), FIG. 5 (1), FIG. 6, step S600). (2) Upon receiving the application, the server NS _1a checks the uniqueness of the registration information in the space A (Check), and FIG. 5 (2)
, FIG. 6 step S610) and register (Register (register), FIG. 5 (2), FIG. 6 step S620).

(3) When the registration is completed, the NS _1a selects one NS _0i from a plurality of higher-order name servers NS _0i (i = 1 to n) by a procedure described later (Select), and Transfer registration information to the server (Tr
ansmit (transmit), FIG. 5 (3), FIG. 6 step S
630).

(4) Server N that received information by transfer
S _0i shares information with other servers by a method described later (Synchronize,
FIG. 5 (4), FIG. 6 step S640)).

On the other hand, a method of referring to information is as follows (FIG. 7). FIG. 8 is a flowchart showing this reference procedure. (1) Query the name server NS _1a in the namespace for object information (Request, FIG. 7)
(1), step S800 in FIG. 8).

(2) The server NS _1a checks the registration information in the space A (Check), FIG. 7 (2), FIG. 8, step S810. In the case of unknown information, one NS selected from a plurality of higher-order name servers NS _0i (i = 1 to n)
_0i , request information of the entire network (Request
(Request), FIG. 7 (2), FIG. 8, step S810).

(3) Information is transferred from the upper server NS _0i (Transmit), and the information is searched (FIG. 7 (3), step S820 in FIG. 8). (1) -3 How to select from multiple name servers.

A method for selecting information from a plurality of name servers NS _0i (i = 1 to n) at the time of registration and reference of information will be described. First, the name server NS _1i (i = a, b, c ‥) that manages each name space first has information on the name server in the upper hierarchy. And this information,
It is appropriately corrected according to the change (movement, increase, decrease) of the upper server. The server selection method is performed as follows using this information.

(1) All upper name servers NS _0i (i
= 1 to n), the message of information update and reference request
Present in Multicast. (2) The presented server responds to the presentation destination with a response time that takes into consideration the load state of its own processing, the network state, and the like.

(3) The server with the quickest response is actually requested to update and refer to the information. (2) A method for sharing management information by a plurality of servers. A method for sharing management information between a plurality of name servers NS _0i (i = 1 to n) at the highest level will be described. Less than,
(1) Classify the network status between each server, and (2) Describe a method of transmitting information between multiple servers based on the classification. (3) As a method of synchronizing information between servers having different data, a method that extends the “ihave / sendme” protocol will be described. (2) -1 Classification of network status.

A method for sharing information among a plurality of higher-level servers is as follows.
It depends on the network status between each server. Therefore, the network state between the servers is classified into the following three modes. (1) Normal mode A state in which the effective communication speed can be sufficiently secured.

(2) Unstable mode A state in which the effective communication speed is slow, but communication is still possible. (3) Disconnection mode The effective communication speed is extremely slow, communication is not possible, or the connection is disconnected.

Note that the detection of the effective communication speed is performed by the ICMP protocol used in the ping program and the like (refer to the document “J.
Postel, "Internet Control Message Protocol," RFC792,
Sep, 1981: Internet Control Message Protocol). The speed is determined based on the Round Trip Time between the servers. (2) -2 An information transmission method between a plurality of name servers.

Under the above-described classification of the network state,
_{Assume that an} arbitrary server NS _0j (1 ≦ j ≦ n) is selected from the upper servers NS _0i (i = 1 to n) in FIG. 4 and a request to update information is _issued from the lower server NS _1a . In this case, information is transferred by the following procedure (FIG. 9).

(1) The server NS _0j is connected to all the other servers N
For S _0i (i = 1 to n, i ≠ j), the network status between the server and the server is detected. (2) For each server, transfer information using the following method based on the detected network status.

(A) In the case of the normal mode, the change information is transferred to the server as it is. The transfer method is asynchronous message transfer (Oneway). (b) In the case of the unstable mode, the change information is directly transferred to the server. The transfer method is synchronous message transfer (Twoway) in order to ensure reliability.

(C) In the case of the disconnection mode, the server is registered in the queue list, and the information transfer is not performed. For the server registered in the queue list, the following (3) and subsequent processes are performed.

(3) For the server registered in the queue list, the network state is periodically detected, and the following processing is performed according to the state. For the server, a registration upper limit time is set when the server is registered in the queue.

(A) When the mode changes to the normal mode or the unstable mode, different data may be managed between the server and the server. Therefore, the extension shown in the following (2) -3
Synchronize information using the ihave / sendme protocol. After the end of the synchronization process, it is deleted from the queue.

(B) If the network status does not change,
That is, if the server remains in the disconnection mode, the network disconnection time with the server is compared with the registration upper limit time, and if it is within the time, the server remains in the queue. If the time has exceeded the upper limit, it is removed from the queue and the server is treated as terminated. (2) -3 Method of synchronizing information between servers.

When the network state between the servers NS _0i and NS _{0j recovers} from a disconnected state to a communicable state (normal mode, unstable mode), each server can hold information on different objects. There is.
Therefore, it is necessary to synchronize data between servers after network recovery. Here, the method is NN
The ihave / sendme protocol, one of the TP protocols (Brain Kantor, Pil Lapsley: "Network News Tra
nsfer Protocol: A Proposed Standard for theStream-B
ased Transmission of the News, "RFC977, Feb, 1986: Network News Transfer Protocol:
A Proposed Standard for the Stream Based Transmission of the News ”etc.) and propose a method that takes into account data deletion. Hereinafter, the data synchronization method will be described (FIG. 10).

(1) Each of the servers NS _0i and NS _0j creates indexes I _i and I _j composed of a name space and its last update time for the information held by itself. Then, NS _0i gives I _i to server NS _0j ((1) ihave
(Ihab)).

(2) Server NS receiving index
_{0j compares} that I _i with its own index I _j ((2) compare (compare)). Then, the following two types of difference indexes D _{i, j} and D _{j, i} are created.

[0080]

(Equation 1)

[0081]

(Equation 2)

In this difference creation, if both servers have data in the same space, the latest one of the last update times is validated, and old data is treated as not existing.

(3) NS _{0j requests} NS _0i to transfer data corresponding to the difference index D _{i, j} ((3) sendme (send me)). At the same time, data corresponding to the difference index D _{j, i} is transferred to NS _0i ((3) sen
dyou)

(4) The NS _0i requested to transfer transfers the corresponding data to NS _0j ((4) sendyou). Through the above process, it becomes possible to synchronize data between the servers.

As described above, according to the embodiment of the present invention, it is possible to provide a plurality of top-level name servers,
Information registration and reference can be performed quickly by selecting a higher-level name server that has the quickest response to message transmission by a name server in the name space in consideration of the network state.

Further, since information is shared among a plurality of top-level name servers by a method according to the state of the network, information can be shared quickly and reliably, and the reliability of information is improved. Also, when sharing information between upper-level name servers, by using an "extended ihave / sendme" protocol,
Information can be shared by a simple communication method.

[0087]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First, referring to FIGS. 11, 12, and 13, information registration and reference processing will be specifically described.

FIG. 11 shows the configuration of a system according to an embodiment of the present invention. This system comprises a plurality of upper computers (corresponding to upper name servers) 50a and 50.
, 50n, and a plurality of lower computers (corresponding to lower name servers) 30a, 30b,..., 30n, and each lower computer and each upper computer are communicably connected by a communication line. I have. That is, this system is a computer network having a two-layer structure having a plurality of upper computers and a plurality of lower computers.

Each of the lower computers 30a, 30b,..., 30n
Is a management unit 10 that performs various processes such as registration and reference
, 100n, memory units 200a, 200b,..., 200n for storing information (corresponding to objects) to which identifiers are attached, and state detection units 160a, 160b for detecting the state of the communication line. , ..., 16
0n.

This state detection unit is used to notify the host computer of all the messages and to select the host computer which has responded to the response message the earliest as the host computer. Then, a packet of a predetermined size is transmitted to the host computer, and a time (response time) until a response is obtained from the host computer can be detected for each host computer.

Each of the host computers 50a, 50b,.
50n, management units 150a, 150b,..., 150n for performing various processes such as registration and reference, and memory units 250a, 250 for storing information provided with identifiers.
, 250n.

Each management unit and each state detection unit include, for example, a ROM having a built-in operation program, a CPU which operates according to the operation program, and an R which functions as a work area.
Various electronic devices such as an AM and an interface IC can be realized, and each memory unit can be realized by, for example, a RAM or a hard disk.

Now, the operation of registering new information will be described with reference to FIG. Now, as shown in FIG. 11, the lower-level computer 30a stores an identifier (ID) in the memory unit 200a.
The host computer 50b stores the information with the identifiers a, b, and d in the memory unit 250b, and sends the information from another computer (not shown) to the computer 30a. It is assumed that new information with the identifier c is transmitted and that this new information (corresponding to the information of the newly created object) is registered together with the identifier c. Further, it is assumed that the shortest response time detected by the state detection unit 160a of the computer 30a is when a communication line connected to the computer 50b is used.

In step S1200, the management unit 100a of the computer 30a receives an application for registering new information with the identifier c from another computer (not shown). At this time, in step S1205, the management unit 100a suspends reception of another process.

Next, in step S1210, state detection section 160a detects the state of the communication line, that is,
Check the response time for each host computer. At this time, since the response time to the upper-level computer 50b is the shortest, the computer 50b is selected as the upper-level computer in the optimal communication line state (step S1215).

Next, at step S1220, the management unit 100a sends the memory unit 250 to the computer 50b.
b transmits a command requesting transfer of the information stored in b. On the other hand, the management unit 150b
The information stored in b is transmitted to the computer 30a via the communication line. As a result, information with the identifiers a, b, and d is transmitted to the computer 30a.

In step S1225, the management unit 100a
Updates the information in the memory unit 200a with the information transmitted from the computer 50b (the updated information is update information). At this time, duplicate information is not newly stored. As a result, only the information with the identifier d is newly stored, and the information in the memory unit 200a is the information with the identifiers a, b, and d.

Next, referring to the identifier of the information in the memory unit 200a and the identifier of the information related to the registration application, it is determined whether or not there is any overlap between the two (confirm the uniqueness). If there is a duplicate, the process proceeds to step S1250, where the registration application is rejected, a message to the effect that a duplicate is present is sent to the other computer, and the process ends (END). If there is no duplicate, the process advances to step S1235 to store the identifier c in the memory unit 200a.
Register new information marked with. As a result, information with identifiers a, b, c, and d is stored in the memory unit 200a.

Then, in step S1240, management unit 100a transfers the information stored in memory unit 200a to computer 50b, resumes accepting processing in step S1245, and ends the processing (END). Thus, the identifiers a, b, c, and d are also stored in the memory unit 250b.
Is stored.

As described above, by providing a plurality of high-level computers and considering the state of the communication line, the high-level computer capable of communicating the most information within a predetermined time is selected, and new information is registered using this. Since the processing is performed, the registration processing can be performed quickly.

The information sharing between higher-level computers performed thereafter will be described later. Next, an operation when referring to information will be described with reference to FIG. Similarly to the description of the registration process, the lower-level computer 3
0a stores information with identifiers a and b in the memory unit 200a, and the host computer 50b stores information with identifiers a, b and d in the memory unit 250b. It is assumed that a command for requesting information reference has been sent from another computer (not shown) to the computer 30a. Further, it is assumed that the shortest response time detected by the state detection unit 160a of the computer 30a is when a communication line connected to the computer 50b is used.

In step S1300, it is assumed that the management unit 100a of the computer 30a receives an information reference application from another computer (not shown), that is, an information inquiry. At this time, in step S1305, the management unit 100a suspends reception of another process.

Next, in step S1310, state detection section 160a detects the state of the communication line,
Check the response time for each host computer. At this time, since the response time to the upper-level computer 50b is the shortest, the computer 50b is selected as the upper-level computer in the optimal communication line state (step S1315).

Next, in step S1320, the management unit 100a sends the memory unit 250 to the computer 50b.
b) transmits a command requesting transfer of the information stored in b. On the other hand, the management unit 150b
The information stored in 50b is transmitted to the computer 30 via the communication line.
Send to a. As a result, information with the identifiers a, b, and d is transmitted to the computer 30a.

In step S1325, the management unit 100a
Updates the information in the memory unit 200a with the information transmitted from the computer 50b (the updated information is update information). At this time, duplicate information is not newly stored. As a result, only the information (unknown information) with the identifier d is newly stored, and the memory unit 2
00a stores information with identifiers a, b, and d.

Next, in step S1330, a reference process is executed. Here, when a reference request command is sent from another computer (not shown), and when the identifier of the information to be referenced exists in the memory unit 200a, the information is searched for, and the search result is stored in another computer. On the other hand, when the identifier of the information to be referred to does not exist in the memory unit 200a, a message to the effect that the information to be referred to does not exist is transmitted to another computer. Specifically, if the identifier of the information to be referred is any of a, b, and d, the corresponding information is searched and the searched information is transmitted to another computer. , B, and d, a message indicating that there is no information to be referred to is transmitted to another computer.

Then, in step S1335, the reception of the processing is resumed, and the processing is ended (END). This allows
The reference processing ends. In this way, by providing a plurality of high-level computers and considering the state of the communication line,
Since a higher-level computer capable of communicating the most information within a predetermined time is selected and new information is referred to using the selected computer, reference processing can be performed quickly.

Next, a description will be given of a synchronization process performed between the host computers. This process is a process for making the information owned by each host computer the same. FIG. 14 shows four host computers 50a connected by a communication line so that they can communicate with each other.
It is a block diagram of 50b, 50c, and 50d. Although the number is described as four for convenience of explanation, it is needless to say that the number is not limited to four.

Each of the host computers 50a, 50b, 50c, 5
0d is a queue list 251a, 251b, 251c, 251d described later in addition to the configuration shown in FIG.
(Provided respectively in the memory units 250a, 250b, 250c, and 250d).
State detectors 170a, 170 for detecting the state of the communication line
b, 170c and 170d. Note that the state detection unit provided in the higher-level computer has the same function as the state detection unit provided in the lower-level computer described above.

Now, the synchronization processing will be described with reference to FIG. 15, but now, in step S1240 in FIG. 12, information in which identifiers a, b, c, and d are assigned from the lower computer 30a to the upper computer 50b. Is transferred,
Description will be made assuming that step S1500 in FIG. 15 is started.

In step S1500, the management unit 15
0b is the identifier a, b, c, d from the lower computer 30a.
In step S1505, the reception of other processes is interrupted.

Next, in step S1510, the management unit 1
An identifier of the information transferred to the memory 50b;
It is checked whether or not all the identifiers of the information already stored in the server overlap with each other. If the identifiers overlap, the process is terminated because new information has not been transferred (end). Is registered in the memory unit 250b as new information with a unique identifier. As described above, the existing information is information with identifiers a, b, and d, and the new information is information with identifier c. b,
Information to which c and d are added is registered (step S1515).

Next, in step S1520, the state detecting section 170b detects the response time via the communication lines corresponding to all the other higher-level computers, and gives the detected response time to the management section 150b.

The management unit 150b determines the status of each network with reference to the response time (step S15).
25). For this purpose, a first time and a second time longer than this are set in advance, and the response time is set at a certain reference time (for example, the time of transmitting a predetermined amount of packets is set as the reference time). If the response time is greater than or equal to the first time and less than the second time from a certain reference time if the response time is less than the first time, the network condition is “unstable”. Mode ", and if the response time is equal to or longer than a second time from a certain reference time, the state of the network is set to" disconnection mode ". That is, the computer 50b
Grasps which mode the communication line with the other computers 50a, 50c, and 50d is in.

In the normal mode (step S1530), the management unit 150b transfers the information stored in the memory unit 250b by the normal transfer method, and in the disconnection mode (step S1540), it waits. The computer name of the transfer destination is registered in the matrix list 251b, and in the case of the unstable mode (step S1535), the transfer is performed by the error correction processing. In addition, transfer by error correction processing
When the management unit 150b transfers the information, it transmits data indicating the size and number of the transfer information to the transfer destination, and confirms that the data indicating the same size and number has been transmitted from the transfer destination. Is what you do.

Then, all the other computers 50a, 50
Transfer processing is performed for c and 50d until the transfer according to the corresponding mode is completed (step S1545). Further, in step S1550, the management unit 150b starts accepting processing, thereby ending a series of processing. Yes (end).

As described above, when a certain higher-level computer synchronizes with another higher-level computer, information transfer is performed in consideration of the state of the network, so that synchronization processing can be performed quickly and reliably.

Next, the operation of the host computer using the queue list will be described with reference to FIG. As an example, the computer 50b performs this processing.
The description will be made assuming that the computer 50c is already registered in the queue list 251b.

First, in step S1600, the management unit 1
50b refers to the registered contents of the queue list 251b, grasps the information (50c) of the registered computer, and drives the state detecting unit 170b to detect the state of the corresponding communication line. This processing may be performed, for example, periodically at a predetermined time period.

In step S1605, the management unit 150b refers to the response time detected by the state detection unit 170b to check whether the communication mode has changed to a mode other than the disconnection mode. If the disconnection mode is maintained, the process ends (ends) in step S1620 while holding the registered contents of the queue list, while the mode is changed to a mode other than the disconnection mode, that is, the normal mode or the unstable mode. If it does, step S161
0, after performing the synchronization process between the computer 50b and the computer 50c, the management unit 150b sets the queue list 251b
From the computer 50c (step S161).
5) End the processing (END).

The synchronization processing in step S1610 is performed by the management unit 150b of the computer 50b.
All the information stored in the memory unit 250b is transferred to the management unit 150c, and the management unit 150c of the computer 50c receiving the information refers to the received information and the information stored in the memory 250c. The information stored only in the memory 250c is stored in the management unit 150b of the computer 50b.
Transfer to The management unit 150b stores the transmitted information in the memory 250b. Thereby, the computer 50
b memory 250b and the computer 250c memory 250c
, The same information is stored, and sharing of the memory information of both computers is realized. In this way, information sharing can be reliably performed by using a simple communication method, and by using a queue list, when a mode other than the disconnection mode is set, synchronous processing can be reliably performed. .

Each of the management units may be configured so that the computer names registered in the queue list are deleted from the list even after a predetermined time has passed since the registration. It is not necessary to perform the processing for the processing that is unlikely to occur, and the processing operation of the entire system can be speeded up.

The synchronization processing described with reference to FIGS. 15 and 16 is performed not only when registering new information but also when referring to information, so that the information possessed by each computer is always updated. It is preferable to match them so that the reliability of information management is improved.

As described above, according to the present invention, even when the amount of management information increases due to the expansion of the network or the performance of the communication line is insufficient, a plurality of higher-level computers are provided. It becomes possible to promptly apply for registration of new information using such a host computer. In addition, since registration of new information and application of information are performed in consideration of the network state, it is possible to perform these processes more quickly. As a result, the consistency of the information possessed by each computer is improved, in other words, the reliability of the information is improved.

In addition, since synchronization is performed between upper-level computers in consideration of the network state, synchronization can be quickly and reliably performed, thereby improving the reliability of information.
Further, by using the queue list, it is possible to reliably perform the synchronization processing when the mode is changed to a mode other than the disconnection mode.

[0126]

As described above, claim 1 or claim 6
According to the present invention, the computer network has a hierarchical structure in which a plurality of higher-order name servers are provided. When the registration is completed, one of the higher-level name servers is selected, the information of the registered new object is transferred to this higher-level name server, and the upper Since the name server shares information with other higher-level name servers, it becomes possible to quickly apply for registration of information on a new object using one of the higher-level name servers. Thereby, the reliability of information owned by each computer is improved.

According to the second or seventh aspect of the present invention, the computer network has a hierarchical structure in which a plurality of high-order name servers are provided. If the information pertaining to the request is unknown information that does not exist in its own registration information, one of the higher-level name servers is selected, and a request is made to this higher-level name server to transfer information of the entire network, The requested higher-level name server transfers the information registered by the higher-level name server to the lower-level name server, so that the object information can be quickly referenced using any of the higher-level name servers.

According to the third or eighth aspect of the invention, in addition to the effects of any one of the first, second, sixth and seventh aspects, the applied or requested lower-level name server is one of the upper-level name servers. If one is selected, the message is fully broadcast to each upper-level name server, and the upper-level name server that responded to the response message earliest is selected as the upper-level name server to be selected. It becomes possible to apply for registration of new object information and to refer to object information more quickly, thereby further improving the reliability of information owned by each computer.

According to the fourth or ninth aspect of the present invention, in addition to the effect of the first or sixth aspect, when the upper-level name server shares information with another upper-level name server, communication is performed. Determine whether the state corresponds to the normal mode, unstable mode, or disconnection mode,
When the communication status is the normal mode, the asynchronous message transfer is performed. When the communication status is the unstable mode, the synchronous message transfer is performed. When the communication status is the disconnection mode, the server waits for another upper-level name server. Since the information is registered in the queue list, it is possible to obtain an effect that the information can be shared between upper-level name servers without fail in consideration of the network state.

According to the fifth or tenth aspect of the present invention, in addition to the effect of the fourth or ninth aspect, the upper-level name server further communicates with the upper-level name server registered in the queue list. When the state changes to a mode other than the disconnection mode, the information sharing with the upper-level name server is performed by the “extended ihave / sendme” protocol, so that the information can be shared by a simple communication method. The effect that it can be performed is obtained.

According to the eleventh or sixteenth aspect of the present invention, the computer network has a hierarchical structure in which a plurality of higher-level computers are provided. Request to transfer all information owned by this host computer, update its own information with the transferred information, and if there is no overlap between the updated information and the new information, And update information as newly owned information, and transfers the registered information to the selected host computer.
The selected higher-level computer transfers all the information owned by itself to the lower-level computer in response to the information transfer request from the lower-level computer, and newly registers the registered information transferred from the lower-level computer. , It ’s your own information,
It is possible to promptly apply for registration of new information using any of the higher-level computers. Thereby, the reliability of information owned by each computer is improved.

According to the twelfth or seventeenth aspect of the present invention, the computer network has a hierarchical structure in which a plurality of high-order computers are provided. Then, the host computer requests transfer of all information owned by the host computer, updates the information owned by itself using the transferred information, performs reference processing by referring to the updated information, and performs selection processing on the other hand. In response to the information transfer request from the lower-level computer, the higher-level computer transfers all the information owned by itself to the lower-level computer, so that the information is quickly referenced using any of the higher-level computers. It becomes possible.

According to the thirteenth or eighteenth aspect, in addition to the effects of any one of the eleventh, twelfth, sixteenth, and seventeenth aspects, the lower-level computer for which registration or reference has been applied is one of the upper-level computers. When selecting one, register the new information in consideration of the network status, because all messages are sent to each host computer, and the host computer that responded to the response message earliest is selected as the host computer to be selected. Applications and reference to information can be performed more quickly, and the reliability of information owned by each computer is further improved.

According to the invention of claim 14 or 19, in addition to the effect of any one of claims 11, 12, 16 and 17, when the selected host computer performs synchronization, The communication status with the host computer of the normal mode,
Unstable mode or disconnect mode is determined.If the communication state is normal mode,
Normal information transfer is performed. When the communication state is in the unstable mode, information transfer is performed by error correction processing. When the communication state is in the disconnected mode, another upper computer is registered in a queue list. Therefore, there is an effect that the information can be shared among the upper computers reliably in consideration of the network state.

Further, according to the invention of claim 15 or 20, in addition to the effect of claim 14 or 19, the communication state of the host computer (registered computer) registered in the queue list is set to the disconnection mode. In the mode other than the above, the selected higher-level computer (selected computer) transfers all the information owned by itself to the registered computer, and transfers the information transferred from the registered computer to the new own computer. On the other hand, the registered computer refers to the information owned by itself and the received information, and transfers the information owned only by itself to the selected computer. The advantage is that information can be shared between computers.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a naming graph in which name servers according to the related art are hierarchically arranged.

FIG. 2 is a block diagram showing a conventional system.

FIG. 3 is a flowchart for explaining the operation of a conventional system.

FIG. 4 is an explanatory diagram of a naming graph in which a plurality of name servers are hierarchically arranged.

FIG. 5 is an explanatory diagram of registration of object information.

FIG. 6 is a flowchart illustrating a registration procedure of object information.

FIG. 7 is an explanatory diagram of reference to object information.

FIG. 8 is a flowchart illustrating a procedure for referring to object information.

FIG. 9 is an explanatory diagram showing a state in which a certain name server transfers information to another name server.

FIG. 10 is an explanatory diagram of a data synchronization method between servers.

FIG. 11 is a block diagram of a system according to an embodiment of the present invention.

FIG. 12 is a flowchart illustrating a procedure of a registration process.

FIG. 13 is a flowchart illustrating a procedure of a reference process.

FIG. 14 is a block diagram illustrating a synchronization process performed by a host computer.

FIG. 15 is a flowchart illustrating a procedure of a synchronization process.

FIG. 16 is a flowchart illustrating a process using a queue list.

[Explanation of symbols]

 30n Lower computers 50a, 50b, ..., 50n Upper computers 100a, 100b, ..., 100n Management units 150a, 150b, ..., 150n Management units 160a, 160b, ..., 160n State detection units 170a, 170b, , 170n State detection unit 200a, 200b, ..., 200n Memory unit 250a, 250b, ..., 250n Memory unit 251a, 251b, ..., 251n Queue list

Claims (20)

[Claims] 1. A method for managing information so that each of a plurality of name servers connected to each other via a network so as to communicate with each other has the same information, wherein a plurality of lower name servers and a plurality of upper name servers are provided. The lower-level name server that has applied for registration of new object information (Apply) confirms the uniqueness of the registration information in its own namespace (Check), and creates a new structure. Register the object information (register), and when the registration is completed, select one of the upper name servers (select) and transfer the registered new object information to this upper name server (transmit). The upper-level name server that has received the transfer of the new object information shares the information with other higher-level name servers (synchronization). Size), the information management method using a computer network. 2. A method for managing information so that each of a plurality of name servers communicably connected to each other via a network has the same information, wherein a plurality of lower name servers and a plurality of upper name servers are provided. The lower-level name server that has made an inquiry (request) for object information confirms (checks) its own registration information, and the information related to the request exists in its own registration information. In the case of unknown information that does not exist, one of the higher-level name servers is selected and a request is made to this higher-level name server to transfer the information of the entire network (request). The requested higher-level name server registers the information registered by itself. Information management method using a computer network, wherein the information is transferred to the lower-level name server. 3. The method according to claim 1, wherein the apply or the requested lower-level name server broadcasts a message to each of the upper-level name servers when selecting one of the higher-level name servers. An information management method using a computer network, wherein an upper-level name server that responds to a response message earliest is selected as an upper-level name server to be selected. 4. The communication system according to claim 1, wherein when the upper-level name server shares information with another upper-level name server (synchronization), the communication status with the other upper-level name server is changed to a normal mode, an unsynchronized mode, or a non-normal mode. Judge whether the communication mode corresponds to the stable mode or the disconnection mode.If the communication status is the normal mode, perform the asynchronous message transfer.If the communication status is the unstable mode, perform the synchronous message transfer. An information management method using a computer network, wherein when the communication state is a disconnection mode, another higher-level name server is registered in a queue list. 5. The high-level name server according to claim 4, wherein the high-level name server further communicates with the high-level name server when the communication state of the high-level name server registered in the queue list changes to a mode other than the disconnection mode. Information sharing is called "extended iha
An information management method using a computer network, which is performed by a "ve / sendme" protocol. 6. A system for managing information so that each of a plurality of name servers communicably connected to each other via a network has the same information, a plurality of lower name servers and a plurality of upper name servers. The lower-level name server that has applied for registration of new object information (apply) confirms the uniqueness of the registration information in the namespace where it exists (check), and creates a new object Means for registering (registering) information, and when the registration is completed, one of the higher-level name servers is selected (selected), and the information of the registered new object is transmitted (transmitted) to this higher-level name server. Means for receiving the transfer of the information of the new object, Comprising means for performing Yuka (synchronizing) information management system using a computer network. 7. A system for managing information so that each of a plurality of name servers communicably connected to each other via a network has the same information, a plurality of lower name servers and a plurality of upper name servers. The lower-level name server that has made an inquiry (request) for object information confirms (checks) its own registration information, and determines that the information related to the request does not exist in its own registration information. In the case of information, there is provided a means for selecting one of the upper-level name servers and requesting (requesting) the upper-level name server to transfer information of the entire network, and the requested upper-level server registers itself. An information management system using a computer network, comprising: means for transferring information to the lower-level name server. 8. The high-level name server according to claim 6, wherein the means for selecting one of the higher-level name servers provided in the apply or the lower-level name server requested includes: An upper-level name server that responds to the response message earliest as a higher-level name server to be selected; and an information management system using a computer network. 9. The system according to claim 6, wherein the means for sharing information (synchronizing) provided in the upper-level name server changes a communication state with another upper-level name server into a normal mode, an unstable mode, and a disconnection mode. Judge which of the modes is applicable. If the communication status is the normal mode, perform the asynchronous message transfer. If the communication status is the unstable mode, perform the synchronous message transfer and disconnect the communication status. An information management system using a computer network, which is a means for registering another higher-level name server in a queue list in a mode. 10. The high-level name server according to claim 9, wherein the high-level name server further communicates with the high-level name server when the communication state of the high-level name server registered in the queue list changes to a mode other than the disconnection mode. Information sharing is called "extended iha
An information management system using a computer network, comprising: means for performing a “ve / sendme” protocol. 11. A method for managing information so that information owned by each of a plurality of computers communicably connected to each other via a network is the same, wherein a plurality of lower computers and a plurality of upper computers are hierarchized. With the structure configured, the lower-level computer for which registration application for new information has been made selects one of the higher-level computers, requests transfer of all information owned by this higher-level computer, and uses the transferred information. If the information owned by the user is updated and there is no overlap between the updated information and the new information, the new information and the updated information are registered as newly owned information, and the registered information is The selected higher-level computer transfers the information owned by itself to the lower-level computer in response to the information transfer request from the lower-level computer. The information the registration that has been transferred from newly, and owned information thereof, information management method using a computer network. 12. A method for managing information so that information owned by a plurality of computers communicably connected to each other via a network is the same, wherein a hierarchy is formed between a plurality of lower computers and a plurality of upper computers. With the structure configured, the lower-level computer that has applied for information reference selects one of the higher-level computers, requests transfer of all information owned by this higher-level computer, and uses the transferred information. And updates the information owned by itself, performs a reference process by referring to the updated information, and the selected upper computer receives all information owned by itself in response to the information transfer request from the lower computer. An information management method using a computer network, wherein the information is transferred to the lower-level computer. 13. The computer according to claim 11, wherein the lower-level computer for which registration or reference has been applied, when selecting one of the higher-level computers, sends a full message to each of the higher-level computers and responds. An information management method using a computer network, wherein a higher-level computer that responds to a message first is selected as a higher-level computer to be selected. 14. The computer according to claim 11, wherein the selected higher-level computer further performs information sharing (synchronization processing) with another higher-level computer, and when performing the synchronization, Determines whether the communication status with the host computer corresponds to the normal mode, unstable mode, or disconnected mode. If the communication status is the normal mode, performs normal information transfer and performs communication. A computer network for transmitting information by error correction processing when the state is unstable mode, and registering another host computer in a queue list when the communication state is disconnection mode; Information management method using. 15. The selected higher-level computer (selected computer) according to claim 14, wherein when the communication state of the higher-level computer (registered computer) registered in the queue list changes to a mode other than the disconnection mode. ) Transfers all the information that it owns to the registering computer, uses the information transferred from the registering computer as new own information, and the registered computer receives the information that it owns. An information management method using a computer network, comprising referring to information and transferring information owned only by itself to the selected computer. 16. A system for managing information so that information owned by a plurality of computers communicably connected to each other via a network is the same, wherein a hierarchy is formed by a plurality of lower computers and a plurality of upper computers. The lower-level computer that has constituted the structure and has applied for registration of new information selects one of the higher-level computers, requests the transfer of all information owned by this higher-level computer, and uses the transferred information Means for updating the information owned by itself and, when there is no overlap between the updated information and the new information, means for registering the new information and the updated information as newly owned information; and To the selected higher-level computer, and the selected higher-level computer transfers all information owned by itself in response to a request to transfer information from the lower-level computer. An information management system using a computer network, comprising: means for transferring to a rank computer; and means for newly setting the registered information transferred from a lower-level computer to own information. 17. A system for managing information so that each of a plurality of computers communicably connected to each other via a network has the same information, wherein a plurality of lower computers and a plurality of upper computers are hierarchized. The lower-level computer that has configured the structure and has applied for information reference selects one of the higher-level computers, requests the transfer of all information owned by this higher-level computer, and uses the transferred information. Means for updating information owned by itself, performing a reference process by referring to the updated information, and wherein the selected higher-level computer receives the information transfer request from the lower-level computer in response to a request to transfer information from the lower-level computer. An information management system using a computer network, comprising: means for transferring all information owned by the computer to the lower-level computer. 18. A method according to claim 16, wherein said means for selecting one of the higher-order computers provided in said lower-order computer for which registration or reference has been applied is for fully transmitting a message to each of said upper-level computers. An information management system using a computer network, which is means for selecting a higher-level computer that has given the response message the earliest as a higher-level computer to be selected. 19. The computer according to claim 16, wherein the selected higher-level computer further includes a synchronization unit that shares information (synchronization processing) with another higher-level computer, and the synchronization unit includes: Judge whether the communication state with other host computers corresponds to the normal mode, unstable mode, or disconnected mode.If the communication state is the normal mode, perform normal information transfer, When the communication state is in the unstable mode, the information is transferred by error correction processing. When the communication state is in the disconnected mode, the other upper computers are registered in the queue list. Information management system using a computer network. 20. The communication system according to claim 19, wherein the selected higher-level computer (selected-side computer) further has a communication state of a higher-level computer (registered-side computer) registered in a queue list other than a disconnection mode. When it becomes, there are means for transferring all information owned by itself to the registered computer, and means for transferring the information transferred from the registered computer as new own information. An information management system using a computer network, comprising: means for referring to information owned by itself and received information to transfer information owned only by itself to the selected computer.