JPH09245004A - Object-oriented distributed processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate the actualization of a system and to increase the execution efficiency by performing a process for matching the value of the attribute of an object stored in a specific area in each computer. SOLUTION: Memory matching processes 1mi and 2mi match the storage contents of a specific area of an application program 1ap on a computer 1c with the storage contents of a specific area of a program 2ap on a computer 2c by sending and receiving data in specific areas of application programs in memories to and from other computers through communication processes 1cm and 2cm. Further, an object attribute matching management table lot is included. Then object attribute matching management processes 1dm and 2dm equalize the values of the attributes of objects of the same kind (class) in different computers by making good use of the memory matching processes to regard those objects as the same object.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an object-oriented distributed processing apparatus in which a plurality of objects on a plurality of computers request processing from each other and proceed as a whole. In particular, the present invention relates to an object-oriented distributed processing device capable of executing processing in cooperation between objects on different computers.

[0002]

2. Description of the Related Art In recent years, computer systems based on so-called object orientation, which describe a program by a set of objects in which data and procedures related thereto are put together, have increased. In an object-oriented system, a plurality of objects call each other's procedures to execute processing as a whole. Furthermore, multiple computers are connected via a network, and objects on those computers are
Object-oriented distributed processing systems that communicate with each other and execute processing in cooperation have come into use.

In a distributed processing system, a method of communicating between programs on different computers and coordinating processing to advance a remote procedure call (Remote Procedure Call)
There is a method called RPC). RPC is, for example, Network Computing System Reference, published by Prentice Hall.
Manual (1990) (Network Computing System
Reference Manual (1990)). RPC
Allows a program on one computer to call a procedure (function) on another computer in the same manner as a normal procedure call (function call). In the remote procedure call, the program that calls the procedure is called the client, and the program in which the called procedure exists is called the server. It is also possible for multiple clients to call the procedure of one server. Remote procedures can be used to communicate between objects on different computers. That is, a process request source object on a computer calls a procedure possessed by a process request destination object on another computer using RPC. This makes it possible to realize an object-oriented distributed processing system.

Further, in Japanese Patent Laid-Open No. 5-12025, a virtual object corresponding to a process request destination object existing in another process is provided in the process in which the process request source object exists, and the process request source object When a process is requested to a virtual process request destination object in the same process, the virtual process request destination object
A method is described in which inter-process communication communication is performed with a process request destination object existing in another process, and the process request destination object executes a process. By using inter-process communication processing that enables communication between processes on different computers, this method enables objects on different computers to request each other for processing and execute processing in cooperation with each other. This makes it possible to realize an object-oriented distributed processing system.

[0005]

In object orientation, in order to request processing to another object, it is necessary to specify an object ID that is an identifier of the object and a procedure name. This is because in object-oriented, there can be multiple objects that have procedures with the same name. However, in normal RPC, the client side can only specify the procedure name on the server side, so in order to realize inter-object communication using RPC, it is necessary to perform processing such as specifying the object ID on the server side with the argument of the procedure. There is. Further, the computer on the server side requires complicated processing for determining which procedure of which object is to be called from the object ID and the procedure name specified by the argument. A conventional RPC is, for example, a network computing system reference, published by Prentice Hall.
Manual (1990) (Network Computing System
As described in the Reference Manual (1990)),
The socket function provided by the operating system (OS) is used.
The processing of the RPC itself is complicated, including the protocol processing such as CP / IP. As described above, the above-mentioned first conventional technique has a problem that the inter-object communication process becomes complicated and it takes time to implement. Further, according to this, there is a problem that the execution efficiency also decreases.

On the other hand, in the method described in Japanese Patent Laid-Open No. 5-12025, a virtual object corresponding to the processing request destination object is provided on the computer in which the processing request source object exists, and the virtual object is processed. The programmer needs to describe the inter-object communication process. Therefore, there is a problem that the programmer has a heavy load and the productivity of the system is not good.

The present invention provides an object-oriented distributed processing device which is easy to implement, has excellent execution efficiency, and which can realize communication between objects on different computers in a system that reduces the burden on the programmer and improves system productivity. The purpose is to do.

[0008]

According to a first aspect of the present invention, in an object-oriented distributed processing device in which a plurality of objects on a plurality of computers cooperate to advance processing,
Object attribute matching management means for designating an area in the memory of each computer in which the attribute of the object is stored as a specific area, an attribute value of the object stored in the specific area on the certain computer, and the other computer An object-oriented distributed processing device comprising: a memory matching processing unit that performs attribute matching processing that matches the attribute value of an object stored in the specific area.

The object attribute matching management means is
An area that includes all the areas that store the attributes of the object that is the object of the attribute matching processing may be designated as one specific area.

The object attribute matching management means is
Each of the areas in which the attributes of the object to be subjected to the attribute matching processing are stored may be individually designated as the specific area.

The object attribute matching management means is
Of the areas that store the attributes of the object that is the object of the attribute matching processing, only the areas that store the separately defined specific attributes may be designated as the specific areas. .

The operation of the first aspect will be described.

The object attribute matching management means on the computer designates, to the memory matching means on the computer, the area on the memory storing the attribute value of the object as the specific area for performing the memory matching. The memory matching means communicates with the memory matching means on the other computer, so that the value of the specific area designated by the object attribute matching management means on the computer and the object object attribute on the other computer. The value of the specific area designated by the matching management means is matched. This makes it possible to match the attribute values of a plurality of objects on different computers. Objects with the same attribute value can be virtually regarded as one object.
As described above, objects on different computers can proceed with processing in cooperation with each other.

As a second aspect of the present invention, in an object-oriented distributed processing device in which a plurality of objects on a plurality of computers advance processing by message communication, each of the computers has a storage area of 1 or 2 or more. A transmission message storage unit that is divided into transmission division areas and stores a message to be transmitted by an object on the computer to an object on another computer by dividing it into each transmission division area according to the destination, and 1 or 2 whose storage area is set corresponding to the transmission division area
A reception message storage unit that is divided into the above-mentioned reception division areas and that delivers the message stored in each reception division area to the object corresponding to the reception division area,
A memory matching process for matching the content stored in each reception segment area of the reception message storage unit with the message stored in the transmission segment area corresponding to the reception segment area of the transmission message storage unit of another computer. An object-oriented distributed processing device is provided.

The transmission division area and the reception division area may be set corresponding to each of the computers. In this case, it is preferable to further have a dispatching means provided for each computer for deciphering the message and allocating which procedure to activate.

The transmission segment area and the reception segment area may be set corresponding to each of the objects. In this case, it is preferable to further have a dispatching means provided for each object for deciphering the above message and allocating which procedure is activated.

The transmission division area and the reception division area may be set corresponding to each type of message.

The operation of the second aspect will be described.

The contents of the message transmitted by the processing request source object on the computer are sorted and stored in the transmission segment area of the transmission message storage means on the computer.

The memory matching means matches the contents of each reception division area in the reception message storage means with the message stored in the corresponding transmission division area. In this way, the contents (message) of the reception divisional area, the contents of which are matched with those of the corresponding transmission divisional area, are passed to the corresponding object. This enables message communication between objects on different computers.

[0021]

Embodiments of the present invention will be described below in detail with reference to the drawings.

(1) First Embodiment of the Present Invention A first embodiment of the present invention will be described. According to the first embodiment of the present invention, the values of the attributes of objects on different computers are made to coincide with each other so that those objects can be regarded as virtually the same object. Is an object-oriented distributed processing device in which processing is performed in cooperation with each other.

First, the outline of the first embodiment of the present invention will be described. FIG. 1 is a conceptual diagram of the first embodiment of the present invention.

There are two objects, object 1 and object 2, in the computer 1c. Object 1 has procedures 1p1, 1p2, etc. and attributes 1d1, 1d2, etc., while object 2 has procedures 2p1, 1p2, etc.
2p2 and the like and attributes 2d1 and 2d2 and the like. There are two objects, object 2 and object 3, in the computer 2c. Object 2 is procedure 2p
1, 2p2, etc. and attributes 2d1, 2d2, etc.,
On the other hand, the object 3 has procedures 3p1, 3p2, etc., and attributes 3d1, 3d2, etc.

The object 1 on the computer 1c calls the procedure of the object 2 on the computer 1c to write or read the attribute value of the object 2.
On the other hand, the object 3 on the computer 2c calls the procedure of the object 2 on the computer 2c to write or read the attribute value of the object 2. The object 2 on the computer 1c and the object 2 on the computer 2c are of the same type (same class, same type). Therefore, the procedure of the object on the computer 1c and the computer 2c
The procedure of the above object is the same.

In the first embodiment of the present invention, the computer 1c
The values of the attributes 2d1, 2d2, etc. of the upper object 2 are made equal to the values of the attributes 2d1, 2d2, etc. of the object 2 on the computer 2c. As a result, the attribute value of the object 2 on the computer 1c and the attribute value of the object 2 on the computer 2c are always kept the same. Therefore, the computer 1c
Object 2 above and object 2 on computer 2c
Can be virtually regarded as one object. That is, in the object-oriented distributed processing device of FIG. 1, both the object 1 and the object 3 call the procedure of the object 2 which is a single object.

As described above, in the first embodiment of the present invention,
Multiple objects on multiple computers can cooperate in processing.

The first embodiment of the present invention will be described in detail below.

FIG. 2 is an overall configuration diagram of the first embodiment of the present invention. The first embodiment of the present invention is a distributed processing device in which a computer 1c and a computer 2c are connected by a network 1n. Each of the computers 1c and 2c has a processor 1
p, 2p, memories 1m, 2m, input / output (I / O) device 1
i, 2i exist. A program executed by a computer exists in the memories 1m and 2m, and the programs are executed by the processors 1p and 2m.
p reads out and executes the process. Operating systems (OS) programs 1os and 2os, object attribute matching management processing programs 1dm and 2dm, and application programs 1ap and 2ap exist in the memories 1m and 2m, respectively. The present embodiment is composed of two computers, but the same applies to the case where it is composed of three or more computers.

An application program is a set of objects. The application program 1ap on the computer 1c includes an object 1 and an object 2. Application program 2a on computer 2c
p includes object 2 and object 3. An object is a unit of a program that collects a set of data and related procedures. In the application program 1ap on the computer 1c, there are two objects, an object 1 and an object 2, and the object 1 has procedures 1p1, 1p2, etc. and an attribute 1d.
Object 1 has a procedure 2
It has p1, p2, etc., and attributes 2d1, 2d2, etc. Application program 1ap on computer 1c
There are two objects, object 2 and object 3, and object 2 is procedure 2p1, 2p2.
p3 has attributes 2d1, 2d2, etc., and the object 3 has procedures 3p1, 3p2, etc., and attributes 3d1, 3
d2 and the like.

An object can call a procedure of another object belonging to the same computer during the processing of the procedure. As a result, objects in the same computer can call each other's procedures and proceed with processing. In this embodiment, each object existing in the distributed processing device has an identifier called an object ID, and the object can be uniquely specified in the distributed processing device by specifying the object ID. To call an object procedure, specify the object ID and the procedure name to call. Arguments can be added if necessary.

Generally, in an object-oriented program, the type of object is called a class and each object is called an instance. A class defines the data structure and behavior of an instance. In the present embodiment, when simply called as an object, it means an instance. Objects belonging to the same class may have the same procedure and data structure, but differ only in the value of the data. The object 2 in the application program 1ap on the computer 1c of this embodiment and the object 2 in the application program 2ap on the computer 2c belong to the same class.

The program written by the programmer is the definition of the class. The method of defining a class is defined by an object-oriented language. Object-oriented languages include, for example, Smalltalk-80 and C ++. In addition,
Smalltalk-80 is described in, for example, Addison Wesley Publishing Company, Smalltalk 80, The Language &It's.
Implementation (1983) (Smalltalk-80 T
he Language and its Implementation (1983)). For C ++, addison
Addison Wesley Publishing Company
Published by The Annotated C + Plus Reference Manual (1990)
+ Reference Manual).

Further, for example, CORBA ("CORB
A "is a trademark of the Object Management Group, and if the interface specification of an object is defined in the interface description language, the procedure and data of the object can be described without using an object-oriented language (for example, C language). Yes, COR
Regarding BA, for example, an object management group and X / Open (Object Management
Group and X / Open) The Common Object Request Broker Architecture: Architecture and Specification (1992) (The Co
mmon Object Request Broker: Architecture and Speci
fication (1992)) (Japanese translation: published by Japan Office Automation Association, Common Object Request Broker-Structure and Specifications-).

The class definition declares the object data structure and procedure. For example, in the language called Smalltalk-80, individual data of objects are called instance variables, and procedures are called methods. Also, C +
In the language called +, individual data are called members and procedures are called member functions. OMG / CORB
In A, individual data are called attributes, and procedures are called operations. In the present embodiment, each data included in the object data is called an attribute, and each procedure is called a function. The class definition described by the programmer is converted into an executable program by a compiler, and it can be loaded on a memory and executed.

The OS is software that executes programs and performs basic services with users. OS1o
s and 2os are schedulers 1sc and 2sc,
Device driver 1dd, 2dd, communication processing 1cm, 2
cm, and memory matching processing 1mi and 2mi are included.
The schedulers 1sc and 2sc schedule execution of application programs. The device drivers 1dd and 2dd control the input / output devices 1i and 2i to perform input / output processing. The communication processes 1 cm and 2 cm transmit / receive data to / from other computers via a network. The memory matching processes 1mi and 2mi perform communication processing 1c on data in a specific area of the application program on the memory.
By sending and receiving with another computer by m, 2 cm,
Storage contents of a specific area of the application program 1ap on the computer 1c and the program 2a on the computer 2c
It has a function of matching the stored contents of a specific area of p.

OS scheduler, device driver,
Communication processing is performed, for example, by Prentis Hall (Prenti Hall).
ce-Hall), The Design of the Unix
Operating System (1986) (The Design
It can be realized by the method described in the UNIX Operating System (1986)).

Memory matching processing 1 mi in the present invention,
2 mi will be described. Memory matching processing 1mi, 2
In the mi program, a matching area identifier for identifying the matching target area, the start address of the matching target area, and the size (number of bytes) of the matching target area are stored in association with each other. It contains a matching area correspondence table. The structure of the matching area correspondence table 1mt is shown in FIG.

In this embodiment, the contents of the matching area correspondence table 1mt are set by the object attribute matching management processing, which will be described later, when the system is started up. When the system is started up, the object attribute matching management processing uses the matching area identifier for identifying the matching target storage area and the start address of the matching target storage area for all memory matching target areas on the computer. The size (number of bytes) of the matching target storage area is specified, and the matching area correspondence table 1mt is initialized.

The memory matching processing program receives the matching data transmission processing for transmitting the storage data in a predetermined storage area on its own computer to another computer and the storage data transmitted from the other computer. And a matching data reception process of storing the data in a predetermined storage area on the own computer.

The matching data transmission process will be described with reference to FIG. The matched data transmission processing 100 is started by designating a matched area identifier for identifying a matched storage area. The activated matching data transmission processing 100 first refers to the matching area correspondence table to obtain the start address and the number of bytes of the matching target area corresponding to the specified matching area identifier (processing 101). Next, the stored contents on the memory corresponding to the number of bytes are read from the start address as matching data (process 102). Then, one communication data is created from the matched area identifier and the matched data (process 103), and the communication data is processed by the communication process 1c.
It is transmitted by m and 2 cm (process 104).

Next, the matching data reception process will be described with reference to FIG. In the matched data reception processing 200, first, communication processing 1c is performed on communication data sent from another computer.
It is received by m and 2 cm (process 201). Next, the matching area identifier and the matching data are extracted from the communication data (process 202). Then, by referring to the matching area correspondence table, the start address and the number of bytes of the matching target area corresponding to the extracted matching area identifier are obtained (process 2).
03). Finally, the sent matching data is stored by the number of bytes from the obtained start address of the matching target area (process 204).

Object attribute matching management processing 1dm,
2 dm will be described. An object attribute matching management table 1ot is included in the programs of the object attribute matching management processes 1dm and 2dm. In this object attribute matching management table 1ot, as shown in FIG. 6, for objects existing on different computers but desired to be regarded as the same object, the object ID and the head of the storage area storing the attributes of the object are stored. The address, the size (the number of bytes) of the storage areas of all the attributes of the object, and the matching area identifier of the object are stored in association with each other.

Object attribute matching management table 1o
The contents of t may be given by the system administrator or may be automatically generated by analyzing the object program with a tool. In the case of the present embodiment, in FIG. 2, since it is desired to consider the object 2 on the computer 1c and the object 2 on the computer 2c to be the same, the object attribute matching management table 1ot on each computer has The object ID of object 2 in
The start address of the storage area storing the attributes of the object, the size (byte number) of the storage area of all the attributes of the object, and the matching area identifier corresponding to the object ID are stored. Matching area identifier is the object ID
To be unique to.

As described above, the object attribute matching management processing program initializes the matching area correspondence table 1mt at system startup. That is, the matching area identifiers corresponding to all the object IDs stored in the object attribute matching management table 1ot, the start address of the matching target storage area, and the size (number of bytes) of the matching target storage area To set. The start address of the matching target storage area is the start address of the storage area that stores the attributes of the object, and the size (byte count) of the matching target storage area is the storage size (byte count) of all attributes of the object. is there.

Further, the object attribute matching management processing program periodically executes the object attribute matching processing. The object attribute matching process will be described with reference to FIG. The object attribute matching process 300 first reads the matching region identifier corresponding to the object ID of one object ID stored in the object attribute matching management table 1ot (process 301). Next, by specifying the read matching area identifier, the matching data transmission processing (see FIG. 4) of the memory matching program is started (processing 302). Then, it is checked whether or not these processes have been executed for all object IDs stored in the object attribute matching management table 1ot (process 303). If not executed, the above processes are executed for the next object ID. To do.

As described above, in this embodiment, the object attribute matching management processes 1dm and 2dm change the attribute values of the objects of the same type (class) on different computers.
By making them identical using the memory matching process, it is possible to regard those objects as the same object. As a result, it becomes possible to proceed with processing in cooperation among a plurality of objects distributed on a plurality of computers.

(2) Effect of the first embodiment of the present invention According to the first embodiment, the object attribute matching process is activated at regular intervals. There is no need to describe, which has the effect of reducing the burden on the programmer.

According to the first embodiment, the description of the message communication process between objects on different computers is not necessary, which reduces the burden on the programmer and facilitates the reuse of the program. is there.

(3) Second Embodiment of the Present Invention A second embodiment of the present invention will be described. Conventionally, message communication between objects has been realized by explicit message communication performed by designating a destination (computer name, object name) and a procedure name. In contrast,
In the second embodiment, the contents of the storage area for storing the message transmitted by the processing request source object and the storage contents of the area for storing the message received by the processing request destination object on another computer are made to match. As a result, message communication is realized between objects on different computers.

First, the outline of the second embodiment of the present invention will be described. FIG. 8 is a conceptual diagram of the second embodiment of the present invention.

The computer 1c has procedures 1p1 and 1p2.
Etc., there is an object 1 having attributes 1d1, 1d2, etc. The computer 2c has procedures 2p1 and 2p2
Object 2 having attributes 2d1, 2d2, etc. exists.

The object 1 on the computer 1c writes the message to the object 2 on the computer 2c in the transmission message storage areas 1n1, 1n2, etc. On the other hand, the object 2 on the computer 2c reads the message written in the received message storage areas 2r1, 2r2, etc. and executes the procedure. Then, the storage contents of the reception message storage areas 2r1, 2r2, etc. on the computer 2c are matched with the storage contents of the transmission message storage areas 1n1, 1n2, etc. on the computer 1. As a result, computer 1
The object 1 on c can send a message to the object 2 on the computer 2c without executing an explicit communication process.

As described above, in the second embodiment of the present invention,
Multiple objects on multiple computers can cooperate in processing.

The second embodiment of the present invention will be described in detail below.

FIG. 9 is an overall configuration diagram of the second embodiment of the present invention. The second embodiment of the present invention is, like the first embodiment, a distributed processing device in which a computer 1c and a computer 2c are connected by a network 1n. Similar to the first embodiment, each of the computers 1c and 2c has a processor 1
p, 2p, memories 1m, 2m, input / output (I / O) device 1
i, 2i exist. A program executed by a computer exists in the memories 1m and 2m, and the programs are executed by the processors 1p and 2m.
p reads out and executes the process. The memories 1m and 2m have operating system (OS) programs 1os and 2os and a transmission / reception message storage area 1m, respectively.
s, 2ms, application programs 1ap, 2a
p exists.

Transmission / reception message storage area 1 ms, 2 ms
Is an area for storing messages exchanged between objects on different computers. The transmission / reception message storage areas 1ms and 2ms are respectively the reception message storage areas 1r1, 1r2 and the transmission message storage area 1n.
1, 1n2, etc., received message storage areas 2r1, 2r2
Etc. and transmission message storage areas 2n1, 2n2, etc.

The OSs 1os and 2os are the same as those in the first embodiment of the present invention.

Memory matching processing in OS 1mi, 2mi
The operation of is similar to that of the first embodiment. However, the target of memory matching is the send / receive message storage area 1 m.
s, 2 ms. Further, the contents of the matching area correspondence table are given in advance by the system administrator or determined at the time of design. In the case of the present embodiment, in FIG. 9, the contents of the reception message storage areas 2r1, 1r2, etc. are made to match the contents of the transmission message storage areas 1n1, 1n2, etc., respectively.
The contents such as 1r2 are stored in the transmission message storage area 2n.
The matching area correspondence table is set so as to match the contents of 1, 2, 2 and the like.

The application program consists of objects and stubs. The application program 1ap on the computer 1c includes an object 1, server stubs 1s1 and 1s2 of the object 1, client stubs 2c1 and 2c2 of the object 2, and the like. Object 1 has procedures 1p1, 1p2, etc., and attributes 1d1, 1d
2 and so on. The application program 2ap on the computer 2c includes the object 2 and the object 2
The server stubs 2c1, 1s2, etc., and the client stubs 1c1, 1c2, etc. of the object 1 are included. Object 2 has procedures 2p1, 2p2, etc., and attributes 2d1, 2d2
Etc.

During the processing of the procedure, the object can send a message for invoking the procedure of the object to the object of another computer.
In this embodiment, each object existing in the distributed processing device has an identifier called an object ID, and the object can be uniquely specified in the distributed processing device by specifying the object ID.

A client stub is a program that creates a message to be sent to an object on another computer. The message source object can create the message to be transmitted by calling the client stub corresponding to the destination object. Also,
The program that interprets the received message and activates the procedure of the corresponding object at the destination of the message is the server stub. In FIG. 9, a computer 1c
When the object 1 above sends a message to the object 2 on the computer 2c to activate the procedure, the client stubs for that are the client stubs 2c1, 2c2, etc. on the computer 1c, and the server stub is the computer 2
These are the server stubs 2s1, 2s2, etc. on c. When the object 2 on the computer 2c sends a message to the object 1 on the computer 1c to activate the procedure,
Client stubs therefor are the client stubs 1c1, 1c2, etc. on the computer 2c, and the server stubs are server stubs 1s1, 1s2, etc. on the computer 1c.

The basic processing of the client stub and the server stub is described in, for example, Network Computing System Reference Manual (199), published by Prentice Hall.
Year 0) (Network ComputingSystem Reference Manual
(1990)) and the same processing as the client stub and server stub in the RPC.
However, the client stub and server stub in the RPC call communication processing such as a socket (socket) function provided by the operating system (OS) to send and receive messages, whereas in the present embodiment, the client stub Differs from the server stub in that the message is written into the transmitted message storage area and the matched data transmission processing of the memory matching processing program is started, and the server stub reads the message from the received message storage area.

The operation of the client stub will be described with reference to FIG. The client stub process 400 is called from the process request source object, and first, a message to the process request destination object is created (process 401).
Next, the created message is written in the transmission message storage area (process 402). Then, the matching data transmission processing of the memory matching processing program is activated (processing 40).
3).

The operation of the server stub will be described with reference to FIG. The process 500 of the server stub is called from the matching data reception process of the memory matching process program.
A message is read from the received message storage area (process 501). Then, the procedure corresponding to the message of the object of the processing request destination is called (processing 502).

In this embodiment, the client stubs 2c1, 2c2, etc. on the computer 1c respectively write a message in the transmission message storage areas 1n1, 1n2, etc., and activate the matching data transmission process of the program of the memory matching process 1mi. Client stub 1 on computer 2c
c1, 1c2, etc. are respectively the transmission message storage area 2n
A message is written to 1, 2, 2n2, etc., and the matching data transmission processing of the program of the memory matching processing 2mi is started.

Server stubs 1s1, 1s on the computer 1c
2 and the like are received message storage areas 1r1 and 1r2, respectively.
Read the message from etc. The server stubs 2s1, 2s2, etc. on the computer 2c read messages from the received message storage areas 2r1, 2r2, etc., respectively.

As described above, in this embodiment, when the client stub of the object of the processing request source writes the created message in the transmission message storage area, it is received by the memory matching processing on the computer where the processing request destination object exists. The same message is generated in the message storage area. The server stub of the transmission destination object can read the message and activate the procedure of the processing request destination object. As a result, it becomes possible to proceed with processing in cooperation among a plurality of objects distributed on a plurality of computers.

(4) Effect of the Second Embodiment of the Present Invention According to the second embodiment of the present invention, the send message storage area and the receive message storage area are prepared for each procedure of the object of the processing request destination. The processing at the computer to which the processing is requested is simplified, and the processing execution efficiency is improved.

(5) Other Embodiments of the Present Invention In the first embodiment of the present invention described above, the object ID is uniquely given to the entire distributed processing apparatus, but it may be given uniquely within the computer. . This gives the object ID
As the object address (pointer) can be used as it is, there is an effect that the execution efficiency is improved.

In the first embodiment of the present invention described above, the object attribute matching process is activated periodically, but it may be activated only when the attribute of the target object is rewritten. This has the effect of improving processing efficiency when the frequency of rewriting the attributes of objects is low.

In the first embodiment of the present invention described above, the object attribute matching process is executed at once for all the attributes of the object for which the matching process is performed, but it may be executed for each individual attribute unit of the object. As a result, the matching process can be executed only for the required attributes,
This has the effect of improving processing efficiency.

In the first embodiment of the present invention described above, the object ID and the matching area identifier are different.
Both may be the same identifier. This has the effect of simplifying the processing and improving the execution efficiency.

In the second embodiment of the present invention described above, the transmission message storage area and the reception message storage area are prepared for each procedure of the object of the processing request destination. However, one of these areas (sent message storage area, received message storage area) may be prepared for each object. Further, as shown in FIG. 12, one may be prepared for each computer. With such a configuration, the memory usage efficiency is improved.
However, in this case, for each object (or,
It is necessary to provide a dispatcher (for each computer).

In the memory matching processing in the first and second embodiments of the present invention described above, the matching area identifiers are used to associate the storage areas for performing matching processing on different computers. However, without using a matching identifier,
The addresses of the storage areas in which the matching process is performed may be the same among the computers. In this way, the memory matching process can be simplified and the execution efficiency can be improved.

[0076]

According to the present invention, a plurality of objects existing on different computers can perform simple message communication without designating a destination (computer name, object name) and a procedure name. By utilizing the memory matching processing, the processing can proceed in cooperation. As a result, the system can be easily realized and the execution efficiency can be improved.

Further, according to the present invention, the object program can be written without being aware of the communication processing between the objects. This has the effect of reducing the burden on the programmer and improving the productivity of the system.

[Brief description of drawings]

FIG. 1 is a diagram showing an outline of a first embodiment of the present invention.

FIG. 2 is a diagram showing an overall configuration of a first embodiment of the present invention.

FIG. 3 is a diagram showing a configuration of a matching area correspondence table according to the first embodiment of this invention.

FIG. 4 is a flowchart of a matched data transmission process according to the first embodiment of this invention.

FIG. 5 is a flowchart of a matched data reception process according to the first embodiment of this invention.

FIG. 6 is a diagram showing a configuration of an object attribute matching management table according to the first embodiment of this invention.

FIG. 7 is a flowchart of object attribute matching processing according to the first embodiment of this invention.

FIG. 8 is a diagram showing an outline of a second embodiment of the present invention.

FIG. 9 is a diagram showing an overall configuration of a second exemplary embodiment of the present invention.

FIG. 10 is a flowchart of client stub processing according to the second embodiment of this invention.

FIG. 11 is a flowchart of server stub processing according to the second embodiment of this invention.

FIG. 12 is a diagram showing an outline of another embodiment of the present invention.

[Explanation of symbols]

1c, 2c ... Computer 1n ... Network 1p, 2p ... Processor 1m, 2m ... Memory 1i, 2i ... Input / output device 1os, 2os ... Operating system 1dm, 2dm ... Object Attribute matching management process 1ap, 2ap ... Application program 1, 2, 3 ... Object 1p1, 1p2, 2p1, 2p2, 3p1, 3p2 ...
・ Procedure 1d1, 1d2, 2d1, 2d2, 3d1, 3d2 ...
Attribute 1sc, 2sc ... Scheduler 1dd, 2dd ... Device driver 1 cm, 2 cm ... Communication processing 1 mi, 2 mi ... Memory matching processing 1 mt ... Matching area correspondence table 1 dm, 2 dm ... Object attribute matching management processing 1ot ... Object attribute matching management table 1ms, 2ms ... Send / receive message storage area 1r1, 1r2, 2r1, 2r2 ... Received message storage area 1n1, 1n2, 2n1, 2n2 ... Transmission message storage area 1c1, 1c2, 2c1, 2c2 ... Client stub 1s1, 1s2, 2s1, 2s2 ... Server stub

Front page continuation (72) Inventor, Shoji Suzuki, 1-1-1, Omika-cho, Hitachi, Hitachi, Ibaraki Hitachi, Ltd. Hitachi Research Laboratory (72) Inventor, Nobuhisa Kobayashi 1070, Igemo, Hitachinaka, Ibaraki Hitachi Factory Mito Factory

Claims (10)

[Claims] 1. In an object-oriented distributed processing device in which a plurality of objects on a plurality of computers cooperate with each other to perform processing, a specific area is stored in the memory of each computer, in which the attributes of the objects on the computer are stored. Object attribute matching management means designated as, and the attribute value of the object stored in the specific area on the certain computer and the attribute value of the object stored in the specific area on the other computer are matched. An object-oriented distributed processing device, comprising: a memory matching processing means for performing an attribute matching processing to be realized. 2. The object attribute matching management means,
2. The object-oriented distribution according to claim 1, wherein an area including all of the areas that store the attributes of the object that is the object of the attribute matching processing is specified as one specific area. Processing equipment. 3. The object attribute matching management means,
The object-oriented distributed processing according to claim 1, wherein each of the areas in which the attributes of the object that is the object of the attribute matching processing are stored is individually designated as the specific area. apparatus. 4. The object attribute matching management means,
Among the areas in which the attributes of the object that is the object of the attribute matching process are stored, only the area in which a separately defined specific attribute is stored is specified as the specific area. The object-oriented distributed processing device according to claim 1. 5. In an object-oriented distributed processing device in which a plurality of objects on a plurality of computers advance processing by message communication, the storage area of each of the computers is divided into one or more transmission partition areas. , A message to be sent by an object on the computer to an object on another computer by allocating the message to each transmission partition area according to the destination, and storing the storage message in the transmission partition area. Received message storage means, which is divided into one or two or more reception division areas set correspondingly, and which passes the message stored in each reception division area to the object corresponding to the reception division area, and the reception message. The contents stored in each reception division area of the storage means are stored in the transmission message of another computer. An object-oriented distributed processing device, comprising: memory matching processing means for matching a message stored in a transmission segment area corresponding to the reception segment area of the storage means. 6. The object-oriented distributed processing device according to claim 5, wherein the transmission partition area and the reception partition area are set corresponding to each of the computers. 7. The object-oriented distributed processing device according to claim 6, further comprising: dispatching means provided for each computer, which dispatches the procedure for decoding the message and allocating which procedure to activate. 8. The object-oriented distributed processing device according to claim 5, wherein the transmission segment area and the reception segment area are set corresponding to each of the objects. 9. An object provided for each object, which decodes the message and distributes which procedure is to be activated,
9. The object-oriented distributed processing device according to claim 8, further comprising dispatching means. 10. The object-oriented distributed processing device according to claim 5, wherein the transmission segment area and the reception segment area are set corresponding to respective message types.