JP2001325237A - Program execution method and device in distributed processing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide program execution method and device capable of controlling the execution of a remote method by a computer unit without changing the constitution of a group or the program of a calling origin. SOLUTION: A system manager specifies constitution information for each group by using a group information operation part and also specifies an operation mode for each computer by using an operation mode operation part. A group communication control part receives the calling request of the remote method, decides the appropriate calling method of the remote method corresponding to the group constitution information and the operation mode and actually performs the calling processing of the remote method by using a communication processing part.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method and an apparatus for executing a program in a distributed processing system implemented by a workstation or a personal computer.

[0002]

2. Description of the Related Art In recent years, distributed computing technology, in particular, CORBA (Common Object Request Broker Architectur)
Systems using e) have become widespread. CORBA is an OMG (Object Manage
ment Group). By conforming to the CORBA architecture, developers can easily build distributed applications that connect between different models and different languages.

[0003] In CORBA, "Fault tolerant CORB
A Using entity Redundancy Request For Proposal, or
As described in “bos / 98-04-01”, support for a fault-tolerant function that provides transparency to an application through multiplexing is being studied. This allows developers to implement complex fault-tolerant processing without having to implement
Availability of a distributed application can be improved (Prior Art 1).

On the other hand, in the technology described in Japanese Patent Application Laid-Open No. 9-212381, in a group RPC or the like, from among a plurality of callable procedure execution sections, the number of procedure execution sections to be actually called and the calling order are dynamically changed. Can be determined. This makes it possible to easily improve the fault tolerance and distribute the load by multiplexing the procedure execution units with the selection logic optimal for the purpose of the user (prior art 2).

FIG. 2 shows an example of a system configuration using the above-mentioned conventional technology. On the three computers 2010 to 2030, the object a forming the group A 2210 and the copies 2080 to 2120 of the object, the object b forming the group B 2220, the copies 2130 to 2160 of the object, the object c forming the group C 2230, and the Duplicate object 2
170 to 2180. Objects 2190 to 2200, which are arranged on two computers 2040 to 2050 and issue a remote method call request, to the objects in the group are duplicated LAN 206
The remote method is called via any one of 0 and 2070.

In a distributed system composed of a plurality of computers, some of the computers are temporarily disconnected from the execution system without stopping the processing of the entire system or modifying the program. In general, a function that enables a program to be processed independently of an execution system in the same system is realized as an important function for the purpose of program maintenance or the like.
For example, in the above configuration, the computers 2010 to 2050
Therefore, in order to temporarily disconnect the computers 2030 and 2050 from the execution system as a test system, a request for calling a remote method from the object 2190 on the computer 2040 is received by the objects on the computers 2010 and 2020 and the The request for calling the remote method from the object 2200 is sent to the computer 203.
It is necessary to make the object on 0 accept. In other words, it is necessary to make it possible to execute different object selection logics for the execution type object and the test type object. However, in the above prior arts 1 and 2, a plurality of object selection logics cannot be designated for one group. Therefore, in order to realize this function, the execution group and the test group are divided. , Different object selection logic must be specified for each group. Therefore, all objects 2110, 2120, 2
150, 2160, and 2180 into groups 2210-2
After excluding each of the groups from 230 and forming three new groups, it is necessary to further modify the program of the object 2200 so as to make a request to call a remote method of the newly formed group.

In a distributed system connected by multiplexed LANs, a function of separating a LAN used for data communication of the active system from a LAN used for data communication of the test system is provided in the network. It is sometimes implemented as a function for the purpose of preventing an increase in traffic. For example, in the above configuration, the computers 2010-2
From 030, the computers 2030 and 2050 are temporarily disconnected from the execution system as test systems, and the communication of test data is performed using a LAN different from the LAN used for communication of execution system data. The data at the time of calling the remote method from the object 2190 on the computer 2040 passes through the LAN 2060,
The data at the time of calling the remote method from the object 2200 needs to pass through the LAN 2070. In other words, the object of the execution system and
For test-based objects, it is necessary to be able to execute different communication path selection logic. However, in the above prior arts 1 and 2, a plurality of communication path selection logics cannot be designated for one group. Therefore, in order to realize this function, the execution system group and the test system Had to be divided, and different communication route selection logics had to be specified for each group.

[0008]

In the above prior art, a group consisting of a set of software modules and a copy of the software module is formed, and at least the group is set in response to a request for calling a remote method for one group. In a method and apparatus for executing a remote method of one or more software modules, a program is executed by receiving a remote method according to the operating state of each computer or communication path without changing the configuration of the group. It was not possible to switch the communication path of the software module to be executed or the remote method. Therefore, for example, when the execution system and the test system are temporarily separated and the program is processed independently, the configuration of the group must be changed each time, and the calling program must be further modified. When the number of arranged groups or software modules is enormous, when the group configuration is complicated, or when the number of calling programs is enormous, there is a problem that the burden on the system administrator is large.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned inconvenience, and an object of the present invention is to execute a remote method according to the operating state of a computer without changing a group configuration or a program that calls a remote method. An object of the present invention is to provide a method and an apparatus for executing a program, which are capable of performing control.

[0010]

In order to achieve the above object, according to the present invention, for each group, a first processing step for inputting information relating to software modules forming the group, A second processing step for inputting selection logic at the time of calling a remote method, a third processing step for inputting or acquiring operation mode information indicating an operation state of the computer for each computer, And a fourth processing step for dynamically determining a method of executing a processing for calling a remote method of the software module based on the information specified by the first, second, and third processing steps. .

By executing these steps, operation mode information indicating the operation state of each computer is designated. When a remote method is called, a software module to be called and a call path are called based on the operation mode information. Determine dynamically. That is, in response to a remote method invocation request from a software module on a computer that specifies the operation mode of the execution system, the remote method of the software module on the computer that specifies the operation mode of the execution system is called, and the operation of the test system is performed. In response to a remote method call request from a software module on a computer that specifies a mode, a remote method of a software module on a computer that specifies a test system operation mode is called. In this way, when some of the computers that make up the system are used separately from the active system, the system administrator only needs to change the operation mode information of the computer to be disconnected and change the group configuration and the program that calls the remote method. It is possible to appropriately switch the call destination and call path of the remote method without making any changes.

[0012]

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

FIG. 1 is a block diagram showing an embodiment of a configuration relating to a program execution method and apparatus of a distributed processing system according to the present invention. Reference numerals 1010 to 1040 denote computers such as workstations and personal computers, to which a plurality of computers are connected via a multiplexed network 1050.

The client computer 1010 includes a communication processing unit 1060 for transmitting and receiving data to and from another computer via a network, a client program 1080 for calling a process via a remote method, and a remote method request from the client program. It comprises a group communication control unit 1070 that controls to accept and communicate with the group appropriately. The server-side computers 1020 to 1030 include a communication processing unit 1060,
It consists of a server program 1090 whose processing is called via a remote method. Computer 1 for system management
040, a communication processing unit 1060, a user input unit 1100 for receiving an input from a user of the computer, a user output unit 1110 for displaying a program execution result and data,
It comprises a storage unit 1140 for storing programs and data, a group information operation unit 1120 for displaying information about groups and changing set values, and an operation mode operation unit 1130 for displaying operation modes and changing set values. Storage unit 114
0, a group configuration storage table 1150 storing configuration information of each group, a group attribute storage table 1160 storing attribute information of each group, and an operation mode storage table 1170 storing operation mode information of each computer.
Is stored.

Subsequently, the group information operation unit 11 shown in FIG.
An example of the process 20 will be described with reference to the flowchart in FIG. First, the group configuration storage table 1150 is read (step 3010), and the group configuration list screen 5
010 is displayed on the user output unit 1110 (step 30).
20), accept input from the user (step 303)
0).

FIG. 4 shows a configuration example of the group configuration table. The group configuration table includes a group identifier 4010 such as a group name, an object identifier 4020 such as an object name or an object reference which is a constituent element of the group, and a computer identifier 40 such as a host name or an IP address of a computer in which the object exists.
30 is a list of lines each including a status change enable / disable flag 4040 indicating whether or not to allow the execution state of the object to be changed according to the operation mode when the operation mode change instruction is issued.

FIG. 5 shows an example of a group configuration list screen. Reference numeral 5020 denotes a menu for performing an operation related to this screen, which can instruct “add group”, “delete group”, and the like. An area 5030 schematically displays the hierarchical structure of the group. Reference numeral 5050 denotes an area for displaying information on the components of the group 5040 selected in the area 5030. The area has an object identifier 5060,
A computer identifier 5070 and a status change enable / disable 5080 are displayed.

After the user's input is started, an area 503 is displayed.
When a specific group among 0 is selected (step 304)
0), the group configuration information is displayed in the area 5050 (step 3050), and the input of the user is accepted again (step 3030). When addition or deletion of a group or an object is instructed from the menu 5020 (step 3060), input of a group identifier, an object identifier, a computer identifier, and a status changeability flag is accepted as necessary, and a line corresponding to the input information is received. After the data is added to or deleted from the group configuration table, the display content of the group configuration list screen is updated (step 3070), and the input of the user is received again (step 3030). When "display group attribute" is selected from menu 5020 (step 3080), group attribute storage table 1160 is read, group attribute screen 7010 is displayed on user output unit 1110 (step 3090), and user input is performed. Accept (Step 3100).

FIG. 6 shows a configuration example of the group attribute storage table. The group attribute storage table includes a group identifier 6010 such as a group name, transmission control logic 6020 for determining an object to be called, retransmission control logic 6030 for calling a remote method again after execution of a remote method, and after completion of execution of a remote method. It is a list of lines consisting of the reception control logic 6040 when determining the remote method execution result. As the transmission control logic, the operation mode is “use” or “standby”
"Execute all remote methods of the object on the computer with no operation mode" stop ", and preferentially execute only one remote method of the object on the computer from the" use "mode side. As the retransmission control logic, "If the execution result is invalid, retransmit the remote method to an object on another computer""If an exception occurs due to a network failure, retransmit the remote method via a different network" specify.
As the reception control logic, "if the execution of the remote method of the object in the" use "mode is completed normally, the execution result is selected; otherwise, the execution result of the remote method of the object in the" standby "mode is selected.""If the execution results on the" use "mode side and the" standby "mode side are the same, select the execution result; otherwise, generate an exception." FIG. 7 shows a screen example of the group attribute screen. 7020 is an area for selecting a transmission control logic, 7030 is an area for selecting a retransmission control logic, and 7040 is an area for selecting a reception control logic.

After the user's input is started, an area 702 is displayed.
When a different control logic from 0 to 7040 is selected (step 3110), the value of the corresponding group attribute storage table is changed, and then the display contents of the group attribute screen are updated (step 3120), and the user inputs again. Is received (step 3100). Also, a setting completion button 705
When 0 is pressed (step 3130), the group attribute screen is hidden (step 3140), and the input of the user is accepted again (step 3030).

On the other hand, when "end" is selected from the menu 5020 (step 3150), the group configuration list screen is hidden (step 3160), and the processing of the group information operation unit is ended.

Subsequently, the operation mode operation unit 113 shown in FIG.
An example of processing 0 is described with reference to the flowchart in FIG. First, the operation mode storage table 1170 is read (step 8010), and the system overall status screen 11 is read.
010 is displayed on the user output unit 1110 (step 80).
20), accept the user's input (step 803)
0).

FIG. 9 shows an example of the configuration of the operation mode storage table. The operation mode storage table contains the host name of the computer and
A computer or path identifier 9010 such as an IP address;
It is a list of lines including the operation mode information 9020 of the computer or the route. The operation mode is "Use"
Specify "standby", "test", "prepare", "stop", etc.

FIG. 10 shows a screen example of the overall system status screen. Reference numeral 10020 denotes a menu for performing operations related to this screen. In the area within 10010, the configuration of the entire system is displayed in a schematic diagram. 10030 indicates the function name of each computer, 10040 indicates the operation mode of each computer, and 10050 indicates the operation mode of each route. In addition, by instructing an area 10030 representing the function name of each computer by the user input unit, a server-specific system status screen 11010 is displayed.

FIG. 11 shows an example of a system status screen for each server. Reference numeral 11020 denotes a menu for performing operations related to this screen. An operation mode 11040 that each computer can take is displayed in an area 11030. Only the symbol representing the current operation mode is displayed in a different color from the symbols representing the other operation modes. When changing the operation mode, a symbol representing the operation mode to be changed is designated by the user input unit.

After the start of user input reception, if a transition to a different operation mode is instructed by the above procedure (step 8040), it is determined whether the transition from the current operation mode to the instructed operation mode is logically possible. Is checked (step 8050). If the transition to the instructed operation mode is possible, the status change enable / disable flag 4040 of all objects executed on the computer is checked (step 806).
0). If the mode transition is not logically possible, or if there is an object whose status change availability flag is "No", an error screen is displayed on the user output unit (step 8080).
The user's input is accepted again (step 8030).
If all the status change permission flags are "permitted", a screen for confirming the change is displayed to the user (step 8).
070), if the user confirms this (step 809)
0), the execution state of all objects executed on the computer is changed to an execution state adapted to the specified operation mode (step 8100). Thereafter, the operation mode storage table, the system overall status screen, and the server-specific system status screen are updated (step 8110), and the input of the user is received again (step 8030). Menu 1
When “end” is instructed from 0020 (step 812)
0), the entire system status screen is not displayed (step 8130), and the processing of the operation mode operation unit ends.

Subsequently, the group communication control unit 10 shown in FIG.
The processing example of 70 will be described with reference to the flowchart of FIG. First, a request from the client program 1080 is received (step 12010). If the request is a bind request (step 12020), an object identifier specified by the client program is obtained (step 12030). Next, the group configuration storage table 1150 is read to check whether a group having the same identifier as the identifier exists in the table (step 12040). If there is, a pseudo object reference is generated in which the group identifier is regarded as an object identifier (step 12050).
For example, an object reference is generated by regarding the group name as an object name. If it does not exist, the binding process is performed using the object identifier specified by the client program as it is, and an object reference is obtained (step 12060). Thereafter, the generated or obtained object reference is returned to the client program (step 1207).
0), the process of the group communication control unit ends. On the other hand, if the request is not a bind request but a method call request (step 12020), the identifier of the object is extracted from the object reference used by the client program for the call (step 12).
080). Next, the group configuration storage table 1150 is read to check whether a group having the same identifier as the identifier exists in the table (step 120).
90). If it does not exist, the method is called using the object reference used by the client program as it is (step 12110),
The call result is returned to the client program (step 12170), and the processing of the group communication control unit ends. If it exists, the group attribute storage table 11
After reading the operation mode and the operation mode storage table 1170, the operation mode of the calling computer is acquired, and a computer and a communication path suitable for the operation mode are selected (step 12100). For example, if the operation mode of the calling computer is “use”, the corresponding operation mode is “
A computer and a communication path which are “used” or “standby” are selected, and among the objects existing on the computer, an object to be actually called and a communication path are determined according to the execution control logic (step 12120). The method is called (step 12130).
At this time, if the call result satisfies the retransmission condition specified by the retransmission control logic (step 12140), the object and the communication path to be called again are determined according to the retransmission control logic (step 12150), and the method is called again (step 12130). ). If the retransmission condition is not satisfied, a call result to be returned to the client program is determined according to the reception control logic (step 12160), the call result is returned to the client program (step 12170), and the processing of the group communication control unit ends.

As described above, in this embodiment, a case has been described in which a management computer exists separately from the processing computer, and the operation mode operation unit and the operation mode storage table are arranged on the management computer. However, the present invention is not limited to this. For each processing computer, an operation mode operation unit for operating the operation mode of the own computer and an operation mode storage table for storing the operation mode of the own computer are arranged. Alternatively, a configuration that does not require a management computer may be adopted.

In this embodiment, the case where the objects are multiplexed has been described. However, the present invention is not limited to this, and the same method can be applied to multiplexing in units of processes, tasks, and the like.

In this embodiment, a case has been described in which the system administrator inputs the status change enable / disable flag for each object from the group configuration list screen. However, the present invention is not limited to this, and the execution program of each object may specify the flag information.

In this embodiment, a case has been described in which a remote method is called from an object which does not form a group to an object which forms a group. However, the present invention is not limited to this, and the same method can be applied to a remote method call between objects forming a group.

[0032]

As described above, according to the present invention, the system administrator dynamically determines the execution method of the process for calling the remote method in accordance with the operation mode information for each computer, so that the system administrator can Even in a distributed environment where many software modules and their copies are arranged in a complicated manner without changing the configuration or modifying the calling program, the execution of remote methods is controlled for each computer, and a specific computer or LAN is controlled. Can be easily separated from the execution system according to the purpose of program maintenance or the like.

Further, by determining whether or not to switch the execution state of the software module on the computer when the operation mode is changed according to the status change enable / disable flag, the execution state of a large number of software modules and their duplicates can be determined. It is possible to prevent the occurrence of inconsistency and incorrect processing of the entire system.

[Brief description of the drawings]

FIG. 1 is a schematic diagram illustrating a block configuration example of a program execution method and apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram schematically illustrating a conventional method of controlling execution of a remote method using a group.

FIG. 3 is a flowchart illustrating an operation example of a group information operation unit in FIG. 1;

FIG. 4 is a diagram showing a configuration example of a group configuration storage table in FIG. 1;

FIG. 5 is a diagram showing a screen example of a group configuration list screen displayed in step 3020 in FIG. 3;

FIG. 6 is a diagram illustrating a configuration example of a group attribute storage table in FIG. 1;

FIG. 7 is a diagram showing an example of a group attribute screen displayed in step 3090 in FIG. 3;

FIG. 8 is a flowchart illustrating an operation example of an operation mode operation unit in FIG. 1;

9 is a diagram showing a configuration example of an operation mode storage table in FIG.

FIG. 10 is displayed in step 8020 in FIG.
It is a figure showing the example of a screen of a system whole situation screen.

FIG. 11 is displayed in step 8040 in FIG.
It is a figure showing the example of a screen of a server individual situation screen.

FIG. 12 is a flowchart illustrating an operation example of a group communication control unit in FIG. 1;

[Explanation of symbols]

1010 to 1030 processing computer, 1040 management computer, 1050 network, 1060 communication processing unit, 1070 group communication control unit, 1080 client program, 1090 server program, 11
00: user input unit, 1110: user output unit, 112
0: group information operation unit, 1130: operation mode operation unit, 1140: storage unit, 1150: group configuration storage table, 1160: group attribute storage table, 117
0: Operation mode storage table.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G06F 15/177 676 G06F 15/177 676G (72) Inventor Junichi Toyouchi 1099 Ozenji Temple, Aso-ku, Kawasaki City, Kanagawa Prefecture F-term in Hitachi, Ltd. System Development Laboratory (reference) 5B045 BB12 BB24 BB28 BB48 BB49 GG09 JJ02 JJ08 JJ13 KK03 5B098 AA10 GC16 GD02 JJ08

Claims (7)

[Claims] 1. A software module for receiving a call of a remote method, which is a function executable via a network, and executing a process, forms a group including a set of the software module and a copy of the software module, In response to a request for calling a remote method for one group, at least when calling a remote method of one or more software modules in the group, input information on software modules forming the group for each group. A first processing step for inputting, for each group, a second processing step for inputting a selection logic at the time of calling a remote method, and for each computer, inputting operation mode information indicating an operation state of the computer Or a third to get And further dynamically determining a method of executing a process for calling a remote method of the software module based on the information specified by the first, second, and third steps. A program execution method, comprising: a fourth processing step. 2. A software module for receiving a call of a remote method which is a function executable through a network and executing a process, forming a group including the software module and a plurality of sets of the software module, In response to a request to call a remote method for one group, a program having the following four processing steps is stored on a storage device as a program for calling a remote method of at least one or more software modules in the group. A program execution device stored therein: a first unit for inputting information on software modules forming the group for each group;
A second processing step for inputting a selection logic at the time of calling a remote method for each group; and a third processing step for inputting or acquiring an operation mode indicating an operation state of a computer for each computer. Processing step; fourth processing step for dynamically determining an execution method of processing for calling a remote method of the software module based on the information specified by the first, second, and third processing steps. . 3. A second processing step for inputting selection logic at the time of calling a remote method for each group, comprising: a logic for determining a software module that receives a remote method call and executes a process. When calling a plurality of remote methods, at least one of a logic for determining an order of calling the remote methods and a logic for determining a network route for calling the remote methods is specified. A program execution method according to claim 1 or claim 2.
The program execution device according to the above. 4. The method according to claim 2, wherein the second processing step for inputting the selection logic at the time of calling the remote method for each group includes the step of calling the same remote method again after calling the remote method at least once. The logic for determining whether or not to execute the remote method again in the above logic, the logic for determining the software module that receives the remote method invocation and executes the processing. Logic for determining the order of invoking the remote methods when it is determined to call a plurality of remote methods, and determining the network path for invoking the remote methods when it is determined in the above logic that the remote method is to be called again Specify at least one of the following logics: Characterized Rukoto, program execution according to claim 1, wherein the method or claim 2
The program execution device according to the above. 5. A third method for inputting or acquiring operation mode information indicating an operation state of a computer for each computer.
3. The program execution method according to claim 1 or the program according to claim 2, wherein the processing step appropriately switches an execution state of a software module that executes processing on the computer according to the operation mode. Execution device. 6. A third method for inputting or acquiring operation mode information indicating an operation state of a computer for each computer.
2. The processing step according to claim 1, wherein when the execution state of the software module executing the processing on the computer cannot be appropriately switched according to the operation mode, the processing step is notified. A program execution method or the program execution device according to claim 2. 7. A software module for receiving a call of a remote method, which is a function executable via a network, and executing a process, forms a group including a set of the software module and a copy of the software module, In response to a request for calling a remote method for one group, at least when calling a remote method of one or more software modules in the group, input information on software modules forming the group for each group. A first processing step for inputting, for each group, a second processing step for inputting a selection logic at the time of calling a remote method, and for each software module, an operation mode indicating an execution state of the software module. A third processing step for inputting or acquiring information; and a method for executing a process for calling a remote method of a software module based on the information specified by the first, second, and third processing steps. A fourth processing step for dynamically determining the program.