JPS6378266A - Distributed interaction processing device - Google Patents

Abstract

PURPOSE:To heighten efficiency of distributed processing between a work station and a host by providing a picture defining body that defines the content of processing and judging whether a program to be started is in its own node or in other node from the picture defining body. CONSTITUTION:Interaction processing environment 1 is made making a work station a base, and a picture defining body 10 that defines the content of processing to operate images and a program independently in each node is provided. A decision means 2 decides designation whether a program to be started is in the work station or in the host side (other node) using the defining body 10. When the program is in the host side, the program of the host side is started 3. At this time, an attribute, a variable extending over nodes, is provided, and the data are transferred 4 together with inputted data. When processing is completed, the variable extending over nodes is received 5 by the work station side together with outputted data. The variable data are updated, and at the same time, outputted 6 to a display 7 of the work station.

Description

DETAILED DESCRIPTION OF THE INVENTION [Overview] The present invention provides a distributed processing interaction device that can execute distributed processing between a workstation and a host computer without changing the program.

The present invention creates a workstation-based interactive processing environment. Then, it is specified whether the program to be started is in the workstation or on the host side (another node), and if it is on the host side, the program on the host side is started. At that time, an attribute called a variable that spans between nodes is provided, and that data is also transferred together with the input data. When the processing is completed, variables spanning between nodes are sent to the workstation along with the output data, and the variable data is updated and output on the workstation display.

In the system of the present invention, the screen and program in the dialog processing system can be made independent without changing the program, and furthermore, it is possible to develop and test programs only on the workstation side. There is.

[Industrial application field]

The present invention relates to the distributed processing function of a complex dialog processing system, and in particular, the screen format is changed by making the screen and program independent, and defining the relationship of the screen hierarchy and the specification of the startup program in the screen definition body. The present invention relates to an extremely flexible distributed dialog processing device that allows not only changes to the hierarchy or addition of screens, but also changes and additions to programs to each workstation.

[Conventional technology]

Distributed processing methods, which are based on the idea that input/output processing and simple calculations are performed at each workstation to reduce the communication load as much as possible, are gaining importance. In addition, the importance of distributed processing systems that use multiple processors to process work in parallel among the processors to increase speed has also increased.

In such a distributed processing system, processing technology for synchronization and order control between each processing step of work executed on each distributed node, technology for distributing data to be processed, communication procedure, etc. Communication technology for performing distributed processing, including protocols that specify this, is required to construct a distributed processing system.

There are several ways to distribute data, such as functional distribution, load distribution, and regional distribution.Recently, the most commonly used databases centrally manage data on a host computer, make it common, and allow various users to Also important is a distributed interactive processing system in which distributed workstations access a common database and use appropriate processing programs on the workstations.

Data distribution methods include the split allocation method and the copy method. The split allocation method is a method of dividing the entire data so that it does not overlap with each other and allocating each piece of data to different nodes. In this case, each piece of data is allocated to a specific node.

On the other hand, the copying method is a method suitable for a case where data that is referenced by a method that has the same data is evenly distributed among multiple nodes.

The latter copying method requires fairly complex update control algorithms to maintain the consistency of duplicate data between nodes.

On the other hand, the former method of split allocation has the problem of high communication overhead when data in other nodes is referenced very frequently, but Since it is possible to allocate nodes with a high value, there is an advantage that the response time of the entire distributed processing can be reduced.

Conventionally, in this type of distributed dialog processing system, there was no distributed dialog processing device for the host computer and workstation, and the processing was performed only by a large-sized computer.

Furthermore, if the program running on the entire distributed processing system is very large, there is no established method for passing variables between programs, so when changing the program processed on each workstation, it is necessary to change the entire program. It needed to be changed.

[Problem that the invention seeks to solve]

The present invention eliminates these conventional drawbacks, and provides distributed interaction processing that allows efficient distributed processing between workstations and hosts without changing programs, and allows efficient program development and testing on each workstation. Provide equipment.

[Means for solving problems]

The present invention creates a workstation-based interaction processing environment 1, as shown in FIG. The main control module 20 is provided with a determining means 2 that uses the screen definition body 10 to determine whether the program to be started is in the workstation or on the host side (another node). If there is, it has a program starting means 3 for starting a program on the host side. At this time, an attribute called a variable that extends between nodes is provided, and a transfer means 4 is provided to transfer the transfer information together with the input data.

Furthermore, when the processing is completed, a receiving means 5 receives the variable across nodes together with the output data as information on the workstation side, and an output data processing means 6 updates the variable data on the workstation side and outputs it on the display. have

[For production]

The present invention can change the screen format of each workstation,
To provide a distributed dialog processing device that does not affect other programs due to changes or additions to the screen hierarchy, and can efficiently exchange data between programs.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 2 is a system configuration diagram of the distributed dialog processing device of the present invention. Generally, there are a plurality of dialogue processing systems, but here the distributed processing method will be explained for two cases, systems A and B.

In complex dialogue processing systems, it is important to use screens and programs as separate as possible.

It is important to be able to specify keys within each screen, call up other screens when data is entered individually, or start a specific program. In the present invention,
Definitions such as processing movement between screens and starting of each program are defined in the screen definition body 10 in the disk to the system. This screen definition body 10 is used to move between screens and operate programs as independently as possible. That is, by defining the relationship between screen hierarchies and designations of startup programs in the screen definition body 10, programs can be made independent depending on the screen. The screen and the program are paired and combined on the control program in the main control module 20, so even if the screen or application program is changed or updated, there is no need to change the entire control program. Each program is processed in a structured manner.

Here, in order to prevent changes in screen format and changes or additions to the screen hierarchy from affecting other programs, data exchange between programs becomes a problem. For example, when inputting data from screen a and calling screen b, screen a may start a processing program, or may simply move the data to screen b without starting the program.

In the present invention, information regarding such program startup and step processing to the next screen is stored in the screen definition body 10 on the screen side.
is defined as Also, when a program is started, the startup program name is defined in the screen definition body. Generally, since there are a plurality of programs to be started, it is necessary to exchange data between the programs, but the present invention has an area for exchanging common data, that is, an inter-program common data area 30 as a mechanism for exchanging data.

This area 30 is located in a fixed area in the common memory, and there are cases where this area is referred to using a symbol, and cases where a named variable is registered in a table and data is retrieved using that name. In this way, by mainly using the screen definition body 10 and the inter-program common data area 30, the format and hierarchy can be changed without each screen being fixed, and screens can also be added.

The distributed dialogue processing device of the present invention is characterized in that programs can be changed or added to each screen element, and extremely flexible programs can be constructed.

Furthermore, even in a large-scale program, changes in each structure of the program do not require changing the entire program (the changes can be made for each screen element or each program, resulting in extremely high flexibility).

When a distributed processing function having such characteristics is incorporated into a system, the main control module 20 of system A follows the procedure shown in the figure. At the beginning of this procedure, the main control program is started using the specified command. Next, input data as shown in step 7 STI. Next, as shown in step ST2, the screen definition body 10 of the display screen is read out and displayed. The input data is edited using the read definition body, and the edited contents are stored in the buffer 201, as shown in step ST3. After editing, step ST4
As shown in FIG. 2, the process content determination, ie, the action determination, is performed based on the input data stored in the buffer 201. When specifying the next screen as case 1) according to the processing contents, step STI is executed by referring to the ID for screen recognition.
Return to processing. Further, if the processing content is processing by a program as in case 2), the program is loaded from the library using the program name as a key. That is, the program is started as shown in step ST5. When processing an application program in the local system, for example, when reading READ, the transfer data edited by each system, for example, system B, is passed through the inter-program common data area 30, and when the processing is completed, as shown in step ST6. Then, the output edit data of the write process is outputted via the buffer 202. Then, in step ST7, the output edited data is output.

When starting a program in step ST5 in system A, there may be a need to pass data to the next program at an intermediate stage from the time the program is called to the end of program processing. For example, data to be passed to other programs includes variable tables and symbols. The data area for transfer in this case is the inter-program common data area 30, which is a common area for each program. This area is generally provided in memory to input and output transfer data. Each system individually receives the handover information through this area and enters the output editing process in step ST6. In step ST6, a stream for display data is created and output. After the output editing process is completed, the data is output in step ST7.

Then, the process returns to step STI and data is input again.

Using such a mechanism, changes such as format changes and additions can be made independently for each screen or program. The present invention specifically constructs such a processing system inside each distributed personal computer to create a workstation-based interactive processing environment. A screen definition module 10 is provided inside each workstation so that it can be specified whether the program to be started is within that workstation or, for example, on the host side. Processing programs that cannot be performed within the workstation can be processed by a program on the host side of system B, for example.

At this time, the workstation node, other nodes,
That is, data transferred from a program that spans the host side is provided with an attribute as a variable that spans between nodes, and that data is also transferred together with input data between nodes. For example, when the process is completed on the host side, variables that span between nodes are sent to the workstation side via the inter-program common data area 30 together with the output data received via the transfer module 40. Updates the variable data and outputs it to the workstation display.

In order to construct such a distributed dialogue processing device, in step ST4, processing contents are determined using input data. In this judgment, when moving to case 2) in which processing is performed by a program, the program name is used as a key to determine whether the program is executed inside the workstation of system A or on another node, for example, system B. Determine whether the process is to be processed on the host side. That is, the determination is made using the program name as a key and the definition in the screen definition body 10, but it is important that the determination information is included in the definition body itself. As a result of this determination, if the program is not a program that should be executed in its own system, the input variable data along with the individual delivery information is sent to the destination system individually via the inter-program common data area 30 and the transmission data buffer 50. Transfer to.

Transfer module 4 assembles transmission data and other transmission information.
Transferred via 0. In other words, using 10 of the other party's system, the name of the program to be executed on the other party's system as well as the transfer data and input (collection data) are sent to the specified system of the other party.The data received by the specified system is the transfer data. and the editing data and program name.

Along with the transfer information, information regarding the reception and decomposition of data is also reflected in the transfer data area, that is, the inter-program common data area 60 in the system B, that is, the partner system. The specified application program 70 is then started. For this program, READ
After /WRITE is executed, the transfer data and result data are sent to the system A side. The main control module 8o of system B controls these data reception and disassembly, program activation, and data assembly and transmission. On the system A side, the sent output data is edited and outputted, and at the same time, the delivered data is updated to make it the same value between systems.

In this way, with the present invention, a specified program within one large program can be moved from one node to another node and executed, and from the user's perspective, it is possible to detect which program is running on which system. Fully distributed processing is performed without Further, according to the present invention, a program created by a user can be assigned to any node, resulting in an extremely efficient and highly flexible system. The definition of which program is assigned to which node can be incorporated into the screen definition body 10. In the system of the present invention, there is no need to change the entire program; what needs to be changed is the definition of which program is assigned to which system. What the control program executes is management of which system is executing each program. At this time, it is generally not possible to define which system a program defined by an arbitrary user should be assigned to, but it may be a program using TSS, a search program, or a bank's online system that consists of a large number of screens. etc. is a program system that allows such definitions.

When creating such a large program, if the method of the present invention is followed, the interface can be made common, and the interface can be easily constructed.

What is important in the system of the present invention described above is the exchange data area between the program common data areas 30 and 60. The configuration of this transfer data area may differ slightly depending on the system configuration. for example,
There are two methods: one is to access the program directly using the program's symbols, and the other is to store the correspondence between names and data as variables in a table and then draw the table. In either case, common data area 30 between programs
and 60 are common areas for exchange data, and input/output data and variables can be freely exchanged between programs via this exchange data area. In particular, if the transfer data area is made into a variable, the amount of data to be transferred can generally be reduced. Furthermore, when variables in the previous process and variables in the next process are different, a method of transferring only the different parts can be effectively used.

There are two types of variables: global variables and local variables. Global variables are variables that remain even after a program has finished processing, and local variables are valid only while the program is processing. Network variables are added as an extension of local variables, and can be used to specify the range of variables to be sent. There are various ways to use a network variable, but one example is to use it as a variable for a network to transfer a variable whose value has changed during program processing.

As explained above, the distributed dialogue processing device of the present invention creates a workstation-based dialogue processing environment,
Specify whether the program to be started is on the workstation or on the host side (another node), and if so, start the program on the host side, set an attribute called a variable that spans nodes, and When the process of transferring data together with input data is completed, the system sends the output data as well as the variables spanning between nodes to the workstation side, updates the variable data of the workstation, and outputs it to the workstation display.

〔Effect of the invention〕

In the system of the present invention, the screen and program in the dialog processing system can be made independent without changing the program, and furthermore, it is possible to develop and test programs only on the workstation side. There is.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a distributed processing system according to the distributed dialog processing device of the present invention, and FIG. 2 is a block diagram of a distributed processing system according to the distributed dialog processing device of the present invention. 1... Dialogue processing environment, 2... Judgment means, 3... Program starting means, 4... Transfer means (delivery information), 5... Receiving means (delivery information), 6... Output Data processing means, 7... Display, 8... Processing program, 10... Screen definition, 20... Main control module, 30... Inter-program common data area, 40... Transfer module, 50... Transmission data buffer, 60... Common data area between programs, 70... Application program, 80... Main control module.

Claims (1)

[Claims] In a distributed processing system where each node is a workstation or a host computer, a screen definition body (10) that defines processing contents for independently operating screens and programs in each node, and a main control module. a determining means (2) for determining whether to perform inter-screen processing or program processing after inputting the screen definition body (10) and data and editing the input data; (2) determines from the screen definition body (10) whether the program to be started when performing program processing is on the own node or on another node;
A program starting means (3) that starts a program, and when the program starting means (3) determines that the program to be started is on another node side, there is a program starting means (3) that starts a program, and when the program starting means (3) determines that the program to be started is on another node side, there is a Transfer means (
4), and after starting the program and receiving a notification of completion of processing from the node that processed the program, receiving means (5) receives the transfer information between the sending and receiving nodes together with the output data from the node that processed the program. ); and an output data processing means (6) for updating the delivery information, editing the received output data, and outputting the received output data.