JPH02230449A - Matrix control system for communication control program - Google Patents

Abstract

PURPOSE:To reduce the developing manhour of a program by preparing a matrix analyzer processing part which analyzes a matrix table via a processing logic which is not dependent on the communication control means and then calls a processing module. CONSTITUTION:A part related directly to a matrix includes a matrix analyzer processing part 13 and a matrix table 17 so as to eliminate the dependence on the protocol procedure, etc. These part 13 and table 17 are completely independent of various processing modules 22a, 22b... which are dependent on the control procedure. Thus it is possible to omit a state checking process, etc., in the processing module and therefore to perform the integrated and universal control of the matrix which does not depend on each communication control procedure. Then the developing manhour of a program is reduced and the overall program constitution is facilitated.

Description

[Detailed Description of the Invention] [Summary] Regarding the matrix control method of a line control program that performs line control with state transitions according to various control procedures in a computer system, a unified and general-purpose method that does not depend on individual line control procedures is provided. Provides a means to control the matrix. The purpose is to reduce program development man-hours and simplify program configuration, depending on the current situation. A control procedure-dependent section consisting of multiple processing modules that performs processes related to state transitions that depend on control procedures, and a module that stores information about state transitions in a tabular form in a predetermined format that does not depend on line control procedures. The L-Risk table and the 1-Matrix table are analyzed using predetermined processing logic that does not depend on the line control procedure. and a matrix analyzer processing section that performs a process of calling a processing module in the control procedure dependent section.

[Industrial application field]

The present invention relates to a matrix control method for a line control program in a computer system that performs line control with state transitions according to various procedures.

In a computer system that communicates with terminals or other systems via lines, various types of line control are required depending on the communication partner's communication status. In one system, multiple types of line control may be required, or control procedures may need to be changed, so it is desirable to be able to deal with these flexibly.

[Conventional technology]

Line control. In particular, programs that control multiple sessions (buses) are controlled according to a matrix (state transition table) that does not indicate the processing to be performed and the state after processing, due to the relationship between states and events. There are many cases. For this reason, developers of line control programs write down a matrix showing state transitions on paper and refer to it while developing their programs.

71. By creating a table of risk information, storing it in a computer, and using that table. Although it has been considered to automatically allocate processing by determining the event and status that have occurred, no concrete method has been established for how to create a table from the information in the matrix. Furthermore, since there is no general-purpose logic for analyzing the matrix table recorded in the computer and distributing the processing, program developers have to create case-high-case logic according to various control procedures. I had to deal with it.

[Problem to be solved by the invention]

In the conventional method, the 71 RIX table and the processing unit that analyzes it are created using separate table configurations and processing logic for each program depending on the control procedure or for each program developer. The problem was that the workload was heavy and the number of development man-hours increased. Another problem was that the developed program was difficult to understand.

The present invention aims to solve the above problems and provides a means for performing unified and general-purpose matrix control that does not depend on individual line control procedures, thereby reducing the number of program development steps and simplifying the overall program configuration. It aims to make it possible.

[Means to solve the problem]

FIG. 1 shows an example of the configuration of the present invention.

In FIG. 1, 10 is a processing unit consisting of a CPU and memory, Ila and Ilb are application processing units that perform data communication and process data, and 12 is a line that controls communication via the line according to protocols and various control procedures. control unit,
13 is a 71 risk analyzer processing unit, and 14 is a mode check analyzer. 15 is type Chesoku analyzer, 1
6 is a processing state check analyzer, 17 is a table 71 for storing information regarding state transitions in the form of a table, 18 is a mode check table, 1 is a type check table, 20 is a processing state tW check table, 21
The control procedure dependent units 22a, 22b and 22c represent processing modules.

The line control unit 12 is composed of programs and tables that perform line control with state transitions. The program of the line control section 12 consists of a control procedure dependent section 21 that depends on various control procedures, a matrix analyzer processing section 13, etc. that does not depend on control procedures. Further, a matrix table 17 is provided as a table handled by the line control unit 12. This table is referred to by the matrix analyzer processing unit 13, and the format of the table is as follows.
It is designed to be independent of control procedures.

In the example shown in FIG. 1, the matrix table 17 is hierarchically structured by a mode check table 18, a type check table 19, and a processing state check table 20.

The mode check table 18 has a pointer to the type check table 19 corresponding to the mode of processing, ``request'' or ``completion.''

Type check table 19 shows “request” and “completed”
For each event, there is a link to the processing status check table 20 corresponding to the event type information.

The process status check table 20 has information on the process to be performed when an event occurs and the status after the process for the current status. In addition, if it is necessary to control the state transition by another matrix, it has a pointer to another matrix table for controlling that matrix.

[MaI Risk Analyzer Processing Unit] 3. When an event occurs, the mode check table 18 is first searched by the mode check analyzer 14. The address of the corresponding type check table 19 is obtained and the type check analyzer 15 is activated. The type of Chesoku Analyzer 15 is. The type checking table 19 is searched, the address of the processing status checking table 20 that handles the type of the event is obtained, and the processing status checking table 20 is activated. The processing status check analyzer 1G searches the processing status check table 2o. Determine the processing for the current state and the state after processing. The processing to be carried out is. The control procedure dependent unit 21 includes a processing module 22a. 22b, . . . are individually registered, and when necessary, one of them is called and activated.

[Effect]

In the present invention, the parts directly related to 71.Risk are excluded so as not to depend on 2Protocols and procedures. It is composed of the matrix analyzer processing section 13 and 71 and the risk stage 17, and various processing modules 22a. 22b, . . . are made completely independent. This eliminates the need for processing such as checking the status within the processing module, making it possible to simplify the program.

The format of the matrix table 17, that is, the field configuration and attributes of the matrix table 17, is predetermined so as not to depend on the line control procedure, and the data is in the same format no matter what matrix. Therefore, the MARIX analyzer processing unit 13, which refers to the data and controls the state transition, can be used for general purposes.

〔Example〕

Fig. 2 is an example of an application system for an embodiment of the present invention, Fig. 3 is a session control system state transition diagram for explaining an embodiment of the present invention, and Fig. 4 is an example of an application system for an embodiment of the present invention. An example of a session controller 1/role matrix for explanation, FIG. 5 is an example of the table configuration of a mode check table according to an embodiment of the invention, and FIG. 6 is a table configuration example of a mode check table according to an embodiment of the present invention. An example of the table configuration of a table, FIG. 7 is an example of a matrix table according to an embodiment of the present invention, FIG. 8 is a processing flow of a mode chenok analyzer according to an embodiment of the present invention, and FIG. FIG. 10, a processing flow of a processing state check analyzer according to an embodiment, shows an example of an application form of the present invention.

An example in which the present invention is applied to a system as shown in FIG. 2 will be described below. In FIG. 2, numeral 30 represents an application program, 31 represents an independent work control section, 32 represents a control dryer, 33 represents a device dryer, and 34 represents an adapter.

This embodiment relates to the software of the control dryer 32 shown in FIG. According to the control driver 32, session control 1, roll system state transition diagram and transmission/reception system state transition diagram. Controls multiple sessions (paths).

Session control system state transition diagram. It is as shown in Figure 3.

In the initial state 40, when a CONNECT request is received, the state changes to the session establishment processing state 41. Upon completion of CONNECT, the state 41 in which one session is being established is transited to the session establishment state 42. Session established state 42, SEND request. Data is sent and received in response to a RECEIVE request. When a DISC request event occurs in the session establishment processing state 41 or session establishment state 42, the state changes to the session release processing state 43, and then returns to the initial state 40 upon completion of DISC.

The state transition diagram shown in Figure 3. The session control matrix shown in FIG. 4 is expressed in a tabular matrix.

Figure 4 shows that, for example, in the initial state, CONNECT
When a request event occurs. This shows that processing A is performed and the state shifts to the session establishment processing state (2).

If a CONNECT request event occurs in the session establishment state, error processing is performed and the current state (■) is maintained.

When a Send/RECEIVB request or SEND/RECEIVE completion event occurs in the session established state,
The process shifts to control using a transmission/reception matrix (not shown).

In the present invention. The information in the matrix shown in Figure 4 is
(1) Check the mode, (2) Check the request type,
(Such as changing the three processing states. Divided by processing stage, in a predetermined format as shown in Figures 5 and 6,
It is stored in a table.

FIG. 5 shows an example of the table structure of the mote check table 18 shown in FIG. 1, which consists of a header section and a matrix section. The number of search counters in the henoder section is the number of table entries in the matrix section, and in this example, the value is "2". Reserve represents unused space. In the next processing analyzer field, a pointer to the mode check analyzer 14 shown in FIG. 1 is set.

The determination unit is set with a load type of "request" or "completion". A pointer to the request or completion type check table 19 is set in the next table pointer.

Although the table structure of the type check table 19 is omitted from illustration. The configuration is almost the same as the configuration of the mode check table 18 shown in No. 5M. However, in the judgment section, CONNE
Event types such as CT and DISC are stored. A pointer to the processing status check table 20 is set in the next table pointer.

The table configuration of the processing status check table 20 shown in FIG. 1 is as shown in FIG. 6.

This is also the configuration of the header section, 5 mode check table 18
It is similar to However, the processing analyzer is a pointer to the processing state check analyzer 16 shown in FIG.

The matrix part consists of a judgment part that indicates the current state, processing module information that points to the processing module when there is processing to be performed in conjunction with state transition, and address information of the new state or next table after processing. It consists of several table entries.

For example, in the processing status check table 20 that is referred to when there is a processing request of "mode request" or [type 1 SEND J], the number of search counters is "2".

At this time, for the first file entry t,
"Initial" state in the judgment section, "error processing" in the processing module. Information of "no transition" is set in the state after processing. In the second table entry, "session established" is set in the determination unit, Null (all O) is set in the processing module, and a pointer to a transmission/reception system matrix (not shown) is set as the state after processing. If analysis by the next analyzer is required, as in this second en-1, you can add Null to the processing module area. represent it.

"Mode request", "Type - SENDJ processing dependence is
This does not occur when the session is being established/released. Since it is impossible on the matrix,
To save memory, no table expansion is performed for that state.

With the above table structure, an example of the session control matrix shown in Figure 4 is created as a table.
Not shown in FIG. 7 is a 71-risk table. In other words.
The matrix table shown in Figure 7 is. Not shown in Figure 471
・Has exactly the same meaning as Rikusu. It is in a format that computers can recognize.

Note that illustration of the header portion of each table shown in FIG. 7 is omitted. Also, to make the diagram easier to understand, mode check table 18 is shown here. Judgment section (
It is represented by a pair of request/completion) and next table pointer. NEXT is pointer information and indicates that the next table needs to be searched. Similarly. Request type check table 19a and completion type check table]. 9b is represented by a pair of determination unit (by type) and next table pointer. Processing status check table 2
0 is represented by a set of a determination unit (current state), a processing model, and a post-processing state (or next table pointer).

The processing status check table 20 shown in FIG. 7 displays entries for all statuses, but as mentioned above,
Table expansion is not necessary for items that cannot exist on the matrix.

The mode check analyzer 14 shown in FIG. 1 performs processing as shown in FIG. Since this process depends only on the format of the mode check table 18 and does not depend on the control procedure, etc., it can be applied not only to the session control 1/roll system.

Hereinafter, the process will be explained according to the processes ① to [phase] shown in FIG.

■ Set the address of the mode check table 18 in the variable schptr (used as a search pointer).

(2) Set the value of the search counter number in the mode check table 18 to the variable schcnt (used as a search counter). In this example, it is "2".

■ Determine whether the variable schptr has become O.

If it becomes 0, perform error handling using process [phase].

(2) Compare the contents of the determination section pointed to by the variable schptr and the mode of the processing request (request or completion).

■ If they do not match, proceed to process ■; if they match, branch to process ■.

■ Increment the variable schptr, and set the next value in the mode check table 18 to 1.
·do.

■ Decrement (=1) the variable schcnt. Thereafter, control is returned to process (2) and the process is repeated in the same manner.

■ When the modes match. Set the next table pointer in the variable nextptr.

(2) Call the next analyzer, that is, the type Chenok analyzer 15 shown in FIG.

[Phase] Based on the judgment of processing ■. If the variable schcnt = 0 is detected, it means that there was an impossible processing request, so 1 It is determined that there is a program error in the self program or the program, and the system down function is called by specifying the predetermined down code. .

The above is the processing by the mode check analyzer 14, but the process by the type check analyzer 15 is exactly the same. The type check analyzer 15 determines a processing status check table 20 corresponding to the type and activates a processing status check analyzer 16 that analyzes the table.

The processing state check analyzer 16 shown in FIG.
Perform the processing shown in the figure. This process depends only on the format of the process status check table 20. Since it does not depend on control procedures, etc., it is not limited to session control 1/roll systems. It is possible to apply. The following is the process shown in Figure 9.
0 will be explained.

(2) Set the address of the processing state check table 20 to be searched in the variable schptr.

■ The variable schcnt contains the processing status check table 2.
Set the search counter number value to 0.

■ Determine whether the variable schptr has become O.

If it becomes 0, error handling is performed using process 0.

(2) Compare the contents of the determination section pointed to by the variable schptr with the state at the time of the processing request.

■ If they do not match, proceed to process ■; if they match, proceed to process ■.

(2) Increment the variable schptr to point to the next en1 in the processing status check table 20;

■ Decrement (-1) the variable schcnt. Thereafter, control is returned to process (2) and the process is repeated in the same manner.

■ If the state of the determination unit matches the state at the time of the processing request, determine whether a processing module exists. If the processing module is registered, the process moves to process (2).

■ If the processing module is Null, there are additional tables to search. So the variable nex
Set the next table pointer in tptr.

[Phase] Calls the next analyzer. In this example, control is shifted from the session control system 71/rix table to control according to the transmission/reception system matrix table.

■ If a processing module is registered in the corresponding entry in the processing status check table 20, call that processing module. This prevents actually necessary processing according to the protocol, procedure, etc. from being performed.

@ When the processing module returns, it obtains post-processing state information from the corresponding entry in the processing state check table 20 and sets it as the current state.

0 If the variable scbcnt-0 is detected in the judgment of process ■, it means that an impossible process request has been made.
It determines that it is a program error in its own program or another program, specifies a predetermined down code, and calls the system down function.

Although the matrix control has been explained with a focus on the session control system, similar 71-risk control can be performed on the transmission/reception system and the like using the same processing logic.

FIG. 10(a) shows an example of an application form when developing a program to control different protocols.

The AC procedure control program 50 and the BC procedure control program 51 are based on different protocols. The method of state transition is different, but here is a state transition diagram. By expressing it in the form of a predetermined matrix table 17 shown in FIG. The micro-I-risk analyzer processing section 13 can be shared.

Regarding differences in processing contents, the individual processing modules 22a. It is absorbed by...

FIG. 10 (exit) shows an example of an application form when programs with different layers to be controlled by the same protocol are developed.

The OS>S layer control program 52 and the OSI-T layer control program 53 have different processing contents because they are in different layers, but their respective state transition diagrams are in the format of the predetermined matrix table 17 shown in FIG. By expressing this as follows, control related to state transition can be performed using a unified processing logic by the matrix analyzer processing unit 13. In this case as well, differences in processing contents are determined by the individual processing modules 22a. It can be absorbed by...

〔Effect of the invention〕

As explained above, the present invention has the following effects6 (1) By creating a 71 risk table from a state transition diagram according to a predetermined format, it is possible to check the state and start necessary processes. etc. can be done automatically, and only the processing modules that depend on protocols and procedures need to be developed individually, reducing the number of program development man-hours (
(2) Since state transitions are controlled by the same matrix control, the program becomes easier to understand.

(3) For example, session controller/J-rule system and transmitting/receiving system. This is possible even when matrices are multiplexed.

(4) It becomes possible to flexibly modify the program in response to matrix changes such as additions and deletions.

In other words, since all state determinations are performed uniformly in the risk analyzer processing section 71, all that is required is to modify the contents of the matrix table. It can be handled flexibly with almost no impact on the actual processing unit, improving maintainability.

[Brief explanation of the drawing]

Figure 1 shows a configuration example of the present invention. Figure 2 shows an example of an application system for an embodiment of the present invention. Figure 3 shows a session controller system state transition for explaining an embodiment of the present invention. 4 is an example of a session control [call system matrix] for explaining an embodiment of the present invention, and FIG. 5 is an example of a table configuration of a mode check table according to an embodiment of the tree invention. FIG. 6 is an example of a table configuration of a processing status check table according to an embodiment of the tree invention. FIG. 7 is an example of a multi-liktable table according to an embodiment of the present invention. FIG. 9 shows a processing flow of a mode check analyzer according to an embodiment of the present invention, and FIG. 10 shows an example of an application form of the present invention. In the figure, 10 is a processing device. 1.1a. ]. 1. b is an application processing unit, 12 is a line control unit, 13 is a matrix analyzer processing unit, 14 is a mode check analyzer, 15 is a type check analyzer, 16 is a processing status check analyzer, 17 is a matrix analyzer, 18 is a mode check table, J9 is a type check table, 20 is a processing state check table, 21 is a control procedure dependent part, 22
a to 22C represent processing modules.

Claims (1)

[Claims] A matrix control method for a line control program that performs line control with state transitions depending on various control procedures, the method comprising: performing processing related to state transitions depending on the control procedure according to the current state. A control procedure dependent unit (21) consisting of a plurality of processing modules; a matrix table (17) that stores information regarding state transitions in a tabular form in a predetermined format that does not depend on line control procedures; and the matrix table described above. (17) using a predetermined processing logic that does not depend on the line control procedure, and a matrix analyzer processing section (13) that performs processing to call the processing module in the control procedure dependent section (21). A matrix control method for line control programs characterized by: