JPH07105116A - Transaction relay method in 3-tier distributed processing system - Google Patents

Abstract

(57) [Abstract] [Purpose] To provide a relay system that does not require modification work of a server machine even if a new application program (AP) is added to a host computer [Configuration] Workstation (WS) 3 The message transfer from the application program (AP) 3a to the AP 1a of the host computer 1 is performed via the host relay process 10 provided in the server machine 2 regardless of the AP. Therefore, even if a new AP is added, the processing in the server machine does not change, and the correction work becomes unnecessary. This is because the interface module provided in the WS stores the message output from the AP of the WS and the destination of the host computer in the message area, the transmission destination is the host relay process, and the message is stored in the message area in the host relay process. This is dealt with by generating new message data to be transmitted to the host computer based on the destination.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a transaction relay system in a three-layer type distributed processing system.

[0002]

2. Description of the Related Art As an example of a distributed processing system in which one host computer (data held by it) is commonly used by a plurality of terminals, as shown in FIG. There is a three-layer type distributed processing system constructed by connecting each workstation 3 of a terminal via 2.

The host computer 1 stores a plurality of application programs (AP1, AP2) 1a for performing various processes. And
When the operator (user) processes a predetermined transaction from the workstation 3, the predetermined application program (APX) 3a installed in the workstation 3 and the application programs AP1 and AP2 to be processed are set. Although it is executed in conjunction with each other, in the case of the three-layer structure, the data is not directly transmitted and received between the workstation 3 and the host computer 1, but it is always executed via the server machine 2.

Therefore, the server machine 2 has a gateway process (“G / W”) for relaying messages.
2a). This G
The / W process 2a is an application program (AP1, AP2) 1 installed in the host computer 1.
In order to make one-to-one correspondence with a, AP1 ',
Two AP2 'will be prepared.

Further, in this kind of three-layer structure distributed processing system, an application program (AP) for performing predetermined transaction processing is also applied to the server machine 2.
W) 2a 'is installed, the message from the workstation 3 is sent to the application program (APW) 2a' installed in the server machine 2 without being sent to the host computer 1, and a predetermined process is performed there. There is also one that is returned from the server machine 2 to the workstation 3 after the operation. An application program (APZ) 3a 'that works in conjunction with the application program 2a' is installed in the workstation 3 as well.

Further, the server machine 2 is provided with a bulletin board 2b, and the bulletin board 2b is configured to store a program name, a transaction identifier, a machine name, etc. of an application program used in transaction processing as a table. . Specifically, the G / W process 2 mounted on the server machine 2
Information about the application program 2a 'processed by the server machine 2 or a is stored.

The host computer 1, the server machine 2 and the workstation 3 have a TP monitor 1
c, 2c, 3c are provided to have various functions required for transaction processing such as transaction scheduling, log collection, failure recovery, and exclusive control, so that messages can be sent and received between connected devices (applications). I have to.

[0008]

However, in the above-mentioned conventional distributed processing system, the host computer 1
Since the G / W process 2a is also installed in the server machine 2 in a one-to-one correspondence with the application program 1a installed in 1., the following problems occur.

That is, for example, as shown in FIG. 1, considering the processing when a new application program (AP3) 1a ″ is newly added to the host computer 1, the application program (AP3) 1
application program (AP
Y) Application program 1 not only for creating 3a "and mounting it on the workstation 3 but also for adding it
A G / W process (AP3 ′) 2a ″ corresponding to a ″ needs to be created on the server machine 2 and registered in the bulletin board 2b, which is complicated.

Moreover, since many server machines 2 are normally connected to one host computer 1, a G / W process must be added and a bulletin board must be updated for each of these server machines 2. However, the work becomes extremely complicated.

Further, since the power is turned off during the adding / updating process for the server machine 2, the workstation connected thereto cannot access the host computer 1. And
Normally, one host computer 1 is used for multiple jobs (sales,
Personnel, accounting, etc.) are processed, so that the added application program (AP3) 1a ″ is not used among operators who use a large number of workstations 3 connected to one server machine 2. In such a case, there arises a situation in which it is impossible to access and process the host computer 1 for maintenance irrelevant to the user, but it is impossible.

The present invention has been made in view of the above background, and its object is to add a new G / W process to a server machine even if a new application program is added to a host computer. The present invention provides a transaction relay method in a three-layer type distributed processing system that can improve the operability of the system by eliminating the need for registration on the bulletin board and the bulletin board.

[0013]

In order to achieve the above object, in the transaction relay method in the three-layer type distributed processing system according to the present invention, a server machine is interposed between a host computer and a terminal processing device. In the transaction relay method in the three-layer type distributed processing system described above, a host relay process is provided in the server machine to collectively transfer messages sent from the terminal processing device to the host computer. On the other hand, for the message to be transmitted to the host computer, the terminal processing device sets a host relay process in the destination storage area and stores the message in the destination storage area of the host computer which is the final transfer destination. An interface module is provided which sets the information to be transmitted and the message to be transmitted to generate transmission message data. Then, in the host relay process, the received transmission message data is analyzed, information regarding the destination of the host computer is set in the transmission destination storage area, and the message to be transferred is set in the message storage area. The new message data is generated and the message data is transferred to the host computer.

[0014]

When a message is sent from the terminal processing device to the application program installed in the host computer in order to perform the desired transaction processing, it is relayed (transferred) through the server machine.
The host relay process in the state where the information about the destination of the host computer as the final transfer destination is paired with the specific message generated by the interface module installed in the terminal processing device (the destination is the host relay process). Send to. Then, in the host relay process, message data having the destination as a new destination is generated and sent to the host computer. That is, regardless of the type of application program, the message data sent from the terminal processing device to the server machine has its destination as the host relay process, and the host relay process also creates a new message based on the received message. Message data is generated, the generation process is the same for any application program.

Therefore, if a new application program or the like is added to the host computer, the terminal processing device performs the necessary maintenance processing correspondingly, but the server machine does not need the maintenance processing. .

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of a transaction relay system in a three-layer distributed processing system according to the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 shows an example of a three-layer type distributed processing system which is the premise of the present invention. As shown in the figure, a plurality of server machines 2 are connected to one host computer 1 as in the conventional case,
Further, a plurality of terminals, workstations 3, are connected to the lower side of the server machine 2. In this example, the host computer 1 is installed in the head office,
A server machine 2 is arranged in each business office, and a workstation 3 is installed in each department of each business office. In this example, the host computer 1 and the server machine 2
Are connected via a private / public line 7, and the server machine 2 and the workstation 3 are connected via a LAN so that they can communicate with each other.

Then, as shown in FIG. 2, a predetermined application program (AP
1, AP2) 1a is installed on the workstation 3 side and a predetermined application
The program (APX) 3a is installed. In addition, an application program (AP
W) 2a 'is also mounted on the server machine 2 and the corresponding application program (APZ)
3a 'is also mounted on the workstation 3. further,
Information about it (APW) is also registered on the bulletin board 2b. The various data (messages) are sent and received by the PT monitors 1c, 2c, 3c provided respectively.
Done through. Such a configuration is similar to the conventional one.

In the present invention, the server machine 2 is
G / W corresponding to each application program (AP1, AP2) 1a installed in the host computer 1
Host relay process 10 instead of providing a process
Is provided so that messages sent from the workstation 3 to the host computer 1 can be collectively sent to the host by the host relay process 10 (transaction relay processing). Information regarding the host relay process 10 is also stored in the server machine 2.
Registered on the bulletin board 2b above.

Further, in the application program (APX) 3a on the workstation 3 side, the message such as the application program name of the message destination and the message are sent in the same manner as before (without changing the function much as usual). In order to output in a paired state (see FIG. 3 (A)), the workstation 3 is provided with an interface module 11 for performing predetermined data conversion. The specific configurations of the host relay process 10 and the interface module 11 are as follows.

First, the interface module 11
As shown in FIG. 4, the application program 3a of the workstation 3 requests the service desired to be linked, that is, the program name of the application program for executing the service and the location of the program (host / server). ) And a user message to be transferred (ST1).

Next, the received data is analyzed to determine the existence location of the program designated as the link destination (ST2). If the location is the host, as shown in FIG. 3B, the destination storage area (protocol
The header) S is set to "host relay process", the identifier (program name) of the application program of the host computer, which is the final destination of the transfer destination, is set to the head portion M1 of the message storage area M, and After that, the message to be transferred is set in the half M2. Then, it is delivered to the TP monitor 3c of the workstation 3 (ST3).

On the other hand, when it is determined to be the server in the process of step 2, the identifier of the application program of the server machine 2 is set in the transmission destination storage area S as shown in FIG. Area M
The message to be transmitted is set in the latter half M2 of the. In this case, the head portion M1 of the message storage area M
Is blank. Then, after that, it is delivered to the TP monitor 3c of the workstation 3 (ST3).

The message delivered to the TP monitor 3c of the workstation 3 is sent to the TP monitor 2c of the server machine 2 based on the destination stored in the transmission destination storage area (protocol header) S of the given data. Via the destination (host relay process 10 / application program 2a ').

On the other hand, the host relay process is performed by the TP monitor 2c of the server machine 2 as shown in FIG.
A message having a data structure as shown in (B) is received (ST5), and the destination host AP identifier stored in the head portion M1 of the message storage area M is referred to (ST6).

Then, as shown in FIG. 3 (D), the destination host AP identifier is set in the transmission destination storage area S '(protocol header) and the latter half M2 passed to the message storage area M'. A new message is generated by setting the message contents stored in (ST7). The generated message is transferred to the host computer 1 (ST8), and the process ends.

Although not specifically shown, the destination stored in the transmission destination storage area (protocol header) S of the message received by the TP monitor 3c of the server machine 2 is the application program of the server machine 2. In the case of 2a ', the program 2a' is handed over, predetermined processing is performed therein, and the processing result is returned to the workstation 3 side as necessary.

By the way, as indicated by the chain double-dashed line in FIG.
When a new application program (AP3) 1a ″ is added to the host computer 1, it suffices to install the application program (APY) 3a ″ which works in conjunction with the workstation 3 on the workstation 3 as shown in FIG. There is no need to perform any processing on the machine 2.

That is, since all messages transferred to the host computer 1 are performed via the host relay process of the server machine 2, the corresponding application program (APY) on the workstation 3 side.
3a ″ is mounted and it is connected to the interface module 11 and a predetermined message is transmitted.
It is sent to the host relay process 10 via the P monitors 3c and 2c.

On the other hand, the host relay process 10 generates a new message based on the received message, but since the generation process is the same for any application program, what kind of application program on the workstation 3 side will be generated? There is no need to modify the function even if there is a modification or change of the message, and the generated new message is sent to the TP monitor 1 of the host computer 1.
You can hand it over to c.

On the host computer 1 side, by registering the existence of the newly added application program 1a ″ in the TP monitor 1c, the predetermined message transferred to the TP monitor 1c can be transferred to the application program. It can be reliably delivered to 1a ″.

That is, regardless of whether such an application program is added, the method of data transmission / reception between the host computer 1 and the server machine 2 and the data transmission / reception between the workstation 3 and the server machine 2 are performed. Since the method and the processing function of the host relay process 10 do not change, maintenance processing for the server machine 2 becomes unnecessary.

As a result, if the host computer 1 is operating, the added application program 1
An operator who does not use a ″ uses the workstation of his / her own computer through the host relay process 10 of the server machine 2 even if the corresponding application is added in another workstation. It is possible to transfer a predetermined message to and perform a desired transaction processing.

In the illustrated example, one application program APX of the workstation 3 performs processing in cooperation with the two application programs AP1 and AP2 on the host computer 1, and the newly added application AP3 is processed. An example of mounting the corresponding application APY on the workstation 3 has been described, but the mode of cooperation between the programs is arbitrary, and for example, all programs are made to correspond one-to-one, or A
Various modes can be adopted such as linking PY with AP1 and AP3, linking APX with AP1 on the host computer as well as APW on the server machine. Furthermore, even if a new application program (AP3) is added to the host computer, if the added AP3 is linked only with the existing application program (APX) on the workstation 3, for example, Just by modifying the function of APX, a new application like the above-mentioned embodiment
No additional programs needed.

[0034]

As described above, in the transaction relay system in the three-layer type distributed processing system according to the present invention, all messages transferred to the host computer are sent through the host relay process provided in the server machine. Even if a new application program is added to the host computer, data transmission / reception between the server machine and the host computer and between the server machine and the terminal processing device does not change. Therefore, it is not necessary to add a new G / W process or perform maintenance processing such as registration work on the bulletin board. As a result, the operability of the system can be improved.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a three-layer type distributed processing system which is a premise of the present invention.

FIG. 2 is a diagram showing a preferred embodiment of a transaction relay system in a three-layer type distributed processing system according to the present invention.

FIG. 3 is a diagram showing an example of a data structure of a message to be communicated.

FIG. 4 is a flow chart diagram illustrating the function of an interface module.

FIG. 5 is a flowchart illustrating the function of a host relay process.

FIG. 6 is a diagram showing a conventional transaction relay system.

[Explanation of symbols]

 1 host computer 1a, 1a ', 1a "application program 2 server machine 3 workstation (terminal processing device) 3a, 3a', 3a" application program 10 host relay process 11 interface module

Claims (1)

[Claims] 1. A transaction relay system in a three-layer type distributed processing system constructed by interposing a server machine between a host computer and a terminal processing device, wherein the server machine transfers the terminal processing device to the host computer. A host relay process for collectively transferring messages sent to the host computer, the terminal processing device sets a host relay process in a destination storage area for a message to be transmitted to the host computer, and The storage area is provided with an interface module that sets information to be hidden in the destination of the host computer, which is the final transfer destination, and a message to be transferred to generate transmission message data, and the host relay process receives the information. The transmission message The data is analyzed and the information about the destination of the host computer is set in the transmission destination storage area, the message to be transferred is set in the message storage area to generate new message data, and the message data is stored in the host computer. A transaction relaying method in a three-layer type distributed processing system, which has a function of transferring data to another.