JPH05158892A - Data management system for distribution processing - Google Patents

Abstract

PURPOSE:To permit multiple processes to share data efficiently and to reduce the processing for the consistency control for data reading and writing of each process even when a process occupies data for a long period. CONSTITUTION:The system is provided with a server 6 managing data and performing the exclusive control of process other than process '1' for data and multiple processes 1 receiving data from the server 6. The server 6 has a list corresponding to the name of data for the exclusive control and the place of storing data. The process 1 has a list indicating the changed state of data corresponding to the name of the data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data management method in distributed processing, and more particularly, to maintain data consistency when a plurality of processes operating in a distributed environment operate by sharing the same data. The present invention relates to a data management method in distributed processing. This data consistency function is an environment where multiple processes operate and share the same data,
For example, when a plurality of computers are connected by a network and a program operating simultaneously on the network updates one data at the same time, the program is used by arbitrating the update by the plurality of processes.

[0002]

2. Description of the Related Art In a conventional database or the like, there is a method of using a lock mechanism as a method of controlling reference or update by a process which is a main body of a plurality of processes for the same data. This locking mechanism prohibits other processes from referencing or updating the data when one process references or updates the data, and maintains the consistency of the data when it is rewritten by the update or the like. ..

[0003]

However, the above locking mechanism is such that if one process prohibits the reference or update of data by another process for a very long time and occupies the data, the other process waits. Therefore, there is a problem that the processing efficiency of the entire process is reduced. Further, when each process refers to or updates data, a mechanism for grasping the status of referencing or updating data of another process is added, which slows down the process of referring to or updating data.

[0004]

In a data management system including a plurality of processes that perform reference or update processing and a server that guarantees data consistency when the processes simultaneously reference or update the same data, an inter-process Of the exclusive control of the process for performing data communication and server communication, performing data reference and update processing and managing data, and the second process other than the first process currently performing reference and update processing. And a server that stores information processed by the first process.

Further, in a data management system in distributed processing, a process communication unit for communicating with a plurality of processes, a data storage unit for storing data in units accessed by the processes, and a data storage unit for storing data. A server including a data management unit that manages information, a data processing unit that refers to or updates data, and a data management unit that manages the data processed by the data processing unit and the storage location of the data in the data storage unit. And a process including a server communication unit that communicates with the server and a process communication unit that communicates with each other. Also, in the data management method in distributed processing, when storing data in a sequential file for each unit accessed by a process, the data name with a unique name for each unit and the data name in the sequential file are stored. A data storage unit of the server including location information, a data management unit of the server including a first relation table showing a correspondence relation between a data name and a storage location of the data name on the data storage unit, a data name, and a data name And a data management unit of the process including a second relation table of information indicating whether or not the data is changed due to the processing of the process.

[0006]

The data management system of the present invention comprises a server that manages data processed by one process and performs exclusive control on the data, and a plurality of processes that communicate with the server and receive data from the server. , The server has a table having a correspondence relationship between data names and data storage locations for exclusive control with respect to other processes,
In addition, the process has a data name, a data storage location, and a relational table showing the state of whether the data has been changed by the update. By referring to the relational table of both the process and the server, the same process can be obtained. Keep data consistent with references or updates from multiple processes to the data. When each process writes data, each process adds the updated data to the data before updating. The process also stores the storage location of the updated data. When data is kept consistent between processes after multiple updates are performed in different processes, information on the storage location of the data name data and the storage location of the data possessed by each process is used between the processes and between the process and the server. Exchange between the two and keep the data consistent between them.

[0007]

1 is a block diagram showing the structure of the present invention. The data management system of the present invention is a server 6 having a function of communicating with a plurality of processes 1 and a function of managing data.
And a process 1 having a communication function with the server 6 and other processes, a data management function, and a data processing function.

The hardware structure of this system is as follows. FIG. 2 is a block diagram showing the hardware configuration of the present invention. The process 1 includes a central processing unit, a main memory 21, and a communication device 22, and the server 6 is composed of a central processing unit and a main memory 27, a communication device 28, and an external storage control device 29, and stores the stored information from the server 6. External storage device 24 is connected to the server 6,
The process 1 and the server 6 are connected by the communication medium 27.

FIG. 3 is a block diagram showing details of FIG. 1 of the present invention. First, the process 1 includes a data processing unit 2 that performs processing such as data reference and update, a data management unit 3 that manages a data name, whether or not the data has been changed, and a management location of the data on the server 6 as a relation table. A server communication unit 4 having a function of communicating with the server 6 and an inter-process communication unit 5 for communicating with other processes are included. Next, the server 6 sequentially processes data as a process communication unit 7 having a function of communicating with a plurality of processes, a data management unit 8 that manages a relationship between data names and data storage locations as a table, and an external storage device. It has a data storage unit 9 for storing it in a file.

FIG. 4 is a diagram showing in detail the data storage unit 9 of the server of the present invention. The data storage unit 9 is a sequential file, and as an example, the storage area is 1 to 100.
Of 100 areas, each stored data is stored in a unit accessed by the process 1, and a name is given to each data unit. This name is called a data name, the data name “A” is stored in the storage location “1”, the data name “B” is stored in the storage location “2”, and the data name “Z” is stored in the storage location “100”. Store.

FIG. 5 shows in detail the data management unit 8 of the server of the present invention. The data management unit 8 of the server manages, as a table, the correspondence between the data name in the data storage unit 9 in FIG. 4 and the storage location of the data on the sequential file corresponding to the data name.

FIG. 6 shows in detail the data manager of the process of the invention. The data management unit 3 of the process 1 manages a data name, a storage location of the data in the data storage unit 9 on the server side, and presence / absence of a change indicating whether or not the data corresponding to the data name has been changed due to update processing or the like.

Next, a data management procedure in the process 1 and the server 6 having the above structure will be described. FIG. 7 shows an example of changing the data of the process of the first embodiment of the present invention. In the first embodiment, the name of process 1 is "TOO".
Let L-A "and" TOOL-B ", and the name of the server 6 be" DATA-SERVER. "FIG. Data management unit, (C) is server DATA
-SERVER, (D) shows the data storage state of the data storage unit 9. TOOL-A and TOOL-B shown in FIG.
And the contents of the server DATA-SERVER are data before updating. In the example of the figure, data “A”,
"B" is process TOOL-A and process TOOL-
It is shared by B and it is necessary to make data consistent with them. When the process TOOL-A changes the data of the data name "A", the process TOOL-A
Then, register that the data A has changed in the table, and further,
Delete the location data that was previously stored.

The results are shown in FIG. Process TOOL
The data in the “storage location” corresponding to the data name “A” of −A is deleted, and the “presence or absence of change” becomes “changed”. The changed data name “A” and the changed data are sent to the server DATA-SERVER, and the data storage unit 9 of the server 6 stores this information.
The name of the data changed by the process TOOL-A and the changed data are sent to the server DATA-SERVER,
Process TOOL-A is a server DATA-SERVER
Receives new "storage location" information for changed data from. As a result, the data management unit 3 of the processes TOOL-A and TOOL-B, the data management unit 8 of the server DATA-SERVER, and the data storage unit 9 have the results shown in FIG.

The "storage location" of the data "A" of the server DATA-SERVER becomes the storage location "3" of the data storage unit 9, and the new data becomes a new storage location.
The data name “A” is stored in the third storage position of the data storage unit 9.
Is stored. However, the process TOOL-A and TOOL
Between LB, the server DATA-SERVER
Since only the process TOOL-A receives the change information from, the data consistency is lost with respect to the process TOOL-B.

The process of keeping the data consistent between the processes TOOL-A and TOOL-B is performed in the following procedure. First, the process TOOL-A informs the process TOOL-B of the data change by using the interprocess communication unit 5 about the data name of the changed data and the new storage location of the changed information. Next, the process TOOL-
B modifies the process TOOL-B table based on the information provided by TOOL-A.

FIG. 10 is a diagram showing an example of maintaining data consistency between processes according to an embodiment of the present invention. The process TOOL-B is notified that the storage location of the data “A” of the process TOOL-A has been changed to “3”. Process TOOL-B has the corresponding data “A” in its own table.
The storage location of is changed from "1" to "3". Further, in the process TOOL-B, when the changed data is needed, the data name and the storage location of the table of the data management unit are designated, and the server is requested, whereby the changed data management unit 3 Data from server D based on table
The data is read from the ATA-SERVER data storage unit 9. Thereby, the data consistency between the processes TOOL-A and TOOL-B is maintained.

In the process of keeping the data consistent, depending on the contents of the process, it may be necessary to immediately reflect the change in the data, or the process may reflect the data when it is needed. is there. The present invention is
Adjustment can be made by changing the time when the second process that receives the information sent from the first process that has made the change reflects the information in the table for data management.

[0019]

As described above, according to the present invention, the process of referring to or updating the data of each process does not affect other processes by adding the changed data to the data before the change. Therefore, unlike the conventional method, the processing of other processes is not interrupted.

Further, since the process for maintaining the data consistency is performed only when each process is required, it can be efficiently performed unlike the conventional method in which arbitration is performed for each data access. it can.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.

FIG. 2 is a block diagram showing a hardware configuration of the present invention.

FIG. 3 is a block diagram showing details of FIG. 1 of one embodiment of the present invention.

FIG. 4 is a diagram showing in detail a data storage unit of the server according to the embodiment of the present invention.

FIG. 5 is a diagram showing in detail a data management unit of the server according to the embodiment of the present invention.

FIG. 6 is a diagram showing in detail the data management unit of the process of the embodiment of the present invention.

FIG. 7 is a diagram (No. 1) showing an example of changing the data of the process according to the embodiment of the present invention.

FIG. 8 is a diagram (No. 2) showing an example of changing the data of the process according to the embodiment of the present invention.

FIG. 9 is a diagram (No. 3) showing an example of changing data of the process according to the embodiment of the present invention.

FIG. 10 is a diagram showing an order of maintaining data consistency between processes according to an embodiment of the present invention.

[Explanation of symbols]

 1 process 2 data processing unit 3 data management unit 4 server communication unit 5 inter-process communication unit 6 server 7 process communication unit 8 data management unit 9 data storage unit 21, 27 central processing unit (CPU) 22, 28 communication device 24 external storage Device 25, 26 Computer 29 External storage controller

Claims (3)

[Claims] 1. A data management system including a plurality of processes for performing a reference or update process and a server for guaranteeing data consistency when the processes simultaneously reference or update the same data. To communicate with the server,
Communication for exclusive control is performed with respect to a process that performs data reference and update processing and manages the data, and a second process other than the first process that is currently performing reference and update processing for exclusive control. And a server for storing information processed by the above process, a data management method in distributed processing. 2. In a data management method in distributed processing, a process communication unit that communicates with a plurality of processes, a data storage unit that stores data in units accessed by the processes, and data in the data storage unit. A server including a data management unit that manages storage information of the data, a data processing unit that refers to or updates data, a data processed by the data processing unit, and a storage location of the data in the data storage unit. The data management in distributed processing according to claim 1, further comprising a data management unit that manages the data, a server communication unit that communicates with the server, and a process including a process communication unit that communicates between the processes. method. 3. A data management method in distributed processing, wherein, when data is stored in a sequential file for each unit accessed by the process, a data name having a unique name for each unit and the sequential file Data of the server including information on the storage location of the data name of the server, and data of the server including a first relationship table indicating a correspondence relationship between the data name and the storage location of the data name on the data storage section A management unit, and a process data management unit including a second relation table of the data name, a storage location of the data name on the data storage unit, and information indicating whether or not the data is changed due to process processing. The data management method in distributed processing according to claim 2, further comprising: