JP2000293416A - Database processor and storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate processing database of after its restoration when abnormality takes place during update processing in a database processor which continuously updates data in a database by plural application programs. SOLUTION: In this database processor 1 which performs a series of update processing updating plural pieces of data in a database 10 by continuously executing plural database update processing, a transaction ID is generated when a first database update processing of transaction is performed with plural pieces of database update processing as one transaction. Then, the transaction ID is inherited each time database update processing is switched during the transaction, and the execution of database update processing that does not inherit the transaction ID is limited. Also, the update contents of the database is determined provisionally during the transaction and determined completely after the transaction is finished normally.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database processing apparatus for performing a database update process,
The present invention relates to a database processing device that restores a database when a database update is not normally processed, and a storage medium that stores the control program.

[0002]

2. Description of the Related Art An RDBMS is used as a database used in processing by a plurality of application programs.
A database called (Relational DataBase Management System) is widely used. This RDBMS
Has a function of restoring a file when an error occurs during the update process and the file is destroyed, as in a normal database management system.

[0003] As a method of restoring a file, a method generally performed so far is to temporarily store an updated content when data in a database is updated by each processing such as deletion and appending by an application program. However, there is a method called rollback for restoring a file based on the stored update content when an error occurs.

The process of updating data in a database, including the above RDBMS, is generally executed in units called "transactions". Usually, a transaction refers to a period from the start of update processing of data in a database by an application program to the end of processing of the application program. During this transaction, the database cannot be operated by another application program.

In a conventional database, recovery by the rollback method is performed for each transaction. That is, recovery is performed by returning the updated portion of the transaction in which the abnormality has occurred to the state before the transaction.

[0006]

However, RDBMS
If the database is restored by the above-mentioned conventional method when the processes in a plurality of application programs are continuously performed on the data in the database, the database is restored in a state before the update process of the application program in which the abnormality has occurred. For this reason, there is a problem that it takes time to continue the processing after the restoration.

[0007] An operator who performs database processing operates a series of processing of updating a plurality of related data in the database by continuously performing processing by a plurality of application programs as one integrated processing. General. However, when the database is restored by the above-described conventional method, it is determined whether or not the transaction has been completed normally for each transaction, and the recovery is performed when an abnormality occurs. For this reason, the database restored after the occurrence of the abnormality reflects the processing of any one of the plurality of application programs, and does not reflect the processing after the application program in which the abnormality occurred. This makes it difficult to know from which application program the processing should be resumed even after the database is restored, resulting in an increase in the work load after the occurrence of the abnormality.

Further, in the conventional database, the transaction ends at the same time when the processing of one application program ends, so that the processing by an unrelated application program can be performed when the processing by one application program ends. For this reason, processing by an unrelated application program is performed between a plurality of application programs to be processed continuously, and data in the database may lose consistency and cause inconsistency in the data.

The present invention has been made in order to solve the above problems.
In a database processing apparatus that updates data in a database by a plurality of application programs, an object is to facilitate processing after database recovery when an abnormality occurs during the update processing.

[0010]

In order to solve the above-mentioned problems, the invention according to claim 1 includes a plurality of update units for updating data in a database, and a plurality of update units designated among the plurality of update units. In a database processing device that performs a series of update processing for updating a plurality of data in the database by an update unit, while the series of update processing is being executed, the update other than the specified plurality of update units is performed. Update control means for restricting update processing by means; storage means for storing update contents in the series of update processing; and when the series of update processing ends normally, according to the update contents stored in the storage means. A data deciding unit for deciding data in the database, and an abnormality stored in the storage unit when an abnormality occurs during the series of updating processes. New content is characterized by comprising: a data recovery unit for recovering the data in said database based on the state of the series of update processing earlier, the.

According to the first aspect of the present invention, when a series of update processes for updating data in the database by a plurality of designated update units are performed, the update unit is designated as the update unit for performing the series of update processes. Since updating of data by an updating unit that is not updated is restricted, it is possible to avoid a situation in which irrelevant data is updated while a plurality of data are continuously updated. Thereby, inconsistency of data in the database can be prevented, and the reliability of the database can be improved. In addition, since the update content updated by the update unit is determined after the series of update processing ends, if an abnormality occurs, the database is restored to a state before the series of update processing is performed. Therefore, since the database is not returned to the state in the middle of the series of update processing, the series of update processing only needs to be executed from the beginning after restoration, and the processing after restoration is very easy, and the work load is reduced. be able to.

According to a fourth aspect of the present invention, there are provided a plurality of updating means for updating data in a database, and a plurality of data in the database are updated by a plurality of designated updating means among the plurality of updating means. In a database processing apparatus that performs a series of update processing, when data is updated by the first update unit of the series of update processing, an identification code corresponding to the series of update processing is generated to update the data. The identification code generating means to be obtained by the means, and when the data updating by the updating means is completed and the data updating is performed by the next updating means during the series of updating processing, the updating is performed first. Identification code control means for causing the next update means to acquire the identification code acquired by the update means, and corresponding to the series of update processing during the series of update processing Update control means for restricting update processing by the update means which has not obtained an identification code, storage means for storing update contents in the series of update processing, and when the series of update processing ends normally, Data determination means for determining data in the database according to the update content stored in the storage means, and, if an abnormality occurs during the series of update processing, the data determination means based on the update content stored in the storage means. Data recovery means for recovering the data in the database to a state before the series of update processing.

According to the fourth aspect of the present invention, when updating is performed by the first updating means in a series of updating processing,
An identification code corresponding to the series of update processing is generated, and processing by an update unit that has not acquired the identification code is limited. Therefore, update of irrelevant data is reliably prevented, and the reliability of the database is improved. be able to. Also,
When the updating means is switched during a series of updating processes,
Since the identification code is passed from the updating unit that has performed the updating to the next updating unit, the irrelevant updating unit can be easily determined. Therefore, the reliability of the database can be easily and reliably improved.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below with reference to FIGS.

[First Embodiment] First, the configuration will be described. FIG. 1 is a diagram showing a configuration of the database processing device 1 according to the first embodiment of the present invention. As shown in FIG. 1, the database processing device 1 includes a CPU (Centra
l Processing Unit) 2, RAM (Random Access Memo)
ry) 3, ROM (Read Only Memory) 4, storage device 5,
The storage device 5 includes a storage medium 6, an input device 7, a display device 8, a printing device 9, and a database 10, and each part except the storage medium 6 is connected by a bus 11.

In the first embodiment, the CP
U2 corresponds to the updating means described in the claims by executing the processing shown in FIG. 2, and similarly, step S12 in FIG.
3 corresponds to the update control means, and performs the processing shown in steps S130 to S131 in FIG. 3 to correspond to the data determination means.
By executing the processing shown in Steps S126 to S128, it corresponds to a data recovery unit. Also, C
Step S124 or step S1 depending on PU2
By executing the processing shown in 35, the storage device 5 corresponds to a storage unit. The CPU 2 corresponds to the identification code generation unit by executing the processing shown in step S134 in FIG. 3, and corresponds to the identification code control unit by executing the processing shown in steps S106 to S107 in FIG.

The CPU 2 reads a basic control program stored in the ROM 4 and various application programs stored in the storage device 5 in accordance with an instruction input from the input device 7, and develops the basic control program in a memory area provided in the RAM 3. Execute. Specifically, the CPU 2
By executing the basic control program stored in M4, control such as shifting each unit of the database processing apparatus 1 to the initial state is performed. Further, the CPU 2 reads and executes various application programs stored in the storage device 5 to execute a database update process (FIG. 2) described later to delete, add, or update data in the database 10. Execute various processes.

The CPU 2 executes a database control process (FIG. 3) described later by reading a database control program described later from the storage device 5 and executing it.
When the data stored in the database 10 is updated with the execution of the application program, backup data of the updated content of the database 10 is generated and stored in the storage device 5, and a provisional commit is performed. further,
A provisional commit is performed whenever data in the database 10 is updated by various application programs until the end of the transaction. And
When a full commit is specified or when the transaction ends, a full commit is executed to completely determine the update content of the provisionally committed data.

Further, the CPU 2 generates display data for displaying various information such as the contents of the database 10 and the progress of the update processing in accordance with an instruction input from the input device 7, outputs the display data to the display device 8, and outputs the display data. It is displayed on the display screen of the device 8. Further, print data for printing the content displayed on the display device 8 is generated and output to the printing device 9,
Execute printing.

The RAM 3 forms a memory area for temporarily storing various programs processed by the CPU 2 and data related to the processing.

The ROM 4 stores various basic control programs for executing a series of processes executed by the CPU.
Each basic control program in the ROM 4 is stored in the form of a program code readable by the CPU 2.

The storage device 5 has a storage medium 6 in which programs, data, and the like are stored.
It is composed of an optical recording medium or a semiconductor memory. The storage medium 6 is either fixedly provided in the storage device 5 or removably mounted.
And various data related to the various application programs to be executed.

The data and the like stored in the storage medium 6 may be received and stored from another device connected via a communication line or the like. A configuration may be adopted in which a storage device having the above storage medium is provided on the other connected device side, and programs and data stored in the storage medium 6 are used via a communication line.

The input device 7 includes a keyboard having cursor keys, numeric input keys, and various function keys, and a pointing device such as a mouse. When the keyboard is operated, the input device 7 generates a press signal corresponding to the pressed key and outputs the signal to the CPU 2, and when a pointing device such as a mouse is operated, together with an operation signal such as a click signal, The relative coordinate data indicating the coordinates designated by this operation is output to the CPU 2.

The display device 8 is a CRT (Cathode Ray Tub).
e) and an LCD (Liquid Crystal Display), and displays various screens based on display data input from the CPU 2.

The printing device 9 includes various printers.
Various documents are printed and output according to the print data input from the CPU 2.

The database 10 is a database in which various types of data processed by the database processing device 1 are stored. Similar to the storage medium 6, a magnetic or optical storage medium or a storage medium such as a semiconductor storage element is used. Is stored in The data in this database 10
The data is updated according to the database control process by the data update process of various application programs executed by the CPU 2. The storage medium storing the database 10 may be configured to be provided in the database processing device 1 separately from the storage device 5 and the storage medium 6, or may be stored in a storage area provided in the storage medium 6. May be stored.

Next, the operation of the database processing device 1 according to the first embodiment of the present invention will be described.

In the operation of the database processing apparatus 1 described below, a series of processes for continuously executing a plurality of application programs and updating a plurality of related data are treated as one transaction. That is, even if the database update processing by one application program ends, if the update processing of related data is subsequently performed by another application program, the transaction is continued without ending. Then, after a continuous update process is performed by a plurality of application programs, 1
One transaction ends.

In order to realize such processing, the database processing apparatus 1 uses a transaction ID. The transaction ID is a code for continuing the started transaction across a plurality of application programs, and is issued when the first application program of the transaction is executed. Also,
In the same transaction, when the processing by the application program ends and the processing of the next application program starts, the transaction ID is inherited between the applications.

Then, when the processing shifts to the processing of the next application program during the transaction, the authentication based on the transaction ID is performed, so that the database can be updated only by the application program that has acquired the correct transaction ID.
This makes it impossible for the application program irrelevant to the transaction to update the database 10 and does not cause inconsistency or inconsistency in the data in the database 10.

In the operation of the database processing device 1 described below, the database 10 updated by the database update process by the application program is used.
The update contents of are temporarily determined in a state where they can be easily recovered until a complete determination is instructed. This tentative confirmation processing is hereinafter referred to as “provisional commit”. In this provisional commit, backup data of the updated contents of the database 10 is generated and stored in the storage device 5.
The database 10 can be easily restored to the state before the provisional commit based on the backup data.

The processing for completely confirming the updated data with respect to this provisional commit is hereinafter referred to as “complete commit”.
Call. This full commit is executed when a full commit is instructed during a database update process described later. After the update content is completely committed, the update content is reflected in the database 10 even if recovery processing such as rollback is performed, so this is useful when it is no longer necessary to restore the state before the update. .

First, an operation when an application program for updating data in the database 10 is executed by the CPU 2 will be described.

FIG. 2 is a flowchart showing a database update process executed by the database processing device 1. The database update process shown in FIG. 2 is a process when various application programs stored in the storage device 5 are executed and the contents of the database 10 are updated.

In the database update process, first, C
PU2 determines whether the application program being executed is the first application program in a transaction (step S101). here,
If the application program being executed is the first application program in the transaction, the process proceeds to step S102, and the data in the database 10 is updated in accordance with the contents input from the input device 7.

In step S103, the CPU 2 designates a provisional commit for the updated contents in step S102. When the provisional commit is specified in step S103, in the database control process (FIG. 3), backup data of the updated contents in step S102 is generated and stored in the storage device 5. If the database update process shown in FIG. 2 is not completed normally, the database 10 is restored based on the backup data of the update contents temporarily stored in the storage device 5.

Thereafter, a transaction ID is issued by the database control process (FIG. 3), and the issued transaction ID is obtained (step S10).
4).

In step S104, the transaction ID
After the is acquired, various commits are specified as needed according to an input instruction from the input device 7 (step S105). For example, when there is data that is to be partially partially committed, or in order to completely commit all data that has been provisionally committed so far, an instruction to perform a complete commit is input from the input device 7 here. .

Then, the transaction ID acquired in step S104 is taken over by the application program to be executed next in the same transaction (step S106), and the process ends.

Subsequently, when the database update process by another application program starts, in the database update process (FIG. 2), the process proceeds from step S101 to step S107, and the transaction ID is acquired from the application program executed immediately before.

Then, the CPU 2 specifies the transaction ID obtained in step S107 (step S107).
108). The transaction ID specified here is
The data is compared with the transaction ID issued in the database control process (FIG. 3), and the data in the database 10 can be updated only when they match.

In step S109, the data in the database 10 is updated according to the instruction input from the input device 7 (step S109), and thereafter, it is determined whether or not the application program being executed is the last application program in the transaction (step S109). Step S11
0). If there is an application program that continues the transaction after the running application program, the process proceeds to step S105.

If the application program being executed is the last application program in the transaction, the flow shifts to step S111 to specify complete commit of the updated data, and ends the processing. As a result, in the database control process (FIG. 3), all of the provisionally committed updates during the transaction are completely committed, and the backup data of the updates stored in the storage device 5 is deleted.

Next, a database control process executed to control data update in the database 10 according to the data update process (FIG. 2) by the application program will be described.

FIG. 3 is a flowchart showing a database control process executed by the database processing device 1. In the database control process shown in FIG.
First, the CPU 2 determines whether a transaction ID has already been issued to the application program for performing the database update process (FIG. 2) (step S1).
21).

If it is determined in step S121 that the transaction ID has not been issued, it means that the transaction has not been started.
The process proceeds to step S122 to start a transaction, and then proceeds to step S124.

If the transaction ID has already been issued, the transaction started when another application program is already executed is continued, and the application program being executed is designated in step S108 (FIG. 2). It is determined whether the transaction ID is a transaction ID issued by the database control process at the start of the transaction (step S123). Here, when the transaction IDs match, the process proceeds to step S124, and when they do not match, the process ends.

As described above, when a transaction is started, a transaction ID is issued by the database control process, and this transaction ID is obtained in the database update process by the application program. This transaction ID is taken over by the next application program in the same transaction when the processing by the application program ends. Therefore, the same transaction ID is sequentially taken over by all application programs executed during one transaction.

Therefore, when the transaction ID specified by the application program and the transaction ID issued in the database control processing do not match, processing by the application program unrelated to the ongoing transaction is about to be executed. Then, the CPU 2 refuses to update the database and ends the processing.

In step S124, step S102
The data in the database 10 is updated corresponding to the operation in (FIG. 2) or step S109 (FIG. 2). The updated content of the database 10 updated in step S124 is temporarily stored in the storage device 5, and is not provisionally committed.

In a succeeding step S125, it is determined whether or not the update processing of the data in the database 10 has been completed normally. If the data update processing has been completed normally, the flow shifts to step S129 described later. If the update of the data in the database 10 has not been completed normally, the process proceeds to step S126, and the database 10 is restored by rollback based on the update contents temporarily stored in the storage device 5. I do.

After restoring the database 10 in step S126, the CPU 2 further determines in step S127 whether there is any data that has been provisionally committed since the transaction was started. Then, when there is the update content in the provisionally committed state, the storage device 5
Then, the database 10 is rolled back based on the updated contents backed up therein (step S128). If there is no data in the provisional commit in step S127, the process ends.

That is, when processing the second and subsequent application programs after the transaction is started, the backup data temporarily stored in the database update process by the application program executed earlier in the same transaction is stored. I have. Therefore, by performing rollback based on the backup data of the update contents during all the provisional commits,
When an abnormality occurs during data update, the database 10 can be restored to a state before all data updates of the transaction are performed.

On the other hand, if it is determined in step S125 that the update processing of the data in the database 10 has been completed normally, the process proceeds to step S129, where the input device 7
It is determined whether or not a complete commit has been designated by the instruction input from the server.

If complete commit is specified in step S111 of the database update process (FIG. 2), the CPU 2 proceeds to step S130 to completely determine the provisionally committed update content and to store the updated data in the storage device. 5 is discarded (step S131). Thereafter, the transaction ends, and a series of processing ends.

Further, in Step S129, Step S
If it is determined in step 111 (FIG. 2) that the complete commit has not been specified, the transaction is continued as it is, and the CPU 2 determines whether a transaction ID has already been issued (step S133). If the transaction ID has already been issued, the transaction ID is handed over to the next application program in step S106 of the database update process (FIG. 2), and the process proceeds to step S135. If the transaction ID has not been issued yet, the transaction ID is issued in step S134, and the process proceeds to step S135.

In step S135, step S124
Tentatively commits the update contents of the database 10 performed in step (1), generates backup data of the update contents, stores it in the storage device 5, and ends the processing.

Thereafter, when the database update process (FIG. 2) by the last application program of the transaction is performed, a complete commit is specified in step S111, and a complete commit is performed in step S129 (FIG. 3). Then, the updated contents of the provisionally committed state are completely determined.

As described above, according to the database processing apparatus 1 in the first embodiment of the present invention, when the database update processing for updating the data in the database 10 is started by the application program, the database update processing is started. , The control process is started, a transaction is started to update the data in the database 10, and the updated contents are temporarily stored in the storage device 5.
To memorize. Then, a unique transaction ID is issued for each transaction, and backup data indicating the updated contents of the database 10 is generated and stored, thereby executing the provisional commit process. When the processing of the running application program is completed and the transaction ID is taken over by the next application program, the transaction ID acquired by the application program is compared with the issued transaction ID to perform authentication. If they do not match, the process ends without updating the database 10. If the transaction IDs match, the database 10 is updated. Further, when the processing by the last application program of the transaction ID is completed, the provisionally committed update content is completely determined and the full commit process is executed. Then, when an abnormality occurs during the update processing of the database 10, all of the update contents temporarily stored in the storage device 5 and the provisionally committed update contents are rolled back, and the database 10 is restored. When the processing of the application program ends, all the update contents temporarily stored in the storage device 5 and the provisionally committed update contents are completely committed, and the processing ends.

Thus, when updating data in the database 10, when updating a plurality of related data by a plurality of applications, it can be executed as one transaction including processing by all application programs. The updated contents during the transaction are backed up as a provisional commit. For this reason, when an abnormality occurs during the update processing of the database 10, the database 10 is restored to the state before the one transaction is performed, not before the update by the application program in which the abnormality has occurred. be able to.

Therefore, even if an abnormality occurs when the contents of the database 10 are updated by a plurality of application programs, the update processing by the plurality of application programs may be executed again from the beginning after the database 10 is restored. The database 10 is not halfway updated.

While a transaction is being executed, updating of the database 10 by an application program that has not acquired the transaction ID issued in the database control process can be refused. Processing is not performed by application programs that are not related in the meantime. For this reason, the database 10
Inconsistency or inconsistency in the content of the content can be reliably avoided.

For this reason, according to the database processing apparatus 1 of the first embodiment, the updating process of the database 10 can be executed efficiently, and the reliability of the data in the database 10 can be improved. Can be.

In the database processing apparatus 1 according to the first embodiment, in the database update processing shown in FIG. 2, when the processing by the first application program of the transaction is completed, all the updated contents are provisionally committed. Although the transaction ID is passed to the next application program, the present invention is not limited to this. For example,
In order to cope with the case where the transaction is completed by one application program, the update contents of the database 10 may be completely committed at the time when the processing by the first application program ends, so that the transaction may be ended.

[Second Embodiment] FIG. 4 is a flowchart showing a database part-by-part control process executed by the database processing apparatus 1 according to a second embodiment of the present invention. Since the database processing device 1 according to the second embodiment has the same configuration as that of the first embodiment, description and illustration of the configuration will be omitted.

In the second embodiment, the CP
U2 corresponds to the confirmation control means described in the claims by executing the processing shown in step S218 of FIG. The input device 7 is also operated by the CPU 2 in step S in FIG.
By executing the processing shown in step 105 and step S217 in FIG. 4, it corresponds to the type designation means.

The database update processing when the application program is executed by the database processing apparatus 1 according to the second embodiment is the same as the database update processing (FIG. 2) according to the first embodiment. Therefore, the description is omitted.

In the database part control process shown in FIG. 4, first, the CPU 2 determines whether or not a transaction ID has already been issued. If the transaction ID has not been issued yet, the application program that performs the database update process (FIG. 2) is the first application program in the transaction, and therefore the process moves to step S202 to start the transaction, and moves to step S204.

In step S204, CPU 2 updates the data in database 10 according to the operation in step S102 (FIG. 2). Note that the updated contents of the database 10 in step S204 are temporarily stored in the storage device 5, and are not provisionally committed. In a succeeding step S205, it is determined whether or not the update of the database 10 in the step S204 has been normally completed.

On the other hand, if it is determined in step S201 that the transaction ID has already been issued, the transaction started when another application program is being executed is continued, and the application program being executed is It is determined whether or not the transaction ID specified in S108 (FIG. 2) is a transaction ID issued by the database control process at the start of the transaction (step S203). If the transaction IDs match, the process proceeds to step S204, and if they do not match, the process ends. Then, step S2
In step S04, the data in the database 10 is updated in accordance with the operation in step S109 (FIG. 2).
Move to 205.

If it is determined in step S205 that the update of the database 10 has not been completed normally, the process proceeds to step S20.
6 and restore the database 10 by rolling back the updated contents in step S204. Further, in step S207, it is determined whether there is backup data in a provisionally committed state. If there is no provisionally committed backup data, the process is terminated. If there is backup data, the backup data is deleted. Rollback is performed to restore the database 10 to the state before the transaction, and the process ends.

Also, in step S205, step S2
If it is determined that the update processing of the database 10 in 04 has been completed normally, the process proceeds to step S209, and it is determined whether or not complete commit has been designated in step S111 of the database update processing (FIG. 2) by the application program. I do. When the complete commit is designated, the update contents temporarily stored in the storage device 5 are completely determined in step S204 (step S210), and then all the update contents in the provisionally committed state are completely updated. Commit (step S211) and end the transaction (step S212).

On the other hand, if it is determined in step S209 that the complete commit has not been specified, the process proceeds to step S2
The process proceeds to 13 to determine whether the transaction ID has already been issued. If the transaction ID has not been issued yet, the transaction ID is issued in step S214, and the process proceeds to step S215.
If the transaction ID has already been issued,
The process moves directly to step S215.

In step S215, step S204
Tentatively commits the update content temporarily stored in the storage device 5 to generate backup data of the update content and store it in the storage device 5. In a succeeding step S216, a commit ID is issued for each provisionally committed update content.
The commit ID is a code used to designate a specific update content from all provisionally committed updates.

Then, the flow shifts to step S217, where it is determined whether or not a partial complete commit is designated by the commit ID in step S105 of the database update process (FIG. 2) by the application program.

The partial complete commit is a process in which only a part of the update contents specified in the provisionally committed update contents is completely committed. This partial complete commit is performed in step S105 in the database update process of FIG.
The specified update content can be prevented from being rolled back even if an error occurs in the processing within the same transaction.

If the contents of the update are provisionally committed, the rollback is performed when an error occurs in the subsequent processing in the same transaction, and the database 10 after recovery will be in the state before the transaction. . However, since the update specified by the partial full commit is completely committed without waiting for the end of the transaction, even if the database 10 is restored by rollback, the specified update is reflected in the content of the database 10. You.

In step S217, step S105
If it is determined in FIG. 2 that the partial complete commit by the commit ID has been designated, the process proceeds to step S218, where the update corresponding to the designated commit ID is completely committed, and the backup data is deleted. Then, the process proceeds to step S219. If it is determined in step S217 that the partial complete commit has not been specified, the process proceeds to step S219.

In step S219, it is determined whether or not the database 10 is to be updated by the same application program. If another data is to be updated, the process returns to step S204 to return to the database by the application program being executed. If the update of 10 has been completed, the processing is terminated.

Thereafter, when the database update processing (FIG. 2) by the last application program of the transaction is performed, a complete commit is specified in step S111, and a complete commit is performed in step S211 (FIG. 4). Then, the updated contents of the provisionally committed state are completely determined, and the transaction is terminated.

As described above, according to the database processing apparatus 1 in the second embodiment of the present invention, when the database update processing is performed by the application program, the database control processing is performed in the database control processing.
In addition to tentatively committing the update content of 0, a commit ID is issued for each tentatively committed update content, and a partial complete commit is designated by the commit ID.
Only the update content corresponding to the specified commit ID is completely committed.

Thus, when updating a plurality of related data in the database 10 by a plurality of application programs, a commit ID is assigned to each update content of the plurality of application programs, so that a transaction is completed before the transaction is completed. Even so,
Only specific updates can be fully committed. Therefore, the state of the database 10 restored when an abnormality occurs during the update processing of the database 10 can be set to a state in which only an arbitrary update processing is reflected. For example, updates that are not closely related to changes made by other application programs in the same transaction and that do not cause a problem even if they are completely committed independently are reflected in the database 10 that is restored when an error occurs. be able to. For this reason, the state of the database 10 that is restored when an abnormality occurs during the update processing of the database 10 can be set to a desired state. Therefore, the work after the database 10 is restored can be reduced, and the database 10 can be updated more efficiently while reliably preventing inconsistencies and inconsistencies in the data in the database 10.

[Third Embodiment] FIG. 5 is a flowchart showing the control processing for each database processing executed by the database processing apparatus 1 according to the third embodiment of the present invention. FIG. 6 is a flowchart showing step S304 in FIG.
7 is a flowchart showing the process type update process in more detail, and FIG. 7 is a flowchart showing the process-specific complete commit process shown in step S317 of FIG. 5 in more detail. Hereinafter, the third embodiment of the present invention will be described with reference to FIGS.
The operation of the database processing apparatus 1 according to the embodiment will be described.

In the third embodiment, the CP
U2 corresponds to the determination means described in the claims by executing the processing shown in step S321 in FIG.
By executing the processing shown in steps S322 to S324, it corresponds to the storage control means, and by executing the processing shown in steps S331 to S334 of FIG. 7, it corresponds to the determination control means for each type. Further, the input device 7 corresponds to a type designation unit by the CPU 2 executing the processing shown in step S105 in FIG. 2 and step S316 in FIG.

As will be described later, in the database processing type control processing according to the third embodiment, the type of the update processing of the database 10 is determined. As a type of processing when updating data in the database 10, for example,
Insert data to add new item data (INSER
T) processing, data update (UPDATE) processing for rewriting old data with new data of the same item, and data deletion (DELETE) processing for deleting unnecessary data. Hereinafter, in the third embodiment, the three types of processing will be described assuming that the control processing for each database processing is executed.

Therefore, in the storage device 5, the INS in which the data insertion processing is registered in association with the above three types of processing.
UPDAT where ERT processing group and data update processing are registered
It is assumed that three tables of an E processing group and a DELETE processing group in which data deletion processing is registered are stored.

Note that the components of the database processing apparatus 1 according to the third embodiment except for the internal configuration of the storage device 5 have the same configuration as those of the first and second embodiments. The description and illustration are omitted. Further, the database update processing when the application program is executed by the database processing apparatus 1 according to the third embodiment is the same as the database update processing (FIG. 2) according to the first embodiment. Description is omitted.

In the control processing for each database processing shown in FIG. 5, the CPU 2 first determines whether a transaction ID has already been issued (step S301). If the transaction ID has not been issued yet, the application program for performing the database update process (FIG. 2) is the first application program in the transaction, and therefore the flow shifts to step S302 to start the transaction, and shifts to step S304.

In step S304, the CPU 2 updates the processing type of the data in the database 10 according to the operation in step S102 (FIG. 2), and executes step S305.
Move to.

The process type updating process in step S304 includes a process of updating the contents of the database processing device 1 as INSERT (insert), UPDATE, as described later.
(Update) and DELETE (delete). Note that the updated content in step S304 is temporarily stored in the storage device 5 without being provisionally committed.

On the other hand, if it is determined in step S301 that the transaction ID has already been issued, the transaction started when another application program is being executed is continued, and the application program being executed is It is determined whether or not the transaction ID specified in S108 (FIG. 2) is a transaction ID issued by the database control process at the start of the transaction (step S303). If the transaction IDs match, the process proceeds to step S304, and if they do not match, the process ends. Then, step S3
In step 04, the process type update process is executed in accordance with the operation in step S109 (FIG. 2), and the flow shifts to step S305.

In step S305, step S304
It is determined whether or not the update of the database 10 has been completed normally. In this step S305, the database 10
If it is determined that the update of the
The process proceeds to step S306, and the contents of the database 10 are rolled back in step S304.
To recover. Further, in step S307, it is determined whether there is backup data in a provisionally committed state. If there is no provisionally committed backup data, the process ends. If there is backup data, the backup data is deleted. Rollback is performed to restore the database 10 to the state before the transaction, and the process ends.

Also, in step S305, step S3
If it is determined that the update process of the database 10 in 04 has been completed normally, the process proceeds to step S109, and it is determined whether or not a full commit has been designated in step S111 of the database update process (FIG. 2) by the application program. I do. If the complete commit is specified, the update contents temporarily stored in the storage device 5 are completely determined in step S304 (step S310), and then all the update contents in the provisionally committed state are completely updated. Commit (step S311) and end the transaction (step S312).

On the other hand, if it is determined in step S309 that a complete commit has not been specified, the process proceeds to step S3
The process proceeds to 13 to determine whether the transaction ID has already been issued. If the transaction ID has not been issued yet, the transaction ID is issued in step S314, and the process proceeds to step S315.
If the transaction ID has already been issued,
Then, control goes to step S315.

In step S315, step S304
Tentatively commits the update content temporarily stored in the storage device 5 to generate backup data of the update content and store it in the storage device 5.

Then, the flow shifts to step S316, where it is determined whether or not a process-specific commit has been designated in step S105 of the database update process (FIG. 2) by the application program. If it is determined in step S316 that a commit by process has been specified, the process proceeds to step S317, executes a complete commit process by process described later, and proceeds to step S318.

Here, the commit specification for each process is the IN designation determined at the time of the process type update process in step S304.
For each type of SERT, UPDATE, and DELETE, whether or not to commit completely is specified. For example, INSERT in step S105 (FIG. 2)
If it is specified that only the process of the above is completely committed,
In step S317, the update content related to the specified INSERT process is completely committed.

If it is determined in step S316 that the complete commit designation for each process has not been performed, the flow shifts to step S318. In step S318, it is determined whether or not the database 10 is updated by the same application program. If another data is updated, the process returns to step S304 to update the database 10 with the application program being executed. Is completed, the process ends.

FIG. 6 is a flowchart showing the processing type update processing executed in step S304 of FIG. 5 in more detail. As shown in FIG. 6, in the process type update process, the CPU 2 first determines the type of update of the database 10 (step S321).

In step S321, step S102
It is determined which of the above three types of processes the process instructed in (FIG. 2) or step S109 (FIG. 2) corresponds to. If it is determined in step S321 that the designated process is a data insertion process, the process proceeds to step S321.
322, the update information indicating the contents of the specified data insertion processing is stored in the INSER stored in the storage device 5.
Register in the T processing group. Thereafter, the CPU 2 proceeds to step S
The process proceeds to the process of 305 (FIG. 5).

If it is determined in step S321 that the designated process is a data update process, the flow advances to step S323 to store update information indicating the content of the designated data update process in the storage device 5. UPD stored in
After registering in the ATE process group, the process proceeds to the process of step S305 (FIG. 5).

Further, if it is determined in step S321 that the designated process is a data deletion process, the flow shifts to step S324 to update information indicating the content of the specified data deletion process in the storage device 5. DE stored in
Registered in the LETE processing group, and step S305 (FIG. 5)
Move to the processing of.

FIG. 7 is a flowchart showing the process-specific complete commit process executed in step S317 of FIG. 5 in more detail. As shown in FIG. 7, in the process-specific commit process, the CPU 2 first determines the type of the process specified in step S105 (step S331).

In step S331, step S105
The processing specified in (FIG. 2) is the above INSERT, UP
It is determined which of the three types of processing is DATE and DELETE. If it is determined in step S331 that the designated process is the data insertion process (INSERT), the process proceeds to step S332, and all the processes registered in the INSERT process group stored in the storage device 5 are executed. Commit the update information completely. Then the CPU
2 shifts to the process of step S319 (FIG. 5).

If it is determined in step S331 that the designated process is a data update process (UPDATE), the process shifts to step S333 to be registered in the UPDATE process group stored in the storage device 5. The update information is completely committed, and the process proceeds to step S319 (FIG. 5).

Further, if it is determined in step S331 that the designated process is a data deletion process (DELETE), the process proceeds to step S334, and the storage device 5
The update information registered in the DELETE processing group stored in the file is completely committed, and the process proceeds to step S319 (FIG. 5).

Therefore, by specifying the type of process in step S105 (FIG. 2) and instructing the process-specific complete commit process, only a specific type of process can be completely committed.

As described above, according to the database processing apparatus 1 of the third embodiment of the present invention, when the contents of the database 10 are updated by the application program, the processing type of the updated contents is changed to the data insertion processing. (INSERT), data update processing (UPDATE),
Alternatively, it is determined which of the data deletion processing (DELETE) corresponds, and update information indicating each update content is registered in each table of the INSERT processing group, the UPDATE processing group, and the DELETE processing group provided in the storage device 5. When a specific processing type is designated from these processing types and a complete commit is instructed, all update contents of the designated processing type are completely committed. Thus, for each type of processing for updating the database 10, a complete commit can be performed even before the transaction is completed. The state can be any state.

Therefore, the work after the database 10 is restored can be reduced, and the database 10 can be updated more efficiently while reliably preventing inconsistencies and inconsistencies in the data in the database 10. .

In the database processing apparatus 1 according to the third embodiment, the type of the update processing of the database 10 is INSERT, UPDATE, DEL
Although the processing is performed by dividing the processing into three types of ETE, the present invention is not limited to this, and the types can be further increased according to the configuration of the database 10.

[0112]

According to the database processing device of the present invention and the storage medium of the present invention, a series of update processing for updating data in the database by a plurality of specified update means is performed. In such a case, data update by an update unit not specified as an update unit that performs this series of update processing is restricted, so that irrelevant data is updated while a plurality of data are continuously updated. Situation can be avoided. Thereby, inconsistency of data in the database can be prevented, and the reliability of the database can be improved. In addition, since the update content updated by the update unit is determined after the series of update processing ends, if an abnormality occurs, the database is restored to a state before the series of update processing is performed. Therefore, since the database is not returned to the state in the middle of the series of update processing, the series of update processing only needs to be executed from the beginning after restoration, and the processing after restoration is very easy, and the work load is reduced. be able to.

According to the database processing device of the present invention, the update contents of each update means in a series of update processing are stored in the storage means, and only the designated update contents among these update contents are determined. be able to. As a result, it is possible to determine only the update contents to be determined first, and not to recover even if an abnormality occurs later. Therefore, when the same update process is executed again, such as when the update content is extremely complicated or the amount of data to be handled is extremely large, the update content that is burdensome can be prevented from being restored at the time of restoration. Since recovery can be performed according to the contents of the processing, it is possible to prevent a decrease in work efficiency while improving the reliability of data in the database.

According to the database processing apparatus of the third aspect of the invention, the update contents in the series of update processing are stored for each type, and the update contents of the specified type are determined from the stored update contents. be able to. Therefore, it is possible to perform an operation such as fixing only the update content for newly adding data and leaving the update content for deleting data in a recoverable state without finalizing the update content. Recovery can be performed in accordance with the characteristics of the type of update content, such as the content and the type of update content that the user wants to be able to recover from in the event of a mistake. Thus, it is possible to prevent a decrease in work efficiency while improving the reliability of data in the database.

According to the database processing apparatus of the invention described in claim 4 and the storage medium described in claim 6, when updating is performed by the first updating means in a series of update processing, the series of update processing is performed. Since the identification code is generated and the processing by the updating means that does not acquire the identification code is restricted, the update of irrelevant data can be reliably prevented, and the reliability of the database can be improved. Further, when the updating unit is switched during a series of updating processes, the identification code is inherited from the updating unit that has performed the updating to the next updating unit, so that the irrelevant updating unit can be easily determined. . Therefore, the reliability of the database can be easily and reliably improved.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of a database processing device 1 according to a first embodiment of the present invention.

FIG. 2 is a flowchart showing a database update process of an application program executed by the database processing device 1 shown in FIG.

FIG. 3 is a flowchart showing a database control process executed by a database processing device 1 in response to the database update process shown in FIG. 2;

FIG. 4 is a flowchart illustrating a database part-by-part control process executed by a database processing apparatus 1 according to a second embodiment of the present invention.

FIG. 5 is a flowchart illustrating a control process for each database process executed by a database processor 1 according to a third embodiment of the present invention.

FIG. 6 is a flowchart showing the process type update process shown in step S304 of FIG. 5 in more detail;

FIG. 7 is a flowchart illustrating the process-specific complete commit process shown in step S317 of FIG. 5 in more detail.

[Explanation of symbols]

Reference Signs List 1 database processing device 2 CPU (update control means; data determination means; data recovery means; determination control means; determination means; storage control means; type-specific determination control means; identification code generation means; identification code control means) 3 RAM 4 ROM Reference Signs List 5 storage device (storage means) 6 storage medium 7 input device (designation means; type designation means) 8 display device 9 printing device 10 database 11 bus

Claims (6)

[Claims] A plurality of updating means for updating data in a database, and a series of updating processes for updating a plurality of data in the database by a plurality of designated updating means among the plurality of updating means. In the database processing device for performing, while the series of update processing is being executed, update control means for restricting update processing by the update means other than the designated plurality of update means, and updating in the series of update processing Storage means for storing the contents; data determination means for determining data in the database according to the update contents stored in the storage means when the series of update processing ends normally; during the series of update processing In the event that an abnormality has occurred, the data in the database is updated based on the update contents stored in the storage means. And a data recovery means for recovering the state before the data processing. 2. The storage device according to claim 1, wherein the storage unit stores update contents in the series of update processing for each of the plurality of designated update units. Specifying means for specifying the update content by the specific update means; and determination control means for determining data in the database by the data determination means in accordance with the update content specified by the specification means, The database processing apparatus according to claim 1, wherein: 3. A determination means for determining a type of an update content in the series of update processing; a storage control means for storing the update content in the storage means for each type determined by the determination means; Type specifying means for specifying a specific type of update content from among the update content stored in the storage means; and the data in the database by the data determination means in accordance with the type of update content specified by the type specification means. 2. The database processing apparatus according to claim 1, further comprising: a type-specific determination control unit that determines the type. 4. A plurality of updating means for updating data in a database, and a series of updating processes for updating a plurality of data in the database by a plurality of designated updating means among the plurality of updating means. In the database processing apparatus, when data is updated by the first updating unit of the series of update processing, an identification code corresponding to the series of update processing is generated to be acquired by the first updating unit. Generating means, during the series of update processing, when the update of the data by the update means is completed and the next update of the data is performed by the update means, the update means which has performed the update earlier has obtained Identification code control means for causing the next update means to acquire the identification code; and acquiring the identification code corresponding to the series of update processing during the series of update processing Update control means for restricting update processing by the update means that has not been performed; storage means for storing update contents in the series of update processing; and storage in the storage means when the series of update processing ends normally. Data determination means for determining data in the database according to the updated content, and, if an abnormality occurs during the series of update processing, data in the database based on the update content stored in the storage means. And a data recovery unit for recovering the state before the series of update processing. 5. A storage medium storing a computer-executable program for executing a series of update processing for updating a plurality of data in the database by continuously executing a plurality of designated update processing programs. And a computer-executable program code for restricting update processing by an update processing program other than the designated update processing program while the series of update processing is being executed; and A computer-executable program code for storing the update content in the storage means, and when the series of update processing ends normally, determine the data in the database according to the update content stored in the storage means. Computer-executable program code for performing the series of update processing When an abnormality has occurred, a computer-executable program code for restoring data in the database to a state before the series of update processing based on the update content stored in the storage means, A storage medium storing a program including the program. 6. A storage medium storing a computer-executable program for performing a series of update processing for updating a plurality of data in the database by continuously executing a plurality of designated data update processing programs. When the first data update processing program of the series of update processing is executed, a computer for generating an identification code corresponding to the series of update processing and causing the computer to acquire the identification code with the data update processing program is Executable program code, acquired by the previously executed data update processing program when the update processing program ends and processing by the next data update processing program is started during the series of update processing Computer for executing the next data update processing program to obtain the obtained identification code Program code, during the series of update processing, a computer-executable program code for limiting the execution of the data update processing program that has not acquired the identification code corresponding to the series of update processing, A computer-executable program code for storing the update content in the series of update processing in the storage unit; and when the series of update processing ends normally, the database according to the update content stored in the storage unit. A computer-executable program code for determining the data in the memory, and if an error occurs during the series of update processing, the data in the database based on the update content stored in the storage unit. Computer-executable program for restoring to a state before the series of update processing Storage medium characterized by storing a program containing a over de, a.