JPH03276351A - Restoration system for contents of shared memory - Google Patents

Abstract

PURPOSE:To ensure the on-line services with high quality and high reliability and with no deterioration of the service performance by restoring the contents of a memory without stopping the on-line service even though the contents of an exclusive memory and a shared memory are broken by a program fault. CONSTITUTION:If a program fault occurs while a business work processing part 4 receives the input data from a terminal 10 and is carrying out the business work processing, a program fault detection part 7 detects the fault and deals with it. Then the part 7 gives a temporary stop request and an initialization request to a business work processing control part 3 together with a shared memory contents restoring request given to a shared memory contents restoring part 8 respectively. The part 3 outputs the stop instructions to a synchronization control part 9 in the order of requests to temporarily stop the part 9. Then the part 8 restores the contents of a shared memory 6 with the stop end information received from the part 9. The part 8 gives a restart request to the part 9 for the part 3 when the storage processing is over. The part 9 outputs a restart instruction to the part 3 for restart of the due control.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to an online system, and in particular, when a plurality of programs exist and operate in parallel, when a shared memory is used, the shared This invention relates to a recovery method for recovering memory contents.

[Conventional technology]

Conventionally, when a program failure occurs for some reason, only the private memory of that program is restored to the point immediately before the data that caused the failure. Therefore, if the shared memory contents between programs were destroyed, the online system would go down, and the online system would be restarted to restore the shared memory contents up to just before the data that caused the failure.

[Problem to be solved by the invention]

In the conventional shared memory content recovery method described above, if a program failure occurs and the shared memory content is destroyed,
The disadvantage is that it takes a long time from an online shutdown to an online restart, recovery of the shared memory contents, and restart of online operations, and the time it takes from the occurrence of a program failure to the restart of online operations. be.

[Means to solve the problem]

The shared memory content recovery method of the present invention detects a program failure, requests a shared memory content recovery unit to restore the shared memory content, requests a plurality of business processing control units to temporarily stop,
a program failure detection unit that requests initial settings from the business processing control unit corresponding to the business processing unit that caused the program failure; and a program failure detection unit that requests the business processing control unit corresponding to the business processing unit that caused the program failure to inform the shared memory content recovery unit that all of the business processing control units have been temporarily stopped. a synchronization control unit that notifies and restarts the business processing control unit in response to a request from the shared memory content recovery unit; and a synchronization control unit that restarts the business processing control unit based on a request from the shared memory content recovery unit; and after all of the business processing control units have temporarily stopped, the shared memory is restored from the shared memory content storage file , a shared memory content recovery unit that requests the synchronization control unit to restart the business processing control unit; and a shared memory content recovery unit that receives input data from a terminal and delivers it as is to the business processing unit.

When a pause request is received, a stop instruction is output to the synchronization control unit to temporarily stop the operation, when a restart request is received, the operation is restarted, and when an initialization request is received, the business processing unit is temporarily deleted. and a business processing control unit that sets necessary initial values after re-registration.

〔Example〕

Next, the present invention will be explained using the drawings.

FIG. 1 is an overall configuration diagram of an embodiment of the present invention. CPU
Also, inside the memory 1 is a data transmitting/receiving section 2. Multiple business processing control units 3. Multiple business processing units4. A plurality of business processing unit private memories 5° shared memory 6, program failure detection unit 7. Shared memory content recovery unit 8. A synchronization control section 9 is present. In addition, C, PU, and memory 1 include terminal 10.
Operation console 11, printer 12. Business file 13. Program storage file 14. Shared memory content storage file 1
5 is connected. Note that the business processing control unit 3. Business processing department 4. Although there are a plurality of business processing unit private memories 5, each memory has a one-to-one one-to-one correspondence.

The data transmitting/receiving unit 2 operates in response to a reception request or a transmission request from the business processing control unit 3, and receives input data from the terminal 10 and transfers it to the business processing control unit 3 at the time of a reception request, and controls business processing at the time of a transmission request. The output data delivered from the unit 3 is transmitted to the terminal 10.

The business processing control unit 3 operates as follows based on input data delivered from the data transmitting/receiving unit 2 or various notifications from other units. When input data is delivered from the data transmitting/receiving unit 2, the corresponding data is delivered as is to the business processing unit 4 for processing. When a temporary stop request is notified from the program failure detection unit 7, a stop instruction is output to the synchronization control unit 9 to temporarily stop the program. When receiving a restart instruction from the synchronization control unit 9, the operation is restarted. When an initialization request is notified from the program failure detection unit 7, the corresponding business processing unit 4 and business processing unit exclusive memory 5 are deleted from memory and re-registered from the program storage file 14, and initialization is performed in the business processing unit 4. Give setting instructions. This is an operation for returning the business processing unit 4 that has caused a program failure and the business processing unit exclusive memory 5 that has not yet been updated to an initial state.

The business processing unit 4 operates only according to instructions from the business processing control unit 3, and when input data from the terminal 10 is transferred, the business processing unit 4 operates as a business file 13. Business processing department exclusive memory 5. Then, it refers to or updates the shared memory 6, performs processing on a business basis, outputs the processing results, and requests transmission of the output data to the business processing control unit 3. Further, when an initial setting instruction is received, necessary initial values are set in the corresponding business processing unit exclusive memory 5.

One business processing dedicated memory 5 exists in each corresponding business processing section 4, and can be referenced and updated only by the corresponding business processing section 4.

The shared memory 6 can be shared by a plurality of business processing units 4 and can be referenced and updated.

The program failure detection unit 7 operates by detecting a program failure, requests the shared memory content recovery unit 8 to restore the shared memory 6, and requests the entire business processing control unit 3 to temporarily stop. In addition, initial settings are also requested to the business processing control unit 3 corresponding to the business processing unit 4 that has caused the program failure.

The shared memory content recovery unit 8 operates in response to a shared memory content recovery request from the program failure detection unit 7 and synchronizes the operating status of all business processing control units 3 with the synchronization control unit 9. Then, when the completion of the stoppage of all business processing control units 3 is notified, the shared memory 6 is restored from the shared memory content storage file 15. It also requests the synchronization control unit 9 to restart the business processing unit 3.

When the synchronization control section 9 receives a stop instruction from the business processing control section 3, it stops the corresponding business processing control section 3. Then, the suspension is canceled in response to a restart request from the shared memory content recovery unit 8, and a restart instruction is notified to the business processing control unit 3. Note that for the operation state adjustment from the shared memory content recovery unit 8, the shared memory content recovery unit 8 is temporarily stopped.
When the stoppage of all the business processing control units 3 is completed, the stoppage of the shared memory content recovery unit 8 is canceled and a notification of the completion of the stoppage is sent.

The status of the online system is constantly output to the console 11. In addition, online execution results are output to the printer 12.

The program storage file 14 stores program contents of the business processing unit 4 and the business processing unit exclusive memory 5.

The shared memory contents are stored in the shared memory contents storage file 15, and are updated by the business processing unit 4 at appropriate times.

FIG. 2 is an explanatory diagram showing the operation from occurrence to detection of a program fault. In the figure, the business processing unit 4 is the terminal 10.
If a program failure occurs while receiving input data from the computer and taking business-related measures, the program failure detection unit 7 immediately detects it and takes measures to deal with the program failure.

FIG. 3 is an explanatory diagram showing the operation after a program failure is detected. Program failure detection unit 7 that detected a program failure
makes a temporary stop request to all business processing control units 3, and makes an initialization request only to the business processing control unit 3 corresponding to the business processing unit 4 that has caused the program failure. It also issues a shared memory content recovery request to the shared memory content recovery unit 8.

The business processing control unit 3 processes the temporary stop request first in the order of requests, and outputs a stop instruction to the synchronization control unit 9 to temporarily stop the process. On the other hand, the shared memory content recovery unit 8 also makes a status inquiry to the synchronization control unit 9 and stops, but when all the business processing control units 3 have stopped, it starts operating again in response to a stop completion notification from the synchronization control unit 9, and shares Performs memory content recovery processing.

FIG. 4 is an explanatory diagram showing the process of restoring the contents of the shared memory. The shared memory content recovery unit 8 reads the shared memory content storage file 15, develops it in the shared memory 6, sets necessary initial values, and recovers the content of the shared memory.

Referring again to FIG. 3 regarding the subsequent operation, the shared memory content recovery unit 8 requests the synchronization control unit 9 to restart the entire business process control unit 3 when the shared memory content recovery process is completed. Upon receiving the restart request, the synchronization control unit 9 outputs a restart instruction to all business processing control units 3 at once. After receiving the restart instruction, the business processing control unit 3 restarts its operation and transfers the input data to the business processing unit 4, and normally online operation is continued.

Note that the business processing control unit 3 corresponding to the business processing unit 4 that has caused the program failure resumes normal online operation after performing the operation according to the initial setting request.

FIG. 5 is an explanatory diagram showing the initial setting process. The business processing control unit 3 deletes the business processing unit 4 and the business processing unit exclusive memory 5 from memory. And Brodarum storage file 1
4, register again, and output initial setting instructions. The business processing unit 4 that has received the initial setting instruction sets necessary initial values in the corresponding business processing exclusive memory 5.

As described above, a shared memory content recovery method is realized when a program failure occurs.

〔Effect of the invention〕

As explained above, even if a program failure occurs and the contents of the program's private memory and shared memory are destroyed, the memory contents can be restored and the online service can be continued without stopping the online service. This has the effect of providing high-quality, highly reliable online services without deteriorating the service.

[Brief explanation of drawings]

Fig. 1 is a configuration diagram showing an embodiment of the present invention, Fig. 2 is an explanatory diagram showing the operation from fault occurrence to detection, Fig. 3 is an explanatory diagram showing the operation after fault detection, and Fig. 4 is shared. FIG. 5 is an explanatory diagram showing the memory recovery operation, and FIG. 5 is an explanatory diagram showing the initial setting process. 2...Data transmission/reception unit, 3...Business processing control unit, 4
... Business processing section, 5... Business processing section exclusive memory, 6
...Shared memory, 7.Program failure detection unit, 8
...Shared memory content recovery section, 9...Synchronization control section, 1
0...Terminal, 13...Business file, 14...Program storage file, 15...Shared memory content storage file.

Claims (1)

[Claims] A business processing control unit that detects a program failure, requests the shared memory content recovery unit to restore the shared memory content, requests a temporary stop to multiple business process control units, and corresponds to the business processing unit that caused the program failure. A program failure detection unit requests initial settings for the program, and a plurality of business processing control units informs the shared memory content recovery unit that all operations have been temporarily stopped, and upon request from the shared memory content recovery unit, the business processing a synchronization control unit that restarts the control unit and the business processing control unit that, after all have temporarily stopped, restores the shared memory from the shared memory content storage file and requests the synchronization control unit to restart the business processing control unit; a shared memory content recovery unit to perform a shared memory content recovery unit, and when input data from a terminal is received, it is delivered as is to the business processing unit, and when a pause request is received, a stop instruction is output to the synchronization control unit to temporarily stop it, and a restart request is made. and a business processing control unit that restarts operation when receiving an initial setting request, temporarily deletes the business processing unit when receiving an initial setting request, and sets necessary initial values after re-registration. Recovery method.