JP2005242652A - Computer backup system and computer backup method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a computer backup system that can prevent a stop of computer processing, reduce the frequency of large volume data transmission to a backup computer and reduce a load on the backup computer. <P>SOLUTION: A transmission part 12d transmits update information in a storage 12 to a storage 22 in a predetermined period and transmits a master DB in the storage 12 to the storage 22 in a period other than the predetermined period. A backup master file update part 22e, if a computer system 1 fails in the predetermined period, updates the master DB transmitted from the transmission part 12d according to the update information transmitted from the transmission part 12d and later updates the master DB according to update command information, or if the computer system 1 fails in the period other than the predetermined period, updates the master DB transmitted from the transmission part 12d according to update command information. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a computer backup system and a computer backup method.

  Conventionally, measures have been proposed for preventing problems caused by computer failures.

  Patent Document 1 (Japanese Patent Application Laid-Open No. 3-271923) describes a magnetic disk device with a failure handling process, which includes a master disk for recording data processed by a computer system and a sub disk for recording backup data. Has been.

  A computer system using the magnetic disk device described in Patent Document 1 as a main recording medium can recover data by using backup data recorded on a sub disk when a failure occurs in a master disk. It becomes.

  However, in the magnetic disk device described in Patent Document 1, when a failure occurs in the computer system itself that uses the magnetic disk device, there is a problem that processing by the computer stops.

  In Patent Document 2 (Japanese Patent Laid-Open No. 2001-297011) and Patent Document 3 (Japanese Patent Laid-Open No. 6-95943), even if a computer executing the process fails, the processing by the computer stops. A computer system for preventing this is described.

  Patent Document 2 describes a non-stop job system comprising an active computer and a standby computer having the same hardware and the same software as the active computer.

  Specifically, the current computer stores processing data generated every time the current computer executes a job. The standby computer copies the processing data stored in the current computer and stores the copied processing data in itself. Therefore, the standby computer stores the backup data of the current computer. When a failure occurs in the current computer, the computer that executes the job is switched from the current computer to the standby computer.

  Patent Document 3 describes a computer backup system including a main computer and a backup computer connected to the main computer via a communication line.

Specifically, the main computer transmits an update journal file indicating only the update contents of the business master file stored therein to the backup computer. Each time a backup computer receives an update journal file, the backup computer updates the business master file built in itself based on the update journal file. Therefore, the backup computer stores the backup data of the business master file of the main computer. When a failure occurs in the main computer, the backup computer executes processing instead of the main computer.
JP-A-3-271923 JP 2001-297011 A Japanese Patent Laid-Open No. 6-95943

  In the magnetic disk device described in Patent Document 1, when a failure occurs in a computer system that uses the magnetic disk device, there is a problem that processing by the computer stops.

  In the non-stop job system described in Patent Literature 2 and the computer backup system described in Patent Literature 3, even if a failure occurs in one computer, the other computer continues to execute the processing, so the processing by the computer is stopped. This makes it possible to avoid the problem of enduring.

  However, in the non-stop job system described in Patent Document 2, since the current computer transmits processing data to the standby computer every time a job is executed, the amount of backup data transmitted to the standby computer at all times is always constant. The problem of becoming larger arises.

  Further, in the computer backup system described in Patent Document 3, every time an update journal file is received, the backup computer must update the business master file built in the backup computer file. The problem that the load becomes large arises.

  An object of the present invention is to provide a computer backup capable of reducing the frequency of transmitting backup data having a large amount of data to the backup computer without reducing processing by the computer and reducing the load on the backup computer. A system and computer backup method is provided.

  In order to achieve the above object, a computer backup system of the present invention includes a main computer, a backup computer that executes processing performed by the main computer when a failure occurs in the main computer, the main computer, and the backup In a computer backup system including an input terminal for performing input to a computer, the main computer includes a master data storage unit for storing master data and a master data stored in the master data storage unit from the input terminal. Update instruction information for updating data is received only within a predetermined period, a master data update unit that updates the master data based on the received update instruction information, and a master data update unit updated by the master data update unit Update information showing only the update part An update information storage unit to be stored, and update information stored in the update information storage unit within the predetermined period is transmitted to the backup computer and stored in the master data storage unit within a period different from the predetermined period A transmission unit that transmits the master data to the backup computer, wherein the backup computer stores the update information transmitted from the transmission unit as backup update information, and the transmission unit The master data transmitted from the backup master data storage unit for storing the master data as backup master data, the failure detection unit for detecting a failure of the main computer, and the failure detection unit within the predetermined period of time, the failure detection unit When detected, the backup update information is stored. Update the backup master data stored in the backup master data storage unit based on the backup update information stored in the backup update data, and then receive the update instruction information, the backup master data is updated based on the received update instruction information When the failure detection unit detects a failure of the main computer within a period different from the predetermined period, the update instruction information is received, and the backup master is based on the received update instruction information. And a backup master data update unit that updates the backup master data stored in the data storage unit.

  Further, the computer backup method of the present invention performs input to the main computer, a backup computer that executes processing performed by the main computer when a failure occurs in the main computer, and the main computer and the backup computer. A computer backup method performed by a computer backup system including an input terminal, wherein the main computer includes a master data storage unit that stores master data and an update information storage unit, and the backup computer includes backup update Update instruction information that includes an information storage unit and a backup master data storage unit, and that the main computer updates master data stored in the master data storage unit from the input terminal only within a predetermined period Acceptance A master data update step for updating the master data based on the attached update instruction information, and update information for storing in the update information storage unit update information indicating only an update portion of the master data updated in the master data update step And storing the update information stored in the update information storage unit within the predetermined period to the backup computer, and master data stored in the master data storage unit within a period different from the predetermined period Transmitting to the backup computer, the backup update information storing step storing the update information transmitted in the transmission step as backup update information in the backup update information storage unit, and the master transmitted in the transmission step Before data as backup master data A backup master data storage step for storing in the backup master data storage unit; a failure detection step for detecting a failure of the main computer; and a failure of the main computer detected by the failure detection step within the predetermined period Updates the backup master data stored in the backup master data storage unit based on the backup update information stored in the backup update information storage unit, then receives the update instruction information, The backup master data is updated based on the update instruction information when the failure detection step detects a failure of the main computer within a period different from the predetermined period. Based on the information of the backup master data A backup master data update step for updating the backup master data stored in the storage unit.

  According to the above invention, the main computer transmits the update information to the backup computer within a predetermined period for receiving the update instruction information, and transmits the master data to the backup computer within a period different from the predetermined period.

  Since the update information is information indicating only the update portion of the master data, the data amount of the update information is smaller than the data amount of the master data. Therefore, it is possible to reduce the frequency of transmitting data having a large amount of data to the backup computer as compared to the case where the updated master data is transmitted to the backup computer every time the master data is updated.

  In addition, since the master data is not updated during the period when the update instruction information is not received, the main computer transmits the master data at least once during the period when the update instruction information is not received, so that the latest master data is transmitted to the backup computer. It becomes possible to provide. Therefore, it is possible to reduce the frequency of transmitting data with a large amount of data to the backup computer.

  The backup computer accepts update instruction information when the failure detection unit detects a failure of the main computer within a period different from a predetermined period, and updates the backup master data based on the received update instruction information. To do. Therefore, the backup computer does not need to update the backup master data based on the backup update information when a failure of the main computer is detected by the failure detection unit within a period different from the predetermined period. The load on the backup computer is reduced.

  In the computer backup system, the main computer includes a main server and a first storage device. The main server includes a transmission operation control unit that controls the operation of the transmission unit, and the master data update unit. The first storage device includes the master data storage unit, the update information storage unit, and the transmission unit, and the backup computer includes a backup server, a second storage device, and the like. The backup server includes the failure detection unit, and an update operation control unit that controls the operation of the backup master data update unit based on a detection result of the failure detection unit, and the second storage device Includes the backup update information storage unit, the backup master data storage unit, and the backup master data update unit. It is desirable

  According to the above invention, transmission of update information, transmission of master data, and update of backup master data are performed by the first and second storage devices based on the control of the main server and the backup server. Therefore, the operation of the computer backup system can be centrally managed by the main server and the backup server, the inconsistency between the data of the main computer and the data of the backup computer can be prevented, and the backup operation can be automated.

Further, since the first storage device transmits information, it is not necessary to use a computer dedicated to information transmission and an application for information transmission. In addition, it is possible to eliminate the need for testing information transmission computers and information transmission applications when the system is modified, and the system test man-hours can be greatly reduced.
In the computer backup system, the main server transmits a main computer failure detection unit that detects a failure of the main computer, and a failure of the main computer detected by the main computer failure detection unit to the backup computer. A failure detection transmission unit, and the update operation control unit is stored in the backup update information storage unit when a failure of the main computer is transmitted from the failure detection transmission unit within the predetermined period. Updating the backup master data stored in the backup master data storage unit based on the backup update information, and then receiving the update instruction information, updating the backup master data based on the received update instruction information, The failure is within a period different from the predetermined period. When a failure of the main computer is transmitted from the detection transmission unit, the update instruction information is received, and the backup master data stored in the backup master data storage unit is updated based on the received update instruction information Thus, it is desirable to control the backup master data update unit.

  According to the above invention, the main computer can notify the backup computer of its own failure, and the backup computer performs the processing performed by the main computer when receiving the notification of the failure of the main computer from the main computer. It becomes possible to take over.

  In the computer backup system, the main computer failure detection unit registers the error information of the main server, the first storage device, and the master data with each other, and the main server and the first When an error that occurs in each of the storage device and the master data corresponds to registered error information, it is desirable to detect a failure of the main computer.

  Further, in the computer backup method, the failure detection step includes a case where each error of the main computer, the master data storage unit, and the update information storage unit corresponds to error information registered in advance in a registration unit, It is desirable to detect a failure of the primary computer.

  According to the above invention, an error can be checked from each of a plurality of viewpoints (for example, the main server, the first storage device, and the master data). For this reason, it becomes possible to perform fine-grained switching of computers for a specific failure.

  According to the present invention, the main computer transmits the update information to the backup computer within a predetermined period for receiving the update instruction information, and transmits the master data to the backup computer within a period different from the predetermined period. For this reason, it is possible to reduce the frequency of transmitting data having a large amount of data to the backup computer as compared to the case where the updated master data is transmitted to the backup computer every time the master data is updated.

  In addition, since the master data is not updated during the period when the update instruction information is not received, the main computer transmits the master data at least once during the period when the update instruction information is not received, so that the latest master data is transmitted to the backup computer. It becomes possible to provide. Therefore, it is possible to reduce the frequency of transmitting data with a large amount of data to the backup computer.

  The backup computer does not need to update the backup master data based on the backup update information when a failure of the main computer is detected by the failure detection unit within a period different from the predetermined period. The load of is reduced.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing a computer backup system according to an embodiment of the present invention.

  In FIG. 1, the computer backup system includes a production computer system 1 as a main computer and a remote standby computer system 2 as a backup computer of the computer system 1.

  The computer system 1 includes a server 11 as a main server and a storage 12 as a first storage device. The server 11 includes a memory (not shown) that stores a program that defines the operation of the server 11, and a computer (not shown) that executes various functions by executing the program stored in the memory in the server 11. Including.

  The server 11 includes an operation management subsystem 11a, an operation unit 11b, and an output unit 11c. The operation management subsystem 11a is realized by a computer in the server 11 executing a program stored in a memory in the server 11.

  The operation management subsystem 11a includes an operation control unit 11a1, a master data update unit 11a2, a reception unit 11a3, a main computer failure detection unit 11a4, a failure detection transmission unit 11a5, and an internal clock 11a6.

  The storage 12 includes a standard disk 12a, a duplicate volume disk 12b, a disk synchronization unit 12c, and a transmission unit 12d. The standard disk 12a includes a master DB storage unit 12a1 and an update journal storage unit 12a2. The duplicate volume disk 12b includes a master DB storage unit 12b1 and an update journal storage unit 12b2. The disk synchronization unit 12c instantly connects the standard disk 12a and the duplicate volume disk 12b, makes the stored contents of the duplicate volume disk 12b equal (synchronizes) with the stored contents of the standard disk 12a, and the standard disk 12a. The duplicate volume disk 12b can be instantaneously disconnected.

  The computer system 2 includes a server 21 as a backup server and a storage 22 as a second storage device. The server 21 includes a memory (not shown) that stores a program that defines the operation of the server 21, and a computer (not shown) that executes various functions by executing the program stored in the memory in the server 21. Including.

  The server 21 includes an operation management subsystem 21a and an operation unit 21b. The operation management subsystem 21a is realized by a computer in the server 21 executing a program stored in a memory in the server 21.

  The operation management subsystem 21a includes a failure detection unit 21a1, an operation control unit 21a2, a reception unit 21a3, and an internal clock 21a4.

  The storage 22 includes a standard disk 22a, a duplicate volume disk 22b, a disk synchronization unit 22c, a reception unit 22d, and a backup master data update unit 22e. The standard disk 22a includes a master DB storage unit 22a1 and an update journal storage unit 22a2. The duplicate volume disk 22b includes a master DB storage unit 22b1 and an update journal storage unit 22b2.

  The server 11 is connected to the server 21 via a communication line such as WAN. The storage 12 is connected to the storage 22 via a communication line such as WAN.

  The operation management subsystem 11 a and the operation management subsystem 21 a have a health check function of the server 11. The operation management subsystem 11a has a function of outputting an instruction for switching to another system in the event of a failure of itself, and the operation management subsystem 21a has a function of switching to the own system.

  The master DB storage unit 12a1 stores a master database (master DB) as master data.

  The operation unit 11b such as a keyboard receives input from the user.

  For example, the operation unit 11b receives update instruction information input by a user operation. In addition, the operation unit 11b receives period information indicating a period in which the server 11 receives update instruction information.

  The reception unit 11a3 receives the update instruction information and the period information input to the operation unit 11b. The receiving unit 11a3 receives update instruction information and period information input from the input terminal 3 connected to the server 11 via a communication line.

  The update instruction information received by the receiving unit 11a3 is update instruction information for updating the master DB stored in the master DB storage unit 12a1. The period during which the server 11 receives the update instruction information is arbitrarily set by the user operating the operation unit 11b or the input terminal 3.

  When the time measured by the internal clock 11a6 is within the period indicated by the period information received by the reception unit 11a3, the master data update unit 11a2 receives the update instruction information received by the reception unit 11a3 and is based on the received update instruction information. The master DB stored in the master DB storage unit 12a1 is updated.

  Furthermore, when the master data update unit 11a2 updates the master DB stored in the master DB storage unit 12a1, update information (hereinafter referred to as “update journal file”) showing only the updated part of the updated master DB. Is stored in the update journal storage unit 12a2.

  When the time measured by the internal clock 11a6 is within the period indicated by the period information received by the receiving unit 11a3, the operation control unit 11a1 determines that the same update journal file as the update journal file stored in the update journal storage unit 12a2 The disk synchronization unit 12c is operated so as to be stored in the storage unit 12b2. For example, the operation control unit 11a1 operates the disk synchronization unit 12c by transmitting an API (Application Program Interface) command to the disk synchronization unit 12c.

  Further, when the time measured by the internal clock 11a6 is within the period indicated by the period information received by the receiving unit 11a3, the operation control unit 11a1 has the same update journal file as the update journal file stored in the update journal storage unit 12b2. The transmission unit 12d is operated so as to be stored in the update journal storage unit 22b2. For example, the operation control unit 11a1 operates the transmission unit 12d by transmitting an API command to the transmission unit 12d.

  When the time measured by the internal clock 11a6 is within the period indicated by the period information received by the receiving unit 11a3 under the control of the operation control unit 11a1, the transmitting unit 12d and the update journal storage unit 12b2 and the updating journal are received via the receiving unit 22d. The storage unit 22b2 is connected, and the update journal file stored in the update journal storage unit 12b2 is transmitted to the storage 22 within the period indicated by the period information received by the reception unit 11a3.

  For example, when the update journal storage unit 12b2 and the update journal storage have a plurality of storage tracks, and the update journal file is stored in these storage tracks, the transmission unit 12d stores the update journal storage unit 12b2. The update journal file that is not stored in the update journal storage unit 22b2 is transferred to the storage 22 in units of storage tracks.

  Therefore, the transfer journal file can be transferred by the function of the storage 12, and it is not necessary to use a transmission-dedicated computer and a transmission application when transferring the update journal file. Therefore, it is possible to greatly reduce the man-hours for system testing at the time of business modification.

  The transmission unit 12d controls the master DB storage unit 12b1 and the master DB storage unit 22b1 when the time measured by the internal clock 11a6 is within the period indicated by the period information received by the reception unit 11a3 under the control of the operation control unit 11a1. Disconnect the.

  The disk synchronization unit 22c operates so that the same update journal file as the update journal file stored in the update volume storage unit 22b2 is stored in the update journal storage unit 22a2.

  Further, when the time measured by the internal clock 11a6 is outside the period indicated by the period information received by the receiving unit 11a3, the operation control unit 11a1 determines that the same master DB as the master DB stored in the master DB storage unit 12a1 is the master DB. The disk synchronization unit 12c is operated so as to be stored in the storage unit 12b1. For example, the operation control unit 11a1 operates the disk synchronization unit 12c by transmitting an API command to the disk synchronization unit 12c.

  The operation control unit 11a1 operates the disk synchronization unit 12c so that the same master DB as the master DB stored in the master DB storage unit 12a1 is stored in the master DB storage unit 12b1, and then the update journal storage unit It is desirable to clear the update journal file stored in 12a2.

  Further, when the time measured by the internal clock 11a6 is outside the period indicated by the period information received by the receiving unit 11a3, the operation control unit 11a1 determines that the same master DB as the master DB stored in the master DB storage unit 12b1 is the master DB. The transmission unit 12d is operated so as to be stored in the storage unit 22b1. For example, the operation control unit 11a1 operates the transmission unit 12d by transmitting an API command to the transmission unit 12d.

  The transmission unit 12d controls the master DB storage unit 12b1 and the master DB via the reception unit 22d when the time measured by the internal clock 11a6 is outside the period indicated by the period information received by the reception unit 11a3 under the control of the operation control unit 11a1. The storage unit 22b1 is connected, and the master DB stored in the master DB storage unit 12b1 is transmitted to the storage 22.

  The transmission unit 12d controls the update journal storage unit 12b2 and the update journal storage unit 22b2 when the time measured by the internal clock 11a6 is outside the period indicated by the period information received by the reception unit 11a3 under the control of the operation control unit 11a1. Disconnect the.

  The disk synchronization unit 22c operates so that the same master DB as the master DB stored in the master DB storage unit 22b1 is stored in the master DB storage unit 22a1.

  The main computer failure detection unit 11a4 detects a failure of the computer system 1. For example, the main computer failure detection unit 11a4 registers error information specific to each of the server level, storage level, and DB level in the main computer failure detection unit 11a4 in advance. The main computer failure detection unit 11a4 acquires and monitors error log information at each of the server level, the storage level, and the DB level, and when the same error log information as the error information registered in advance appears, the computer The system 1 detects that a failure has occurred.

  The failure detection transmission unit 11a5 transmits the failure of the computer system 1 detected by the main computer failure detection unit 11a4 to the server 21 via the output unit 11c. Furthermore, the failure detection transmission unit 11a5 provides a switching instruction that instructs the server 21 to execute processing on behalf of the server 11 when a failure of the computer system 1 is detected by the main computer failure detection unit 11a4. To 21.

  The output unit 11c transmits the period information received by the receiving unit 11a3 to the server 21.

  The failure detection unit 21a1 detects a failure in the computer system 1.

  When the failure detection unit 21a1 detects a failure in the computer system 1 within the period indicated by the period information supplied from the server 11, the operation control unit 21a2 updates the update journal file stored in the update journal storage unit 22a2. Based on the above, the master DB stored in the master DB storage unit 22a1 is updated. By this update, the master DB stored in the master DB storage unit 22a1 becomes the latest state. Thereafter, the operation control unit 21a2 receives update instruction information supplied from the input terminal 3 or the like via the reception unit 21a3, and updates the master DB based on the received update instruction information.

  In addition, when the failure detection unit 21a1 detects a failure in the computer system 1 outside the period indicated by the period information supplied from the server 11, the operation control unit 21a2 receives the reception unit 21a3 from the input terminal 3 or the like. The update instruction information supplied via the master DB is received, and the master DB is updated based on the received update instruction information.

  The operation control unit 21a2 also updates the update journal file stored in the update journal storage unit 22a2 when a failure of the computer system 1 is transmitted from the server 11 within the period indicated by the period information supplied from the server 11. Based on the above, the master DB stored in the master DB storage unit 22a1 is updated. By this update, the master DB stored in the master DB storage unit 22a1 becomes the latest state. Thereafter, the operation control unit 21a2 receives update instruction information supplied from the input terminal 3 or the like via the reception unit 21a3, and updates the master DB based on the received update instruction information.

  In addition, when the failure of the computer system 1 is transmitted from the server 11 outside the period indicated by the period information supplied from the server 11, the operation control unit 21 a 2 receives from the input terminal 3 or the like via the reception unit 21 a 3. The supplied update instruction information is received, and the master DB is updated based on the received update instruction information.

  Next, the operation will be described.

  It is assumed that the period set by the period information is set to an online period in which the computer system 1 is connected to the input terminal 3.

  First, the operation during the online period when no failure has occurred in the computer system 1 will be described.

  When the measured time of the internal clock 11a6 is within the period indicated by the period information received by the receiving unit 11a3, the master file update unit 11a2 receives the update instruction information received by the receiving unit 11a3, and based on the received update instruction information The master DB stored in the master DB storage unit 12a1 is updated.

  Furthermore, when the master file update unit 11a2 updates the master DB stored in the master DB storage unit 12a1, the update journal file indicating only the updated part of the updated master DB is sequentially stored in the update journal storage unit 12a2. To do.

  When the time measured by the internal clock 11a6 is within the period indicated by the period information received by the receiving unit 11a3, the operation control unit 11a1 determines that the same update journal file as the update journal file stored in the update journal storage unit 12a2 The disk synchronization unit 12c is operated so as to be stored in the storage unit 12b2. For example, the operation control unit 11a1 operates the disk synchronization unit 12c by issuing an API command to the disk synchronization unit 12c.

  Further, when the time measured by the internal clock 11a6 is within the period indicated by the period information received by the receiving unit 11a3, the operation control unit 11a1 has the same update journal file as the update journal file stored in the update journal storage unit 12b2. The transmission unit 12d is operated so as to be stored in the update journal storage unit 22b2. For example, the operation control unit 11a1 operates the transmission unit 12d by issuing an API command to the transmission unit 12d.

  The disk synchronization unit 22c operates so that the same update journal file as the update journal file stored in the update volume storage unit 22b2 is stored in the update journal storage unit 22a2.

  Therefore, in the online period when no failure has occurred in the computer system 1, the stored content of the update journal storage unit 12a2, the stored content of the update journal storage unit 12b2, the stored content of the update journal storage unit 22b2, and the update journal The contents stored in the storage unit 22a2 are the same.

  Next, the operation during a non-online period (for example, at night) when no failure has occurred in the computer system 1 will be described.

  When the time measured by the internal clock 11a6 is outside the period indicated by the period information received by the receiving unit 11a3, the master data updating unit 11a2 does not receive the update instruction information supplied from the receiving unit 11a3. Therefore, the master DB stored in the master DB storage unit 12a1 is not updated during this period.

  The operation control unit 11a1 stores the same master DB as the master DB stored in the master DB storage unit 12a1 as soon as the time measured by the internal clock 11a6 is outside the period indicated by the period information received by the reception unit 11a3. The disk synchronization unit 12c is operated so as to be stored in the unit 12b1. For example, the operation control unit 11a1 operates the disk synchronization unit 12c by issuing an API command to the disk synchronization unit 12c.

  Further, as soon as the time measured by the internal clock 11a6 is outside the period indicated by the period information received by the receiving unit 11a3, the operation control unit 11a1 acquires the same master DB as the master DB stored in the master DB storage unit 12b1. The transmission unit 12d is operated so as to be stored in the DB storage unit 22b1. The operation control unit 11a1 operates the transmission unit 12d by issuing an API command to the transmission unit 12d.

  Therefore, in the non-online period when no failure has occurred in the computer system 1, the stored content of the master DB storage unit 12a1, the stored content of the master DB storage unit 12b1, the stored content of the master DB storage unit 22b1, and the master The stored contents of the DB storage unit 22a1 are the same.

  Thereafter, the operation control unit 11a1 clears the update journal file stored in the update journal storage unit 12a.

  Note that the transmission unit 12d desirably transmits the master DB to the storage 22 only once during the time period outside the period indicated by the period information received by the reception unit 11a3. In this case, since the master DB is not updated while the time measured by the internal clock 11a6 is outside the period indicated by the period information received by the receiving unit 11a3, if the master DB is transmitted to the storage at least once during this period, This is because the latest master DB can be provided. Therefore, it is possible to reduce the frequency of transmitting the master DB having a large amount of data to the backup computer.

  Next, an operation when a failure occurs in the computer system 1 within the online period will be described.

  When a failure of the computer system 1 is detected by the failure detection unit 21a1 within the period indicated by the period information supplied from the server 11, the operation control unit 21a2 switches the server 21 to a production server.

  Further, the operation control unit 21a2 updates the master DB stored in the master DB storage unit 22a1 based on the update journal file stored in the update journal storage unit 22a2. Thereafter, the operation control unit 21a2 receives update instruction information supplied from the input terminal 3 or the like via the reception unit 21a3, and updates the master DB based on the received update instruction information.

  When a failure of the computer system 1 is transmitted from the server 11 within the period indicated by the period information supplied from the server 11, the operation control unit 21 a 2 switches the server 21 to a production server.

  Further, the operation control unit 21a2 updates the master DB stored in the master DB storage unit 22a1 based on the update journal file stored in the update journal storage unit 22a2. Thereafter, the operation control unit 21a2 receives update instruction information supplied from the input terminal 3 or the like via the reception unit 21a3, and updates the master DB based on the received update instruction information.

  Next, an operation when a failure occurs in the computer system 1 within the non-online period will be described.

  When a failure of the computer system 1 is detected by the failure detection unit 21a1 outside the period indicated by the period information supplied from the server 11 (non-online period), the operation control unit 21a2 sets the server 21 as a production server Switch to. Furthermore, when the period indicated by the period information supplied from the server 11 is reached, the operation control unit 21a2 receives the update instruction information supplied from the input terminal 3 or the like via the reception unit 21a3, and the received update instruction information Based on this, the master DB stored in the master DB storage unit 22a1 is updated.

  Further, when a failure of the computer system 1 is transmitted from the server 11 outside the period indicated by the period information supplied from the server 11, the operation control unit 21a2 switches the server 21 to a production server. Furthermore, when the period indicated by the period information supplied from the server 11 is reached, the operation control unit 21a2 receives the update instruction information supplied from the input terminal 3 or the like via the reception unit 21a3, and the received update instruction information Based on this, the master DB stored in the master DB storage unit 22a1 is updated.

  The computer system 2 does not need to update the master DB based on the update journal file when a failure occurs in the computer system 1 within the non-online period, so the load on the computer system 2 is reduced.

  According to the present embodiment, the following effects can be obtained.

  The first effect is that control of data synchronization at the level between storages for production switching between remote locations, control at the level between servers, and recovery processing at the DB level (using an update journal file) Master DB update) can be centrally managed in the operation management subsystem. By centrally managing the data synchronization and recovery processing at the storage level, server level, and DB level, data inconsistency can be prevented and automatically handled.

  The second effect is that it is possible to simplify trouble handling training and education for the operation manager. The reason is that the computer backup system automatically performs backup processing, so even if the operation administrator does not have to master storage technology, DB technology, and server business technology in advance, it will acquire a certain level of technology. This is because the computer backup system can be supported.

  The third effect is that the production system switching process can be performed for detailed specific failures at the server level, storage level, and DB level in the operation management subsystem. The reason is that in the event of a specific failure between servers, the other server is instructed to switch to the production system, which automatically switches to the production system at the storage level and DB level, and the operation management subsystem. In this case, the specific error information at the server level, storage level, and DB level is registered in advance, and the error log information at each level is acquired and monitored. This is because the system switching instruction is issued via the network.

  The fourth effect can be realized by the functions of the storage 12 and the storage 22 for transferring data between the storage 12 and the storage 22. Therefore, it is not necessary to use a transmission-dedicated computer and a transmission application when transferring data between storages. Therefore, it is possible to greatly reduce the man-hours for system testing at the time of business modification.

  The fifth effect is that the frequency of transmitting backup data with a large amount of data to the backup computer can be reduced and the load on the backup computer can be reduced without stopping the processing by the computer. It is.

  In the embodiment described above, the illustrated configuration is merely an example, and the present invention is not limited to the configuration.

  For example, the computer system 1 may not include the main computer failure detection unit and the failure detection transmission unit.

  Further, when the computer system 1 is restored, data may be returned from the storage 22 to the storage 12. In this case as well, data can be transferred by the storage function.

It is the block diagram which showed the computer backup system of one Example of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Computer system 11 Server 11a Operation management subsystem 11a1 Operation control part 11a2 Master file update part 11a3 Reception part 11a4 Main computer failure detection part 11a5 Failure detection transmission part 11a6 Internal clock 11b Operation part 11c Output part 12 Storage 12a Standard disk 12a1 Master DB Storage unit 12a2 Update journal storage unit 12b Replica volume disk 12b1 Master DB storage unit 12b2 Update journal storage unit 12c Disk synchronization unit 12d Transmission unit 2 Computer system 21 Server 21a Operation management subsystem 21a1 Failure detection unit 21a2 Operation control unit 21a3 Accepting unit 21a4 Internal clock 21b Operation unit 22 Storage 22a Standard disk 22a1 Master DB storage unit 22a2 Update journal Part 22b copy volume disc 22b1 master DB storage section 22b2 update journal storage unit 22c disk synchronization portion 22d receiving portion 22e backup master file updater

Claims (6)

In a computer backup system comprising: a main computer; a backup computer that executes processing performed by the main computer when a failure occurs in the main computer; and an input terminal that performs input to the main computer and the backup computer ,
The main computer is
A master data storage for storing master data;
Master data for receiving update instruction information for updating the master data stored in the master data storage unit from the input terminal only within a predetermined period and updating the master data based on the received update instruction information Update section,
An update information storage unit for storing update information indicating only an update part of the master data updated by the master data update unit;
The update information stored in the update information storage unit is transmitted to the backup computer within the predetermined period, and the master data stored in the master data storage unit is transmitted to the backup computer within a period different from the predetermined period. A transmission unit for transmitting to a computer,
The backup computer is:
A backup update information storage unit that stores the update information transmitted from the transmission unit as backup update information;
A backup master data storage unit for storing the master data transmitted from the transmission unit as backup master data;
A failure detection unit for detecting a failure of the main computer;
When a failure of the main computer is detected by the failure detection unit within the predetermined period, the failure is stored in the backup master data storage unit based on the backup update information stored in the backup update information storage unit When the backup master data is updated and then the update instruction information is received, the backup master data is updated based on the received update instruction information, and the failure detection unit performs the main computer within a period different from the predetermined period. A backup master data update unit that receives the update instruction information and updates the backup master data stored in the backup master data storage unit based on the received update instruction information. Including a computer backup system Temu. The computer backup system according to claim 1,
The main computer includes a main server and a first storage device,
The main server includes a transmission operation control unit that controls the operation of the transmission unit, and the master data update unit,
The first storage device includes the master data storage unit, the update information storage unit, and the transmission unit,
The backup computer includes a backup server and a second storage device,
The backup server includes the failure detection unit, and an update operation control unit that controls the operation of the backup master data update unit based on a detection result of the failure detection unit,
The computer storage system, wherein the second storage device includes the backup update information storage unit, the backup master data storage unit, and the backup master data update unit. The computer backup system according to claim 2,
The main server is
A main computer failure detection unit for detecting a failure of the main computer;
A failure detection transmission unit that transmits a failure of the main computer detected by the main computer failure detection unit to the backup computer;
The update operation control unit, when a failure of the main computer is transmitted from the failure detection transmission unit within the predetermined period, the backup operation information based on the backup update information stored in the backup update information storage unit Update the backup master data stored in the master data storage unit, then receive the update instruction information, update the backup master data based on the received update instruction information, and within the period different from the predetermined period When a failure of the main computer is transmitted from the failure detection transmission unit, the update instruction information is received, and the backup master data stored in the backup master data storage unit is updated based on the received update instruction information The computer that controls the backup master data update unit Data backup system. The computer backup system according to claim 3,
The main computer failure detection unit registers error information of each of the main server, the first storage device, and the master data, and stores the error information in each of the main server, the first storage device, and the master data. A computer backup system for detecting a failure of the main computer when an error that occurs corresponds to registered error information. A computer backup system including a main computer, a backup computer that executes processing performed by the main computer when a failure occurs in the main computer, and an input terminal that inputs data to the main computer and the backup computer A computer backup method to perform,
The main computer includes a master data storage unit that stores master data and an update information storage unit, and the backup computer includes a backup update information storage unit and a backup master data storage unit,
The main computer receives update instruction information for updating the master data stored in the master data storage unit from the input terminal only within a predetermined period, and the master data is received based on the received update instruction information. A master data update step to be updated;
An update information storage step of storing update information indicating only an update part of the master data updated in the master data update step in the update information storage unit;
The update information stored in the update information storage unit is transmitted to the backup computer within the predetermined period, and the master data stored in the master data storage unit is transmitted to the backup computer within a period different from the predetermined period. A sending step for sending to a computer;
A backup update information storage step of storing the update information transmitted in the transmission step in the backup update information storage unit as backup update information;
A backup master data storage step of storing the master data transmitted in the transmission step in the backup master data storage unit as backup master data;
A failure detection step of detecting a failure of the main computer;
When a failure of the main computer is detected in the failure detection step within the predetermined period, the failure is stored in the backup master data storage unit based on the backup update information stored in the backup update information storage unit Update the backup master data, then receive the update instruction information, update the backup master data based on the received update instruction information, and in the failure detection step within the period different from the predetermined period, the main computer A backup master data update step of receiving the update instruction information when a failure is detected, and updating the backup master data stored in the backup master data storage unit based on the received update instruction information Computer backup method. The computer backup method according to claim 5, wherein
The failure detection step detects a failure of the main computer when errors in the main computer, the master data storage unit, and the update information storage unit correspond to error information registered in advance in a registration unit, Computer backup method.

Images