JP2014106580A - Information processing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem of being unable to combine reduction of load and assurance of consistency of transaction processing.SOLUTION: An information processing device includes: update means for performing predetermined update processing on a master database to update a database stored in the master database; update information extraction means for extracting update information based on the database updated by the update means; and update information transfer means for including a plurality of update information extracted by the update information extraction means into one update information group and transferring the update information group at a time to a replica database in which replicated data of a database stored in the master database is stored. The update information transfer means includes updated a plurality of update information updated by the same transaction processing that manages a plurality of related update processes as a sequence of processing into one update information group and transfers the update information group to the replica database at a time.

Description

  The present invention relates to an information processing apparatus, and more particularly to an information processing apparatus that transfers update information when data is updated to another storage device.

  In order to provide redundancy to a database, a replication system including a master database that stores data and a replica database that stores replica data of the master database is known (see, for example, Patent Document 1). In a replication system, when a master database is updated, update information associated with the update is extracted, and the extracted update information is transferred to a replica database, thereby realizing database replication.

JP 2002-318718 A

  Here, the master database may be updated by a transaction process that manages a plurality of related processes as a series of processes. In the replication system described above, consistency between the master database and the replica database can be ensured by transferring update information to the replica database one by one for each transaction process. However, in the case of transferring update information for each transaction process, the update information transfer count increases as the number of updates to the master database increases, the master database side process, the replica database side process, the network, There was a problem that the load of each increased.

  In the replication system described above, a process for transferring update information when the number of update information extracted by updating the master database reaches a fixed number is known in order to suppress the load. However, when the update information is transferred when the number of update information reaches a certain number, all of the update information based on the same transaction processing may not be transferred at a single timing. As a result, there is a problem in that consistency between the master database and the replica database cannot be temporarily achieved, and the consistency of transaction processing of updates performed on the master database cannot be guaranteed in the replica database. It was.

  Therefore, an object of the present invention is to provide an information processing apparatus that can solve the problem that it is impossible to achieve both the suppression of load and the guarantee of transaction processing consistency. .

In order to achieve the above object, an information processing apparatus according to an aspect of the present invention provides:
Updating means for performing a predetermined update process on the master database and updating the database stored in the master database;
Update information extraction means for extracting update information based on the database updated by the update means;
Update information transfer means for transferring update information groups, which are a plurality of update information extracted by the update information extraction means, to a replica database that stores duplicate data of the database stored in the master database at a single timing When,
With
The update information transfer means manages a plurality of related update processes as a series of processes, and includes the update information extracted based on the database updated by the same transaction process in the update information group.
Take the configuration.

Moreover, the program which is the other form of this invention is:
In the information processing device,
Updating means for performing a predetermined update process on the master database and updating the database stored in the master database;
Update information extraction means for extracting update information based on the database updated by the update means;
Update information transfer means for transferring update information groups, which are a plurality of update information extracted by the update information extraction means, to a replica database that stores duplicate data of the database stored in the master database at a single timing When,
And realize
The update information transfer means manages a plurality of related update processes as a series of processes, and includes the update information extracted based on the database updated by the same transaction process in the update information group.
It is a program for realizing this.

In addition, an information processing method according to another aspect of the present invention includes:
Perform a predetermined update process on the master database, update the database stored in the master database,
Extract update information based on the updated database,
The update information group including the update information extracted based on the database updated by the same transaction process, which manages a plurality of related update processes as a series of processes, of the database stored in the master database Transfer to the replica database that stores the replicated data once
Take the configuration.

  By configuring as described above, the present invention has an excellent effect that consistency of transaction processing of updates performed on the master database can be guaranteed in the replica database while reducing the load. .

It is a block diagram which shows the outline of a structure of the information processing system in this invention. It is a block diagram which shows the structure of a master server. It is a block diagram which shows the structure of a slave server. It is a functional block diagram which shows the structure of the information processing system in 1st Embodiment of this invention. It is a figure for demonstrating the example of an update log intermediate | middle table. It is a flowchart for demonstrating an update process. It is a flowchart for demonstrating update information extraction processing. It is a flowchart for demonstrating update information control processing. It is a flowchart for demonstrating update information transfer processing. It is a flowchart for demonstrating a reflection process. It is a flowchart for demonstrating a batch reflection process. It is a functional block diagram which shows the structure of the master server in 2nd Embodiment of this invention.

<First Embodiment>
A first embodiment of the present invention will be described with reference to FIGS. 1 to 5 are diagrams for explaining the configuration of the information processing system. 6 to 11 are diagrams for explaining the operation of the information processing system.

[Constitution]
First, the configuration of the information processing system 1 will be described. As shown in FIG. 1, the information processing system 1 in this embodiment includes a master server 11, a slave server 12, and a communication network 13. The master server 11 (information processing apparatus) includes a master DB (database) 21 that stores (stores) master data. Further, the slave server 12 includes a replica DB 22 that stores duplicate data of the master data stored in the master DB 21. Furthermore, the master server 11 and the slave server 12 are connected via a wired or wireless communication network 13.

  The information processing system 1 controls so that the contents of the master data stored in the master DB 21 and the contents of the replicated data stored in the replica DB 22 are always synchronized. Thereby, when a failure or the like occurs in the master server 11, the slave server 12 functions in place of the master server 11, so that the functions of the entire information processing system 1 can be maintained. In addition, although one slave server 12 is described in the information processing system 1 in FIG. 1, the information processing system 1 may include two or more slave servers 12.

  Next, the configuration of the master server 11 and the slave server 12 will be described. First, the configuration of the master server 11 will be described with reference to FIG. FIG. 2 is a block diagram illustrating a configuration example of the master server 11. As shown in FIG. 2, the master server 11 includes a calculation unit 41, a storage unit 42, an input / output unit 43, and a communication unit 44.

  The computing unit 41 is configured by, for example, a CPU (Central Processing Unit). The calculation unit 41 reads various programs stored in the storage unit 42 and executes the programs to execute various functions for controlling the master server 11. The functional configuration of the calculation unit 41 will be described later with reference to FIG.

  The storage unit 42 has a master DB 21 that stores master data. The master data is, for example, a database that manages bank accounts, a database that manages product orders and delivery, and the like. In particular, in the present embodiment, the master data is updated by a transaction process that manages a plurality of related update processes as a series of processes.

  The input / output unit 43 receives input of database reflection information for updating the master DB 21. When the input / output unit 43 receives an input of database reflection information, the update unit 71 described later updates the master DB 21 based on the database reflection information received by the input / output unit 43. Subsequently, the update log extraction unit 72 described later extracts update information (update log) based on the update of the master DB 21. Then, the communication unit 44 transfers the generated update information to the slave server 12 via the communication network 13.

  Next, the configuration of the slave server 12 will be described with reference to FIG. FIG. 3 is a block diagram illustrating a configuration example of the slave server 12. As shown in FIG. 3, the slave server 12 includes a calculation unit 51, a storage unit 52, and a communication unit 53. In addition, the calculation unit 51, the storage unit 52, and the communication unit 53 of the slave server 12 have the same configuration as the calculation unit 41, the storage unit 42, and the communication unit 44 of the master server 11. Therefore, differences from the master server 11 will be described.

  The storage unit 52 includes a replica DB 22 that stores duplicate data of master data stored in the master DB 21. The communication unit 53 receives the update information transferred from the master server 11. When the communication unit 53 receives the update information, an update log reflection unit 91 described later reflects the received update information on the replica DB 22 and updates the replica DB 22. Thus, since the content stored in the master DB 21 is reflected in the content stored in the replica DB 22, a replication function can be realized in the information processing system 1 and the entire function of the information processing system 1 can be maintained.

  Next, with reference to FIG. 4, the detail of each structure of the master server 11 and the slave server 12 is demonstrated. FIG. 4 is a block diagram illustrating a functional configuration example of the master server 11 and the slave server 12.

  The calculation unit 41 of the master server 11 executes a program to update the unit 71 (update unit), the update log extraction unit 72 (update information extraction unit), and the update log control unit 73 (update information control unit). , An update log transfer unit 74 (update information transfer unit) and a table acquisition unit 75 (database transfer control unit). The storage unit 42 of the master server 11 includes a master DB 21 and an update log intermediate table A81. The update log intermediate table A81 will be described later with reference to FIG.

  Further, the calculation unit 51 of the slave server 12 includes an update log reflection unit 91, a batch reflection control unit 92, and a batch reflection execution unit 93 by executing a program. Furthermore, the storage unit 52 of the slave server 12 includes a replica DB 22 and an update log intermediate table B101. The update log intermediate table B101 of the slave server 12 has the same configuration as the update log intermediate table A81 of the master server 11. Hereinafter, specific processing of the master server 11 and the slave server 12 will be described.

  First, a process in which the master server 11 updates the master data in the master DB 21 will be described. The update unit 71 performs a predetermined update process on the master DB 21 to update the database stored in the master DB 21. Specifically, first, the update unit 71 starts transaction processing. That is, the update unit 71 declares the start of a transaction and starts update processing. Subsequently, the update unit 71 acquires database reflection information for updating the master DB 21 via the input / output unit 43. The database reflection information is generated, for example, when the user performs an operation of transferring money to a bank account.

  And the update part 71 starts the update of the master data memorize | stored in master DB21 based on the acquired database reflection information. At this time, the update unit 71 issues a trigger to the update log extraction unit 72. This trigger is information for causing the update log extraction unit 72 to start the extraction process, and is issued for each update process. Specifically, the update unit 71 issues a trigger every time predetermined data included in the master data is updated based on the database reflection information. For example, when moving a predetermined amount (for example, 100,000 yen) from Mr. a's bank account to Mr. b's bank account, processing (update processing) for subtracting 100,000 yen from the amount of Mr. a's bank account has started. Sometimes the update unit 71 issues a trigger. And when the process which adds 100,000 yen to the amount of a bank account of b is started, the update part 71 will issue a trigger again.

  Subsequently, the update log extraction unit 72 starts the update information extraction process based on the database updated by the update unit 71 by receiving the trigger issued by the update unit 71. Then, each time the update log extraction unit 72 receives a trigger, the update log extraction unit 72 updates the update information based on the master data before update stored in the master DB 21 and the updated master data updated based on the update information. (Update log) is extracted, and the extracted update information is stored in the update log intermediate table A81. Here, the details of the update log intermediate table A81 will be described with reference to FIG.

  FIG. 5 is a diagram for explaining the details of the update log intermediate table A81. As illustrated in FIG. 5, the update log intermediate table A81 includes a trigger ID (Identification), a time, a transaction ID, a reflection ID, and an update each time an update process for reflecting the database reflection information in the master DB 21 is performed. One record having information is registered. The trigger ID is information for identifying update information that is set for each trigger issued by the update unit 71. For example, the update log extraction unit 72 generates a trigger ID incremented by 1 each time a trigger is received. The time is information indicating the time when the trigger is issued (that is, the time when the update is started), and is acquired by the update log extraction unit 72. The transaction ID is information for identifying transaction processing. The update log extraction unit 72 sets the same transaction ID in the update information generated by the same transaction process. For example, the update log extraction unit 72 generates a transaction ID incremented by 1 each time a transaction is switched. The reflection ID is information for identifying an update information group that is transferred to the slave server 12 at a single timing and reflected to the replica DB 22 at a single timing. The reflection ID is information set by an update log control unit 73, which will be described later, and is blank (or NULL) when the update log extraction unit 72 stores update information. The update information is information including data before update and data after update.

  For example, in the process of moving 100,000 yen from Mr. a's bank account to Mr. b's bank account, the process of subtracting 100,000 yen from the amount of Mr. a's bank account, and the amount of Mr. b's bank account to 100,000 The same transaction ID is set for each update information extracted based on each of the process of adding a circle. Then, the update log extraction unit 72 registers two records in the update log intermediate table A81.

  Also, as shown in FIG. 5, the update information includes the target table, the number of stored rows, and the update information on the Y (Y is a natural number (X <Y)) row from the update information on the X (X is a natural number) row. And have. The target table is information for specifying a table updated by the update unit 71 among the tables (databases) stored in the master DB 21. The number of stored rows is information indicating the number of records (rows) stored in the target table.

  The update information of the X-th row includes the target row primary key, the number of storage columns, and the pre-update data of the N (N is a natural number (M <N)) column from the pre-update data of the M (M is a natural number) column. Data and updated data in the Nth column from updated data in the Mth column. The target row primary key is information for specifying one updated row of the target table. The number of stored columns is information indicating the number of columns stored in the target table. The M-column pre-update data includes an M-column position indicating the position where the pre-update data is stored, an M-column length indicating the size of the pre-update data, and pre-update data indicating the data before the update. ,have. Similarly, the updated data in the Mth column includes the M column position indicating the position where the updated data is stored, the M column length indicating the size of the updated data, and the updated data indicating the updated data. And have.

  Returning to the description of FIG. When the update by the update unit 71 of FIG. 4 is completed and the update log extraction unit 72 stores all update information based on the database reflection information in the update log intermediate table A81, the update unit 71 outputs a success of transaction processing. That is, the update process by the update unit 71 is confirmed. Thereby, the update information based on the update of the master DB 21 can be extracted and stored in the update log intermediate table A81.

  Next, a process in which the master server 11 sets a reflection ID in the update information stored in the update log intermediate table A81 will be described. First, the update log control unit 73 refers to the update log intermediate table A81 and reads update information for which the reflection ID is not set (for example, blank or NULL). Subsequently, the update log control unit 73 determines a plurality of update information (update information group) to be transferred to the replica DB 22 at a single timing and reflected to the replica DB 22 based on a preset transfer condition. For example, when the number of cases is set in advance as the transfer condition, the update log control unit 73 sets the reflection IDs up to the set number, and determines a plurality of pieces of update information set with the same reflection ID as an update information group.

  Specifically, if the set number of cases is 100, the update log control unit 73 determines that the update information is within 100 cases and before the transaction ID is switched (that is, all the update information with the same transaction ID set). ) Set the same reflection ID. For example, the update log control unit 73 issues reflection IDs in order from 1. For example, when the transaction ID of the update information of the 90th to 98th items (trigger ID “90” to “98”) is “10” and the transaction ID of the update information of the 99th to 104th items is “11”, The update log control unit 73 sets the same reflection ID (for example, “1”) for the update information of the first to 98th items, and sets the update information of the first to 98th items for which the reflection ID “1” is set as one update information. Determine as a group. In addition, the next reflection ID (for example, “2”) is set in the update information after the 99th (less than the 199th), and a plurality of pieces of update information with the reflection ID “2” are set as the next update information group. decide. In this way, the update log control unit 73 assigns a reflection ID to each update information in the update log intermediate table A81 based on the unit number of update information.

  That is, in the process of moving 100,000 yen from Mr. a's bank account to Mr. b's bank account, the process of subtracting 100,000 yen from the amount of Mr. a's bank account and the amount of Mr. b's bank account to 100,000 The same transaction ID and the same reflection ID are set for each update information extracted based on the process of adding a circle.

  In addition, when the number of cases is not set in advance, the update log control unit 73 sets the same reflection ID until a time set in advance as a transfer condition (step S44). For example, when the preset time is “1 second”, the update log control unit 73 refers to the time in the update log intermediate table A81 until the update information within 1 second and before the transaction ID is switched. Set the same reflection ID. In this way, the update log control unit 73 can also assign a reflection ID to each update information in the update log intermediate table A81 based on the time interval unit.

  Next, a process in which the master server 11 transfers update information stored in the update log intermediate table A81 to the slave server 12 will be described. The update log transfer unit 74 performs a process of transferring the update information group extracted by the update log extraction unit 72 and determined by the update log control unit 73 to the replica DB 22 at a single timing.

  First, the update log transfer unit 74 refers to the update log intermediate table A81. When the number of cases is set, the update log transfer unit 74 acquires update information that is less than the set number (or more) and before the reflection ID is switched. If the set number is 100, the update log transfer unit 74 acquires update information of the 1st to 98th cases, for example. Note that the preset number of records read by the update log control unit 73 (for example, 100 cases described above) and the preset number of cases read by the update log transfer unit 74 may be different values (for example, 1000). The update log transfer unit 74 acquires all the update information stored in the update log intermediate table A81 when the number of cases is not set.

  Then, the update log transfer unit 74 transfers the acquired update information to the slave server 12 via the communication unit 44 and the communication network 13. When the slave server 12 receives the update information via the communication unit 53, the slave server 12 stores the update information in the update log intermediate table B101 of the storage unit 52. The update log intermediate table B101 has the same configuration as the update log intermediate table A81. Subsequently, the update log transfer unit 74 of the master server 11 instructs the update log reflection unit 91 to perform reflection processing. Further, the update log transfer unit 74 deletes the transferred update information (including the record) from the update log intermediate table A81.

  In this way, update information set with the same reflection ID (that is, update information set with the same transaction ID) is transferred at a single timing, so that consistency of update transaction processing performed on the master DB 21 is guaranteed. As it is, the update information can be transferred to the slave server 12.

  Next, a process in which the slave server 12 reflects the update information stored in the update log intermediate table B101 to the replica DB 22 will be described.

  When the reflection process is instructed by the update log transfer unit 74 of the master server 11, the update log reflection unit 91 first determines whether or not the reflection process is locked. The reflection process is locked when performing a batch reflection process described later. If the reflection process is locked, the update log transfer unit 74 waits until the lock of the reflection process is released. When it is determined that the reflection process is not locked, the update log reflection unit 91 refers to the update log intermediate table B101, and starts update transaction processing when update information is stored. Then, the update log reflecting unit 91 reflects the update information stored in the update log intermediate table B101 to the replica DB 22. Specifically, the update log reflection unit 91 acquires update information in which the same transaction ID is set at one time, and updates the replicated data stored in the replica DB 22. For example, in the process of moving 100,000 yen from Mr. a's bank account to Mr. b's bank account, the process of subtracting 100,000 yen from the amount of Mr. a's bank account, and the amount of Mr. b's bank account to 100,000 Each update information extracted based on each of the process of adding a circle is acquired at a single timing, and the replica DB 22 is updated.

  The update log reflection unit 91 outputs the success of the transaction process when the update information is successfully reflected on the replica DB 22. As a result, it is possible to guarantee the consistency of the update transaction processing for the replica DB 22.

  On the other hand, the update log reflection unit 91 performs rollback when the update information is not reflected on the replica DB 22. For example, when the load on the slave server 12 increases, reflection of update information may fail. When the update information reflection fails in this way, the update log reflection unit 91 performs the process of reflecting the update information to the replica DB 22 again. If the update information is not successfully reflected on the replica DB 22 even if the rollback is repeated (for example, the rollback is performed more than a preset number of times), the update log reflection unit 91 outputs a transaction processing failure. Then, the update log reflection unit 91 instructs the batch reflection execution unit 93 to perform batch reflection processing.

  Here, processing when the slave server 12 reflects update information to the replica DB 22 in a batch will be described. First, the batch reflection control unit 92 acquires the load of the CPU of the slave server 12 at a preset timing (for example, periodically), and the acquired load value is less than a preset threshold value. , Start the batch reflection process. The batch reflection control unit 92 can also start batch reflection processing when the number of update information (records) stored in the update log intermediate table B101 is less than a preset threshold.

  When starting the batch reflection process, the batch reflection control unit 92 locks the reflection process described above. If the reflection process is being executed, the batch reflection control unit 92 waits until the reflection process is completed, and locks when the reflection process is completed. After locking the reflection process, the batch reflection control unit 92 acquires table information to be batch reflected based on the update information that failed to be reflected by the update log reflection unit 91. For example, the batch reflection control unit 92 acquires information on the database itself that is the update source of the update information. This table information can be specified by the “target table” included in the update information stored in the update log intermediate table B101. Then, the batch reflection control unit 92 instructs the batch reflection execution unit 93 to execute the reflection process.

  Subsequently, the batch reflection execution unit 93 transmits the table information to be batch reflected acquired by the batch reflection control unit 92 to the table acquisition unit 75 of the master server 11 via the communication unit 53 and the communication network 13. In addition, by transmitting table information from the slave server 12, the master server 11 is notified that the update information has not been reflected in the replica DB 22. And the table acquisition part 75 acquires table (database) itself (whole) shown with the received table information from master DB21. For example, in the process of moving 100,000 yen from Mr. a's bank account to Mr. b's bank account, the table acquisition unit 75 stores a database including information on Mr. a's bank account and information on Mr. b's bank account. Get each with the containing database. Subsequently, the table acquisition unit 75 transfers the acquired table itself to the batch reflection execution unit 93 of the slave server 12 via the communication unit 44 and the communication network 13. Then, the batch reflection execution unit 93 reflects the received table on the replica DB 22. That is, the batch reflection execution unit 93 replaces the entire table stored in the replica DB 22 with the received table. As a result, even if the load on the slave server 12 is large and the update information reflection process fails, the update information reflection process is performed collectively when the load on the slave server 12 becomes small. Can be executed.

  Further, the batch reflection execution unit 93 records the start time for starting batch reflection, and when the batch reflection processing ends, the batch reflection in which the time before the start time is set from the update log intermediate table B101. Delete the update information contained in the updated table. As a result, the batch reflection process ends, and the batch reflection control unit 92 notifies the update log reflection unit 91 of the release of the lock of the reflection process. Then, when the update log reflection unit 91 is notified of the lock release of the reflection process, the update log reflection unit 91 executes the reflection process again.

  As described above, the operation of finely and arbitrarily adjusting the timing of transferring update information by setting the reflection ID of the update log control unit 73 and the transfer timing of the update log transfer unit 74 is possible. In addition, when the reflection by the update log reflection unit 91 fails, the batch reflection control unit 92 reflects the update information at a time when the load on the slave server 12 is relatively small, so that the consistency of the database is guaranteed. be able to. Therefore, the consistency of the update transaction processing performed on the master DB 21 can be guaranteed in the replica DB 22.

[Operation]
Next, the operation of the above-described information processing system 1 will be described in detail with reference to FIGS. FIG. 6 is a flowchart for explaining the update process. FIG. 7 is a flowchart for explaining the update log extraction processing. FIG. 8 is a flowchart for explaining the update information control process. FIG. 9 is a flowchart for explaining the update information transfer process. FIG. 10 is a flowchart for explaining the reflection process. FIG. 11 is a flowchart for explaining the batch reflection process.

  First, the update process of the update unit 71 and the update log extraction process of the update log extraction unit 72 will be described with reference to FIGS. First, in the update process of FIG. 6, the update unit 71 starts a transaction process (step S1). Subsequently, the update unit 71 acquires database reflection information (step S2). And the update part 71 starts the update of the master data memorize | stored in master DB21 (step S3).

Subsequently, the update unit 71 issues a trigger to the update log extraction unit 72 (step S4).
Upon receiving the trigger issued by the update unit 71, the update log extraction unit 72 starts the update log extraction process of FIG. Next, the update unit 71 determines whether or not the update is completed (step S5). When it is determined that the update has not been completed (step S5: No), that is, when the database reflection information still exists, the process returns to step S2, and the subsequent processes are repeated. On the other hand, when it is determined that the update has been completed (step S5: Yes), the update unit 71 notifies the update log extraction unit 72 of the completion of the update (step S6).

  Here, the update log extraction processing of the update log extraction unit 72 will be described with reference to FIG. In the update log extraction process of FIG. 7, first, the update log extraction unit 72 receives the trigger issued in the process of step S4 of FIG. 6 (step S21). Subsequently, the update log extraction unit 72 generates update information (update log) based on the master data before update stored in the master DB 21 and the updated master data updated based on the update information. (Extract) (Step S22).

  Then, the update log extraction unit 72 stores the update information generated in the process of step S22 in the update log intermediate table A81 (step S23). Next, the update log extraction unit 72 determines whether or not an update completion notification output from the update unit 71 has been received in the process of step S6 of FIG. 6 (step S24). If it is determined that the update completion notification has not been received (step S24: No), the process returns to step S21 and the subsequent processes are repeated.

  On the other hand, if it is determined that an update completion notification has been received (step S24: Yes), the update log extraction unit 72 determines whether all update information has been stored (step S25). If it is determined that not all update information is stored (step S25: No), the process returns to step S21 and the subsequent processes are repeated.

  On the other hand, when it is determined that all the update information has been stored (step S25: Yes), that is, when all the update information generated within the same transaction is stored in the update log intermediate table A81, the update log extraction unit 72 The storage completion is notified to the update unit 71 (step S26). After the process of step S26, the update log extraction process ends. Thereby, while updating master DB21 by transaction processing, update information of master DB21 can be extracted.

  Next, in step S7 of FIG. 6, the updating unit 71 receives the storage completion notified in the process of step S26 of FIG. 7 (step S7). Then, the update unit 71 outputs the success of the transaction process (Step S8). After the process of step S8, the update process ends.

  Next, the update log control process of the update log control unit 73 will be described with reference to FIG. The update log control process operates independently of the operation of the update unit 71.

  First, the update log control unit 73 refers to the update log intermediate table A81 to determine whether there is update information for which no reflection ID is set (step S41). If it is determined that there is no update information for which no reflection ID is set (step S41: No), the process returns to step S41, and the same process is repeated. On the other hand, when it is determined that there is update information for which the reflection ID is not set (step S41: Yes), the update log control unit 73 determines whether or not the number of cases is set (step S42).

  When it is determined that the number of cases has been set (step S42: Yes), the update log control unit 73 sets the reflection ID up to the set number (step S43). Specifically, if the number of set cases is 100, the update log control unit 73 sets the same reflection ID within 100 cases and the update information before the transaction ID is switched. For example, the update log control unit 73 issues reflection IDs in order from 1. Also, the same reflection ID is set for update information having the same transaction ID. For example, when the transaction ID of the 90th to 98th update information is “10” and the transaction ID of the 99th to 104th update information is “11”, the update log control unit 73 sets 1 to 98 cases. The same reflection ID is set for the update information of the eye.

  On the other hand, when it is determined that the number of cases has not been set (step S42: No), the update log control unit 73 sets the reflection ID until a preset time (step S44). For example, when the preset time is “1 second”, the update log control unit 73 refers to the time in the update log intermediate table A81 until the update information within 1 second and before the transaction ID is switched. Set the same reflection ID.

  After the processes in steps S43 and S44, the process returns to step S41, and the subsequent processes are repeated. In this manner, the update log control unit 73 assigns a reflection ID to each update information in the update log intermediate table A81 based on the number of update information items or the time interval unit.

  Next, the update log transfer process of the update log transfer unit 74 will be described with reference to FIG. The update log transfer process starts at a preset timing (for example, periodically).

  First, the update log transfer unit 74 refers to the update log intermediate table A81 and determines whether update information exists (step S61). When it is determined that update information exists (step S61: Yes), the update log transfer unit 74 determines whether or not the number of cases is set (step S62). When it is determined that the number of cases is set (step S62: Yes), the update log transfer unit 74 determines whether there is update information equal to or more than the set number (step S63).

  When it is determined that the number of cases is not set in the process of step S62 (step S62: No), and when it is determined that there is no update information greater than the number of cases set in the process of step S63 (step S63: No), the update log transfer unit 74 acquires all the update information stored in the update log intermediate table A81 (step S64). On the other hand, if it is determined that there are more update information than the number set in the process of step S63 (step S63: Yes), the update information to be transferred is acquired (step S65). Specifically, the update log transfer unit 74 acquires update information that is greater than or equal to the set number (or less) and before the reflection ID is switched.

  Then, the update log transfer unit 74 transfers the update information acquired in the processes of steps S64 and S65 to the update log intermediate table B101 (step S66). The update log intermediate table B101 has the same configuration as the update log intermediate table A81 and is the same as that shown in FIG. Subsequently, the update log transfer unit 74 instructs the update log reflection unit 91 to perform reflection processing (step S67). The reflection process will be described later with reference to FIG. Next, the update log transfer unit 74 deletes the transferred update information (including records) from the update log intermediate table A81 (step S68).

  Then, the update log transfer unit 74 waits for a preset time (step S69), and then repeats the processing after step S61. Even when it is determined in the process of step S61 that there is no update information (step S61: No), the update log transfer unit 74 waits for a preset time (step S69).

  Next, the reflection process of the update log reflection unit 91 will be described with reference to FIG. The reflection process starts when the update log transfer unit 74 instructs the reflection process (for example, step S67 in FIG. 9).

  First, the update log reflection unit 91 determines whether the reflection process is locked (step S81). The reflection process is locked when performing a batch reflection process described later. The batch reflection process will be described later with reference to FIG. When it is determined that the reflection process is not locked (step S81: No), the update log reflection unit 91 determines whether update information exists in the update log intermediate table B101 (step S82).

  If it is determined that update information exists (step S82: Yes), the update log reflection unit 91 starts transaction processing (step S83). Then, the update log reflecting unit 91 reflects the update information stored in the update log intermediate table B101 to the replica DB 22 (Step S84). Specifically, the update log reflecting unit 91 reflects the update information in which the same transaction ID is set in the replica DB 22 and updates the replicated data stored in the replica DB 22.

  Subsequently, the update log reflection unit 91 determines whether the reflection is successful (step S85). When it is determined that the reflection is successful (step S85: Yes), the update log reflection unit 91 outputs the success of the transaction process (step S86). Thereby, reflection of the update information with respect to the replica DB 22 can be confirmed. Then, the update log reflection unit 91 deletes the reflected update information (including the record) from the update log intermediate table B101 (step S87). After the process of step S87, the process returns to step S81, and the subsequent processes are repeated.

  On the other hand, when it is determined in the process of step S85 that the reflection is not successful (step S85: No), the update log reflection unit 91 determines whether or not the number of rollbacks is equal to or greater than a preset threshold value. (Step S88). When it is determined that the number of rollbacks is not equal to or greater than the threshold (step S88: No), the update log reflecting unit 91 performs rollback (step S89). Then, after the process of step S89, the process returns to step S84 and the update information is reflected again.

  On the other hand, when it is determined in step S88 that the number of rollbacks is equal to or greater than the threshold (step S88: Yes), the update log reflection unit 91 outputs a transaction processing failure (step S90). Then, the update log reflection unit 91 instructs the batch reflection execution unit 93 to perform batch reflection processing (step S91). The batch reflection process will be described with reference to FIG. After the process of step S91, the process returns to step S81, and the subsequent processes are repeated.

  Further, when it is determined that the reflection process is locked in the process of step S81 (step S81: Yes), and when it is determined that the update information does not exist in the process of step S82 (step S82: No). The reflection process in FIG. 10 ends.

  Next, the batch reflection process will be described with reference to FIG. First, the batch reflection control unit 92 acquires load information indicating the CPU load of the slave server 12 at a preset timing (for example, periodically) (step S111). Subsequently, the batch reflection control unit 92 determines whether or not the acquired load is less than a preset threshold value (step S112). If it is determined that the acquired load is not less than a preset threshold value (greater than or equal to the threshold value) (step S112: No), the process returns to step S111, and the subsequent processes are repeated.

  On the other hand, when the acquired load is less than the preset threshold value in step S112 (step S112: Yes), the collective reflection control unit 92 locks the reflection process (step S113). After locking the reflection process, the update log reflection unit 91 acquires table information to be collectively reflected based on the update information that failed to be reflected (step S114).

  Subsequently, the batch reflection execution unit 93 transmits the table information to be batch reflected acquired by the batch reflection control unit 92 to the table acquisition unit 75 of the master server 11 via the communication unit 53 and the communication network 13 (Step S1). S115). The table acquisition unit 75 of the master server 11 acquires the table (database) itself indicated by the received table information from the master DB 21. Subsequently, the table acquisition unit 75 transfers the acquired table itself to the batch reflection execution unit 93 of the slave server 12 via the communication unit 44 and the communication network 13.

  Then, the batch reflection execution unit 93 receives the transferred table (step S116), and updates the replica DB 22 (step S117). That is, the batch reflection execution unit 93 replaces the entire table stored in the replica DB 22 with the received table. As a result, even if the load on the slave server 12 is large and the update information reflection process fails, the update information reflection process is performed collectively when the load on the slave server 12 becomes small. Can be executed.

  Next, the batch reflection execution unit 93 deletes the update information stored in the update log intermediate table B101 (and the update log intermediate table A81) (step S118). Specifically, the batch reflection execution unit 93 records the start time for starting batch reflection, and when the batch reflection processing is completed, a time before the start time is set from the update log intermediate table B101. Delete update information. Finally, the batch reflection control unit 92 unlocks the reflection process to the update log reflection unit 91 (step S119). After the process of step S119, the batch reflection process ends. Then, when the update log reflection unit 91 is notified of the lock release of the reflection process, the update log reflection unit 91 executes the reflection process again.

  In this embodiment, when there are a plurality of replica DBs 22 (slave servers 12), a plurality of update log intermediate tables A81 of the master server 11 are prepared as many as the number of replica DBs 22. Then, the update log extraction unit 72 simultaneously writes to each of the plurality of update log intermediate tables when a trigger is issued by the update unit 71. Then, the update log transfer unit 74 transfers the update information for each replica DB 22 unit. As described above, when a plurality of replica DBs 22 are prepared and divided into remote locations, the transfer timing can be made independent.

Second Embodiment
Next, a master server according to the second embodiment of the present invention will be described with reference to FIG.
The master server 201 (information processing apparatus) according to the second embodiment
An update unit 212 (update means) that performs a predetermined update process on the master database (master DB) 211 and updates the database stored in the master DB 211;
An update log extraction unit 213 (update information extraction means) that extracts update information based on the database updated by the update unit 212;
A plurality of update information extracted by the update log extraction unit 213 is included in one update information group, and the update information group is transferred to a replica database that stores replica data of the database stored in the master DB 211 at a single timing. An update log transfer unit 214 (update information transfer means) to transfer,
With
The update log transfer unit 214 manages a plurality of related update processes as a series of processes, includes a plurality of update information updated by the same transaction process in one update information group, and includes the update information group in the replica database. And transfer at once.

  With the above configuration, when the update unit 212 updates the database by transaction processing, the update log extraction unit 213 includes a plurality of update information in one update information group. Then, the update log transfer unit 214 transfers the update information group to the replica database at a single timing. Therefore, consistency of transaction processing performed on the master DB 211 can be ensured in the replica database while suppressing a load on the master server 201 and the replica database side device.

  Although the present invention has been described with reference to the above embodiment, the present invention is not limited to the above-described embodiment. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

<Appendix>
Part or all of the above-described embodiment can be described as in the following supplementary notes. The outline of the configuration of the information processing apparatus and the like in the present invention will be described below. However, the present invention is not limited to the following configuration.

(Appendix 1)
Updating means for performing a predetermined update process on the master database and updating the database stored in the master database;
Update information extraction means for extracting update information based on the database updated by the update means;
A plurality of the update information extracted by the update information extraction unit is included in one update information group, and the update information group is once stored in a replica database storing replica data of the database stored in the master database. Update information transfer means for transferring at the timing of
With
The update information transfer means manages a plurality of related update processes as a series of processes, includes a plurality of update information updated by the same transaction process in one update information group, and includes the update information group. Transfer to the replica database at a single timing,
Information processing device.

  According to this, when the database is updated by transaction processing, a plurality of update information is included in one update information group, and the update information group is transferred to the replica database at a single timing. It is possible to guarantee the consistency of transaction processing performed on the master database in the replica database while suppressing the load on the replica database side device.

(Appendix 2)
An information processing apparatus according to attachment 1, wherein
Based on a preset transfer condition, the update information group is determined so that the plurality of update information extracted based on the database updated by the same transaction process is included in one update information group. Further comprising update information control means,
The update information transfer means transfers one update information group determined by the update information control means to the replica database at a single timing.
Information processing device.

  According to this, based on a preset transfer condition, the update information group is determined so that a plurality of update information extracted based on the database updated by the same transaction process is included in one update information group. Since the determined update information group is transferred to the replica database at a single timing, the update information group that guarantees the consistency of the transaction processing of the update performed on the master database is determined, and the replica database is determined. Can be transferred to. Therefore, the consistency of update transaction processing can be ensured in the replica database.

(Appendix 3)
An information processing apparatus according to appendix 2, wherein
The update information extraction means sets the same transaction identification information to each of the plurality of update information extracted based on the database updated by the same transaction processing,
The update information control means determines one update information group including all the update information in which the same transaction identification information is set by the update information extraction means.
Information processing device.

  According to this, the same transaction identification information is set in each of a plurality of update information extracted based on the database updated by the same transaction processing, and all update information in which the same transaction identification information is set is included. Since one update information group is determined, each update information based on the same transaction process can be easily managed, and the consistency of the update transaction process can be more reliably ensured in the replica database.

(Appendix 4)
An information processing apparatus according to attachment 3, wherein
The update information control means sets the same transfer identification information to each of the plurality of update information including all the update information for which the same transaction identification information is set by the update information extraction means, and the same transfer identification information A plurality of the update information set as is determined as the update information group,
Information processing device.

  According to this, transfer identification information is set for each of a plurality of update information including all update information set with the same transaction identification information, and the plurality of update information set with the same transfer identification information is set as an update information group. Therefore, the update information group transferred to the replica database at a single timing can be easily managed, and the consistency of update transaction processing can be more reliably ensured in the replica database.

(Appendix 5)
An information processing apparatus according to appendix 4, wherein
The update information control means is the update information equal to or less than a preset number, or the update information extracted by the update information extraction means within a preset time, and the same transaction identification information Set the same transfer identification information for all the set update information,
Information processing device.

  According to this, since the number of update information transferred to the replica database at a single timing can be limited, it is possible to suppress a load on other information processing apparatuses including the replica database.

(Appendix 6)
An information processing apparatus according to any one of appendices 1 to 5,
When another information processing apparatus including the replica database is notified that the update information is not reflected on the replica database, the database itself that is the update source of the update information is acquired from the master database. Database transfer control means for transferring the acquired database itself to the replica database;
An information processing apparatus further comprising:

  According to this, when another information processing apparatus fails to reflect the update information to the replica database, the database itself as the update source of the update information is acquired and transferred to the replica database. Therefore, the other information processing apparatus can replace the database stored in the replica database with the transferred database and update it. Therefore, it is possible to prevent the update of the replica database by a simple process, and to reliably replace the replica database. Can be updated.

(Appendix 7)
In the information processing device,
Updating means for performing a predetermined update process on the master database and updating the database stored in the master database;
Update information extraction means for extracting update information based on the database updated by the update means;
A plurality of the update information extracted by the update information extraction unit is included in one update information group, and the update information group is once stored in a replica database storing replica data of the database stored in the master database. Update information transfer means for transferring at the timing of
And realize
The update information transfer means manages a plurality of related update processes as a series of processes, includes a plurality of update information updated by the same transaction process in one update information group, and includes the update information group. Transfer to the replica database at a single timing,
A program to make things happen.

(Appendix 8)
The program according to appendix 7,
Based on a preset transfer condition, the update information group is determined so that the plurality of update information extracted based on the database updated by the same transaction process is included in one update information group. Further comprising update information control means,
The update information transfer means transfers one update information group determined by the update information control means to the replica database at a single timing.
program.

(Appendix 9)
Perform a predetermined update process on the master database, update the database stored in the master database,
Extract update information based on the updated database,
Managing a plurality of related update processes as a series of processes, including a plurality of update information extracted based on the database updated by the same transaction process in one update information group, , Transferred to the replica database storing the replicated data of the database stored in the master database at a single timing,
Information processing method.

(Appendix 10)
An information processing method according to attachment 9, wherein
Based on a preset transfer condition, the update information group is determined so that the plurality of update information extracted based on the database updated by the same transaction process is included in one update information group. ,
Transferring the determined one update information group to the replica database at a single timing;
Information processing method.

1 Information processing system 11 Master server 12 Slave server 13 Communication network 21 Master DB
22 Replica DB
41 calculation unit 42 storage unit 43 input / output unit 44 communication unit 51 calculation unit 52 storage unit 53 communication unit 71 update unit 72 update log extraction unit 73 update log control unit 74 update log transfer unit 75 table acquisition unit 81 update log intermediate table A
91 Update log reflection unit 92 Batch reflection control unit 93 Batch reflection execution unit 101 Update log intermediate table B
201 Master server 211 Master DB
212 Update Unit 213 Update Log Extraction Unit 214 Update Log Transfer Unit

Claims (10)

Updating means for performing a predetermined update process on the master database and updating the database stored in the master database;
Update information extraction means for extracting update information based on the database updated by the update means;
A plurality of the update information extracted by the update information extraction unit is included in one update information group, and the update information group is once stored in a replica database storing replica data of the database stored in the master database. Update information transfer means for transferring at the timing of
With
The update information transfer means manages a plurality of related update processes as a series of processes, includes a plurality of update information updated by the same transaction process in one update information group, and includes the update information group. Transfer to the replica database at a single timing,
Information processing device. The information processing apparatus according to claim 1,
Based on a preset transfer condition, the update information group is determined so that the plurality of update information extracted based on the database updated by the same transaction process is included in one update information group. Further comprising update information control means,
The update information transfer means transfers one update information group determined by the update information control means to the replica database at a single timing.
Information processing device. An information processing apparatus according to claim 2,
The update information extraction means sets the same transaction identification information to each of the plurality of update information extracted based on the database updated by the same transaction processing,
The update information control means determines one update information group including all the update information in which the same transaction identification information is set by the update information extraction means.
Information processing device. The information processing apparatus according to claim 3,
The update information control means sets the same transfer identification information to each of the plurality of update information including all the update information for which the same transaction identification information is set by the update information extraction means, and the same transfer identification information A plurality of the update information set as is determined as the update information group,
Information processing device. The information processing apparatus according to claim 4,
The update information control means is the update information equal to or less than a preset number, or the update information extracted by the update information extraction means within a preset time, and the same transaction identification information Set the same transfer identification information for all the set update information,
Information processing device. An information processing apparatus according to any one of claims 1 to 5,
When it is notified from another information processing apparatus including the replica database that the update information has failed to be reflected in the replica database, the database itself that is the update source of the update information is acquired from the master database. Database transfer control means for transferring the acquired database itself to the replica database;
An information processing apparatus further comprising: In the information processing device,
Updating means for performing a predetermined update process on the master database and updating the database stored in the master database;
Update information extraction means for extracting update information based on the database updated by the update means;
A plurality of the update information extracted by the update information extraction unit is included in one update information group, and the update information group is once stored in a replica database storing replica data of the database stored in the master database. Update information transfer means for transferring at the timing of
And realize
The update information transfer means manages a plurality of related update processes as a series of processes, includes a plurality of update information updated by the same transaction process in one update information group, and includes the update information group. Transfer to the replica database at a single timing,
A program to make things happen. The program according to claim 7,
Based on a preset transfer condition, the update information group is determined so that the plurality of update information extracted based on the database updated by the same transaction process is included in one update information group. Further comprising update information control means,
The update information transfer means transfers one update information group determined by the update information control means to the replica database at a single timing.
program. Perform a predetermined update process on the master database, update the database stored in the master database,
Extract update information based on the updated database,
Managing a plurality of related update processes as a series of processes, including a plurality of update information extracted based on the database updated by the same transaction process in one update information group, , Transferred to the replica database storing the replicated data of the database stored in the master database at a single timing,
Information processing method. An information processing method according to claim 9,
Based on a preset transfer condition, the update information group is determined so that the plurality of update information extracted based on the database updated by the same transaction process is included in one update information group. ,
Transferring the determined one update information group to the replica database at a single timing;
Information processing method.

Images