TWI805544B - Database operation method and device - Google Patents

Abstract

The invention provides a database operation method and device. A method includes: during the process of the application program server executing the target transaction, sequentially obtaining the database operation instructions in the target transaction executed by the application program server; performing predictive execution on the database operation instructions, and returning the prediction execution result to the application The program server is used for the application program server to determine the next database operation command that needs to be executed, and to record the database operation command locally, and the database operation command belongs to the modification database type instruction and the locked non-modification database type When one of the instructions is recorded, the predicted execution data generated by the predicted execution is recorded; when the transaction submission command in the target transaction is obtained, the actual execution target of the database corresponding to the application server is controlled according to the locally recorded database operation command and the predicted execution data affairs. The invention can improve transaction execution efficiency and throughput.

Description

Database operation method and device

The invention relates to the technical field of databases, in particular to a database operation method and device.

Database Transaction (Database Transaction) refers to a series of operations performed as a single logical unit of work, either completely executed or not executed at all. A transaction consists of a series of database operation instructions, such as statements.

In the database operation, in order to ensure the correctness of concurrent read data, the concept of transaction isolation level is proposed. In the standard Structured Query Language (SQL) specification, four transaction isolation levels are defined, among which, Read Committed (Read Committed) is the preferred isolation level for most applications, which can avoid dirty reads, And it has better concurrency performance, but still encounters the problem of non-repeatable read.

For the non-repeatable read problem faced by the read committed isolation level transaction, you can temporarily increase the isolation level of a certain data in the read committed isolation level transaction through the for update statement to ensure that the data can be read repeatedly, but this will seriously reduce the The concurrent performance of the system leads to low execution efficiency and low transaction throughput.

Aspects of the present invention provide a database operation method and device, which are used to ensure the concurrent performance of the database system, improve transaction execution efficiency and increase transaction throughput.

In one aspect of the present invention, a method for operating a database is provided, including: during the process of the application server executing the target transaction, sequentially obtaining the database operation instructions in the target transaction executed by the application server; The database operation instruction performs predictive execution, returns the predicted execution result to the application server for the application server to determine the next database operation instruction to be executed, and records the database operation locally instruction and record the predicted execution data generated by the predictive execution when the database operation instruction belongs to one of the modified database type instruction and the locked non-modified database type instruction; when the transaction in the target transaction is obtained When submitting an instruction, control the database corresponding to the application server to actually execute the target transaction according to the locally recorded database operation instruction and the predicted execution data.

Another aspect of the present invention provides a method for operating a database, including: during the process of executing a target transaction by an application server, sequentially acquiring database operation instructions in the target transaction executed by the application server; Perform predictive execution on the database operation command, return the predictive execution result to the application program server for the application program server to determine the next database operation command to be executed, and record the database locally Operating instructions and recording the predicted execution data generated by the predicted execution when the database operation instruction belongs to the locked non-modified database instruction; when the transaction commit instruction in the target transaction is obtained, according to the local record The database operation instruction and the forecast execution data control the database corresponding to the application server to actually execute the target transaction.

Still another aspect of the present invention provides a database operation device, including: an acquisition module, used to sequentially acquire the target transaction executed by the application server during the process of executing the target transaction by the application server The database operation instruction; the predictive execution module is used to perform predictive execution on the database operation instruction, and return the predictive execution result to the application program server for the application program server to determine the next one that needs to be executed Database operation instructions, and record the database operation instructions locally and record the results generated by the speculative execution when the database operation instructions belong to one of the modified database instructions and the locked non-modified database instructions predictive execution data; a control execution module, configured to control the application server corresponding to the The database actually executes the target transaction.

Still another aspect of the present invention provides a database operation device, including: an acquisition module, used to sequentially acquire the target transaction executed by the application server during the process of executing the target transaction by the application server The database operation instruction; the predictive execution module is used to perform predictive execution on the database operation instruction, and return the predictive execution result to the application program server for the application program server to determine the next one that needs to be executed Database operation instruction, and record the database operation instruction locally and record the prediction execution data generated by the prediction execution when the database operation instruction belongs to the locked non-modified database instruction; control the execution module , used to control the database corresponding to the application server to actually execute the target according to the locally recorded database operation command and the predicted execution data when the transaction commit command in the target transaction is obtained affairs.

In the present invention, the database operation device and the application server cooperate with each other, adding a predictive execution process in the process of executing the target transaction on the application server, obtaining and recording in advance all the database operation instructions that the target transaction needs to execute, and only for Database operation instructions that belong to the database operation instructions that belong to the modified database type instructions and the locked non-modified database type instructions, or only record the speculative execution data generated by the speculative execution for the database operation instructions that belong to the locked non-modified database type instructions , to adapt to the locking scheme under the read-committed isolation level, providing conditions for the actual execution of transactions, and then controlling the database corresponding to the application server to actually execute the target transaction according to the recorded database operation instructions and predicted execution data , which is conducive to improving execution efficiency and thus increasing transaction throughput.

10‧‧‧Application server

20‧‧‧database

30‧‧‧Database operation device

51‧‧‧Acquiring modules

52‧‧‧Forecast Execution Module

53‧‧‧Control Execution Module

61‧‧‧Acquiring modules

62‧‧‧Forecast Execution Module

63‧‧‧Control Execution Module

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings that need to be used in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following descriptions are some of the present invention. Embodiments, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without paying creative labor.

Fig. 1 is a deployment diagram of a database application system in the prior art; Fig. 2 is a deployment diagram of a database application system provided by an embodiment of the present invention; Fig. 3 is a schematic flowchart of a database operation method provided by another embodiment of the present invention ; Fig. 4 is a schematic flow chart of a database operation method provided by another embodiment of the present invention; Fig. 5 is a schematic structural diagram of a database operation device provided by another embodiment of the present invention; Fig. 6 is provided by another embodiment of the present invention Schematic diagram of the structure of the database operation device.

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is a part of embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

FIG. 1 is a deployment diagram of a database application system in the prior art. As shown in FIG. 1 , the system includes: an application server 10 and a database 20 . Wherein, there may be one application server 10 or multiple application servers. There can be one database 20, or multiple databases. FIG. 1 only shows an application server 10 and a database 20 as examples. When it is necessary to access the database 20 in a transactional manner, the application server 10 executes the transactional logic, and sequentially applies the database operation commands in the transaction to the database 20 . Since the database operation instructions that need to be executed in the transaction cannot be known in advance, the execution efficiency of the transaction is low, especially in the scheme that locks the data through the for update statement under the configured read committed isolation level, the execution efficiency will be lower , the transaction throughput is low.

Aiming at the defects existing in the prior art, the present invention provides a new database application system, as shown in FIG. 2 . In this system, a database operation device 30 is added between the application server 10 and the database 20 . The database operation device 30 is used to execute the database operation method provided by the present invention, and implement a new transaction execution logic in the locking scheme under the read-committed isolation level, that is, perform predictive execution on the transaction first, and obtain the transaction in advance The database operation instructions that need to be executed to realize the prediction of the execution path, and combined with the read submitted isolation level and locked, only for the instructions that belong to the modified database class and the locked non-modified database class instructions Database operation instructions, or only for the database operation instructions that belong to the locked non-modified database instructions, record the predicted execution data generated by predictive execution, and then obtain the database operation instructions and predictive execution based on the predicted execution The data actually executes the transaction on the database 20, which is beneficial to improve the transaction execution efficiency and increase the transaction throughput. Transaction throughput refers to the number of transactions processed per unit time.

It should be noted that the database operation device 30 is actually a logic processing device, which can be deployed independently, located between the application program server 10 and the database 20, or can be deployed at the end of the application program server 10 for implementation, or also It can be deployed at one end of the database 20 for implementation.

The following examples will specifically illustrate the method flow of the technical solution of the present invention.

FIG. 3 is a schematic flowchart of a database operation method provided by another embodiment of the present invention. As shown in Figure 3, the method includes:

301. During the execution of the target transaction by the application server, sequentially acquire database operation instructions in the target transaction executed by the application server.

302. Predictively execute the database operation command, return the predictive execution result to the application server for the application server to determine the next database operation command to be executed, and record the database operation command locally and store it in the database When the operation instruction belongs to one of the modified database instruction and the locked non-modified database instruction, the predicted execution data generated by the predicted execution is recorded.

303. When the transaction submission instruction in the target transaction is obtained, control the database corresponding to the application server to actually execute the target transaction according to the locally recorded database operation instruction and predicted execution data.

Firstly, for the convenience of description, in this embodiment, the transaction to be executed by the application server is referred to as the target transaction. The target transaction mainly includes operation instructions for operating the database. In addition to the database operation instructions, it also includes some control instructions for controlling the execution status of the target transaction, such as transaction start instructions, transaction commit instructions, and transaction rollbacks. ) instructions, etc. These instructions are actually statements written in the language of the database. Depending on the language of the database, these instructions can be statements written in different languages. For example, if SQL is adopted, the above-mentioned database operation instructions and control instructions are actually a series of SQL statements.

In addition, in some application scenarios, transaction processing needs to select some records first, and then process these records, so it is necessary to add an exclusive lock to the selected records. This means that the database operation command in this embodiment may be a locked database operation command. For example, assuming that 100 yuan is transferred from Bob's account, this operation is completed in three steps, query Bob's account, determine whether the account is enough for 100 yuan, and reduce the money in Bob's account by 100 yuan. If there are two transfer operations in parallel, when the first transfer operation finds out the balance of Bob's account, it is found that there is 100 yuan. If there is another transfer operation at this time, it also checks Bob's account, and it is also found in Bob's account There are 100 yuan, so the two transfer operations will transfer money, so the first operation subtracts 100 yuan from Bob's account, and the other operation also subtracts 100 yuan from Bob's account. At this time, Bob's account The balance may be negative. In order to avoid this situation, you need to use for update to lock Bob's account to ensure that during the first transfer operation, the second transfer operation cannot obtain this lock, and thus cannot be queried.

The method provided in this embodiment is suitable for operating the database whose isolation level is read committed. Among them, some databases default to the read-committed isolation level. For example, the Oracle database defaults to the read-committed isolation level. Some databases default to the read-committed isolation level. For example, the MySQL database defaults to Repeatable read is the isolation level. Based on this, before executing the method provided by this embodiment, it can be determined whether the isolation level of the database is read committed, and if not, the isolation level of the database needs to be preconfigured as read committed. For example, you can configure the database to read-committed isolation level before the application server executes the target transaction.

On the premise that the isolation level of the database is read committed, the method provided by this embodiment can be executed to execute the target transaction in the database. The specific process is as follows: the application server controls the execution logic of the entire target transaction, and can execute the target transaction in an existing manner. The database operation device sequentially obtains the database operation instructions in the target transaction executed by the application server during the execution of the target transaction by the application server.

In an optional implementation manner, the application program server sends database operation instructions to the database in an existing manner, so as to realize access to the database. The database operation device can monitor the communication between the application program server and the database, so as to intercept the database operation commands sent by the application program server to the database.

In another optional implementation manner, the processing logic of the application server is slightly modified, and the original processing logic of sending database operation instructions to the database is changed to sending to the database operation device. Based on this, the database operation device can receive the database operation instruction issued by the application server.

It is explained here that if the target transaction is an explicit transaction, the first command in the target transaction is a transaction start command, such as begin transaction. Based on this, the database operation device obtains the data executed by the application server Before the library operation command, the transaction start command in the target transaction sent by the application server to the database can be intercepted, and according to the transaction start command, it is known that the target transaction needs to be executed. Alternatively, before the database operation device obtains the database operation command executed by the application server, it may receive the transaction start command in the target transaction actively issued by the application server, and learn that the target transaction needs to be executed according to the transaction start command. . If the target transaction is an implicit transaction, the first command of the target transaction is the database operation command, excluding the step of obtaining the transaction start command.

After the database operation device obtains the database operation instructions executed by the application program server, it predictively executes the obtained database operation instructions, and returns the prediction execution results to the application program server for the application program server to determine A database operation command to execute. Wherein, the predicted execution result determines the execution path of the target transaction, where the execution path refers to the jump logic between database operation instructions. By returning the predictive execution result to the application server, the application server controls the execution logic of the entire target transaction.

In addition, the database operation device also needs to record the obtained database operation instruction locally, and when the database operation instruction belongs to one of the database modification instruction and the locked non-modification database instruction, the record is correct. The speculative execution data generated by the speculative execution of the database operation instruction. Described modification database class instruction refers to the database operation instruction that database has modification, such as the instruction beginning with update, insert etc.; Generally the instruction beginning with select does not belong to modification database class instruction (referred to as non-modification database class instructions). Although in essence, the read-committed isolation level only guarantees that writing multiple data is atomic and does not lock when reading, but in order to ensure transaction consistency and reduce the probability of return, some non-modifying database instructions It may be necessary to lock through the for update statement. For example, in the above account transfer example, the select command needs to be locked to form a database operation command in the form of select...forupdate. This kind of command can be called a locked non- Modify the database class directive. The predictive execution data mainly refers to some data during the predictive execution process of the database operation instruction, such as the data operated by the database operation instruction, the primary key ID of the data and the version number information (version) corresponding to the primary key ID, etc.

It is worth noting that, for the same database operation command, either it belongs to the database modification command, or it belongs to the locked non-modification database command, or it belongs to the unlocked non-modification database command. However, from the point of view of the entire target transaction, the predictive execution data generated by the speculative execution of all database-modifying instructions and all locked non-modifying database-type instructions in the target transaction will be recorded locally.

For example, the database operating device can create a memory bank locally, and store the obtained database operation instructions and predicted execution data into the memory bank. Further, if the predictive execution data does not include the predictive execution result, the predictive execution result may also be stored in the memory bank.

In order to realize the simulation of the locking scheme under the read-committed isolation level on the basis of the core idea of predictive execution, this embodiment is aimed at the instructions belonging to the modified database class and the locked non-modified database class instructions. One of the database operation instructions records the speculative execution data generated by its speculative execution, but does not record the speculative execution data corresponding to the database operation instructions that belong to the unlocked non-modifying database instructions. It can be seen that in the case of implementing the locking scheme under the isolation level of read committed, the amount of recorded data is relatively small, so that the subsequent database operation instructions and predicted execution data based on local records will be stored in the database. In the process of actually executing the target transaction, the amount of data to be processed is relatively small, which is conducive to further improving execution efficiency and performance, and increasing transaction throughput.

Wherein, the operation of the database operation command on the database is mainly to access the data in the database, so the data environment of the database operation command can be simulated, and the database operation command can be predicted and executed based on the simulated data environment. Further, considering that the database operation instructions belonging to non-modifying database instructions will not cause changes in the database, so the database operation instructions can be directly predicted and executed in the database, which can save the operation of the simulated data environment, The implementation is relatively simple, which is conducive to saving resources and improving execution efficiency. However, as for the database operation command belonging to the database modification command, because it will cause the change of the database (mainly refers to the change of the data in the database), it is necessary to simulate the data environment and perform the operation in the simulated data environment. Predictive execution.

Based on the above analysis, a method of predictive execution of database operation instructions can be: identify the type of the acquired database operation instructions; The database simulates the data environment of the database operation command in the database, and predicts and executes the database operation command based on the simulated data environment; if the database operation command is a non-modified database command, execute the database operation command in the database, to perform speculative execution on the database operation instruction.

Further, a data environment for simulating database operation instructions in a locally created memory bank, the implementation of predictive execution of database operation instructions based on the simulated data environment includes: splitting the obtained database operation instructions into A read command and a write command, run the read command on the real database, that is, execute the read command in the database to obtain the read data set (ReadSet), store the read data set in the local memory bank, and Simulate the data environment required by the database operation command. Then, the write command acts on the memory bank to realize the predictive execution of the database operation command, that is, execute the write command in the memory bank to modify the read data set, such as updating or querying the relevant data in the read data set processing etc. Wherein, executing the write command to modify the read data set can generate a result data set (affectRowInMemdb), and the result data set includes the predicted execution result.

In the above-mentioned predictive execution process, if the database operation instruction belongs to the database modification instruction, the database operation device can locally record the read data set read by the read command and the result data set generated by executing the write command. It can be seen that the predicted execution data may include the read data set and the result data set, or may also include part of the data in the read data set and the result data set, for example, some data that can have a beneficial effect on the actual execution of the target transaction in the database. Data, such as various indexes of some values, such as primary index, secondary index, primary key ID, version number, etc.

Preferably, in order to further reduce the amount of recorded data, when the database operation instruction is a locked non-modified database instruction, only the primary key ID and version number of the data operated by the database operation instruction can be recorded as a prediction Executive data. In this way, in the process of actually executing the target transaction in the database based on the locally recorded database operation instructions and predicted execution data, the amount of data to be processed will be further reduced, which is conducive to further improving execution efficiency and performance, and increasing the number of transactions. throughput. In addition, sometimes the data may be modified many times during the transaction process, but eventually returns to the initial value, but the primary key ID and version number of the data will not change. Therefore, if the method of recording the data itself is used, the data may be different when the predicted execution is compared with the actual execution, and a reversal will occur, and the cost of the reversal is relatively high. However, this embodiment adopts the method of only recording the primary key ID and version number of the data, and the probability of inconsistency is low when the predicted execution is compared with the actual execution, and the probability of return is low.

When the database operation device obtains the transaction submission command in the target transaction (that is, the application program server executes the transaction submission command), it can control the corresponding application program server according to the locally recorded database operation command and predicted execution data. The database actually executes the target transaction.

An optional implementation manner of controlling the database corresponding to the application server to actually execute the target transaction according to the locally recorded database operation instructions and predicted execution data includes: the database operation device sends the locally recorded data to the database library operation instruction to instruct the database to execute the database operation instruction, and receive the actual execution result of the database operation instruction returned by the database; then, compare the actual execution result of the database operation instruction with the predicted execution result, if the actual execution If the result is the same as the predicted execution result, issue a transaction commit command to the database for the database to submit the target transaction; if the actual execution result is different from the predicted execution result, issue a transaction return command to the database for the database Forward the target transaction.

In the above embodiments, considering the atomicity of the transaction, by comparing the actual execution result with the predicted execution result, errors in the transaction execution process can be avoided, which is beneficial to improve the success probability of the transaction execution.

Furthermore, in the prior art, since it is impossible to know in advance all the database operation instructions that the transaction needs to execute, the database operation instructions can only be executed sequentially according to the execution logic of the transaction, which will lead to frequent interaction between the application server and the database , which will consume a lot of network resources in remote application scenarios. However, in this embodiment, all database operation instructions that need to be executed by the target transaction, that is, locally recorded database operation instructions, have been obtained in advance through the predictive execution process, so the locally recorded database operation instructions are sent to the database to During the process of instructing the database to execute the database operation instruction, the database operation instructions recorded locally can be sent to the database at the same time (or together), which is beneficial to save network resources. In addition, for some parallel database operation instructions, the database can also be operated in parallel, which is conducive to further improving transaction execution efficiency and increasing transaction throughput.

Furthermore, all the database operation instructions recorded locally together with the predicted execution data recorded locally can also be issued to the database.

Further, in an optional implementation mode in which the database corresponding to the application server is controlled to actually execute the target transaction according to the locally recorded database operation instruction and the predicted execution data, the database operation device may use the locally recorded database operation instruction , to judge whether the target transaction is a stand-alone transaction; if the judgment result is yes, then according to the locally recorded database operation instructions and predicted execution data, the stand-alone transaction processing logic is used to control the database to actually execute the target transaction; if the judgment result is no, then according to The locally recorded database operation instructions and forecast execution data control the database to actually execute the target transaction with the distributed transaction processing logic.

Wherein, it can be identified whether the target transaction is a stand-alone transaction or a distributed transaction based on the database operation instructions obtained during the predictive execution process. For example, it can be judged whether the operation object of the locally recorded database operation command acts on the same physical device, if the judgment result is yes, it can be determined that the target transaction is a stand-alone transaction; if the judgment result is no, it can be determined that the target transaction for distributed transactions.

Because the processing logic of distributed transactions is different from that of stand-alone transactions, the processing logic of stand-alone transactions is relatively simple, for example, it does not involve issues such as read-write locks, so the resources consumed are relatively small. Based on this, in this embodiment, by identifying whether the target transaction is a stand-alone transaction, when it is identified as a stand-alone transaction, the processing logic of the stand-alone transaction is used for processing, instead of uniformly adopting the distributed transaction processing logic for processing as in the prior art, which is beneficial to improve the processing efficiency. efficiency and save processing costs.

In an optional implementation, if the database operation device obtains the transaction return command in the target transaction (that is, the application server executes the transaction return command), it can delete the locally recorded database operation command and predictive execution data, for example, the local memory bank can be cleared directly to realize the transfer operation. In this case, since the target transaction is not actually executed in the database, the database does not need to perform a rollback operation. It can be seen that in the case of transaction rollback, the method of this embodiment can also improve transaction execution efficiency.

It can be seen from the above that in the method provided by this embodiment, the database operation device and the application server cooperate with each other, and the predictive execution process is added to the process of executing the target transaction on the application server. During the predictive execution process, the target can be obtained in advance. The execution path of the transaction, that is, the database operation command that really needs to be executed, and only for the database operation command that belongs to the database modification command, or only for the database operation command record that belongs to the locked non-modification database command The predicted execution data generated by the predicted execution is adapted to the locking scheme under the read-committed isolation level, which provides conditions for the real execution of the transaction, and then when the target transaction is actually executed according to the recorded database operation instructions and the predicted execution data, For example, based on predictive execution data, relevant data information can be obtained in advance and jumps between instructions can be reduced, so execution efficiency can be improved and transaction throughput can be increased.

Taking the inventory reduction business as an example, the workflow of the technical solution of the present invention will be described in detail below. Among them, the isolation level of the database is the isolation level of read committed.

The SQL code for the inventory reduction transaction is as follows, and the text in brackets is a comment: begin transaction (start transaction) select * from inventory where itemId=? for update (use the product id to view the current inventory, note that the external display of for update is used here to lock) if (item.inventory>0) item.inventory--; update inventory set item.inventory=$item.inventory where itemId=? commit; else rollback; (if the current inventory is greater than 0, reduce and update the commodity inventory and submit, otherwise return)

In the actual execution process, the application server is responsible for the entire logic of the inventory reduction transaction, and the database operation device (or it can also be called the execution server) only obtains the initial transaction executed by the application server, and the SQL to be executed statement, and commit/return instructions.

First, after starting the transaction, the database operation device creates a memory bank locally; when the database operation device obtains select * from inventory where itemId=? When the SQL statement for update, it is recognized that this SQL statement will not modify the database, so this SQL statement is directly applied to the actual database, that is, the SQL statement is executed in the database to obtain the query result, that is Read the data set, and return all query results to the application program server, so that the application program server can determine the next SQL statement to be executed, and record this SQL statement in the local memory bank. Although it is the read isolation level, because the statement is locked, it is necessary to record the relevant data of executing the SQL statement. Preferably, only the primary key ID and the version number (version) of the data queried by the SQL statement can be recorded.

After the application server receives the query result returned by the database operation device, it judges whether the current inventory in the query result is greater than 0. If the judgment result is yes, it reduces the current inventory and executes the update statement, that is, update inventory set item .inventory=$item.inventory where itemId=? ; If the judgment result is no, execute the rollback statement.

In the first case, assuming that the application server executes the rollback statement, the database operation device will obtain the rollback statement. Once the database storage device obtains the rollback statement, clear the local memory bank, mainly referring to clearing the previously recorded select * from inventory where itemId=? for update this SQL statement. In this case, there is no need for the database to perform a rollback because the transaction is not executed in the database.

In the second case, suppose the application server executes update inventory set item.inventory=$item.inventory where itemId=? With this statement, the database operation device will obtain update inventory set item.inventory=$item.inventory where itemId=? this sentence. Then the database operation device obtains update inventory set item.inventory=? where itemId=? When this SQL statement is recognized, the SQL statement will modify the database, so this SQL statement is disassembled into a read command, that is, select * from inventory where itemId=? and a write command, update inventory set item.inventory=? where itemId=? ; Next, select * from inventory where itemId=? This statement is sent to the database to obtain the read data set, and write the data in the read data set to the local memory bank; wherein, the read data set includes the primary key list (itemId) and the version number (version) corresponding to the primary key. Then, according to update inventory set item.inventory=? where itemId=? This statement performs an update operation in local memory to obtain a result dataset and returns the result dataset to the application server. The result data set can be modification record information such as how many records have been modified, for example, affectRow=1 means that a record has been modified; or, the result data set can also be modification details information that records modification details. Among them, the result data set and the read data set constitute the forecast execution data.

The application server executes the commit command at this time.

After the database operating device obtains the commit command, it executes the inventory reduction transaction in the database based on the SQL statement in the local memory database and the predicted execution data. This transaction execution process is real. For example, the locally recorded SQL statements and predicted execution data are as follows: {select * from inventory where itemId=? update inventory set item.inventory=? where itemId=? [readVersion: itemId=xxx and version=xxx /affectRow=1] commit()}

Preferably, the database operating device can submit all the SQL statements in the local memory database to the database at one time, and the database executes based on this, which is beneficial to save network resources consumed by transmitting SQL statements.

This actual execution will also return the execution result. You can compare the predicted execution result with the actual execution result. For example, you can first compare the primary key ID and version number in the two results, and then compare the affectRow in the two results; if the comparison result is If they are the same, the transaction can be submitted in the database, and if the comparison result is not the same, the submission is considered to have failed. In this case, the transaction itself has not actually been submitted in the database, so The entire request can simply be redirected and the application server resubmitted.

Among them, after the above-mentioned predictive execution, we have obtained all the SQL statements that need to be executed, as well as the corresponding predictive execution data, such as all segmentation conditions, numerical indexes, version numbers, etc., and these data are enough for us to know in advance whether the transaction is Stand-alone transactions to make judgments. Based on this, corresponding transaction processing logic can be used for processing, which is beneficial to save resources.

FIG. 4 is a schematic flowchart of a database operation method provided by another embodiment of the present invention. As shown in Figure 4, the method includes:

401. During the process of executing the target transaction by the application server, sequentially acquire database operation instructions in the target transaction executed by the application server.

402. Predictively execute the database operation command, return the predictive execution result to the application server so that the application server can determine the next database operation command to be executed, and record the database operation command locally and store it in the database Record the speculative execution data generated by speculative execution when the operation command belongs to the locked non-modifying database command.

403. When the transaction submission instruction in the target transaction is acquired, control the database corresponding to the application server to actually execute the target transaction according to the locally recorded database operation instruction and predicted execution data.

This embodiment is similar to the embodiment shown in FIG. 3 , the difference is that only the predicted execution data generated by the predicted execution of the locked non-modified database type instruction in the target transaction is recorded, and the predicted execution data generated by the modified database type instruction is recorded. and the speculative execution data generated by the speculative execution of unlocked non-modified database instructions are not recorded. In addition, this embodiment is applicable to the situation where the target transaction includes locked non-modified database instructions, and the database to be operated can be not only the database scenario whose isolation level is read committed, but also various other database scenarios .

If the method provided by this embodiment is applied to a database whose isolation level is read and submitted, then it is considered that some databases default to read and submit this isolation level, for example, the Oracle database defaults to read and submit this isolation level Yes, some databases do not set the read-committed isolation level by default. For example, the MySQL database defaults to the repeatable read isolation level, so before executing the method provided by this embodiment, it can be determined whether the isolation level of the database is read. Submitted, if not, you need to pre-configure the database to read submitted this isolation level. For example, you can configure the database to read-committed isolation level before the application server executes the target transaction.

In an optional implementation manner, the application program server sends database operation instructions to the database in an existing manner, so as to realize access to the database. The database operation device can monitor the communication between the application program server and the database, so as to intercept the database operation commands sent by the application program server to the database.

In another optional implementation manner, the processing logic of the application server is slightly modified, and the original processing logic of sending database operation instructions to the database is changed to sending to the database operation device. Based on this, the database operation device can receive the database operation instruction issued by the application server.

It is explained here that if the target transaction is an explicit transaction, the first command in the target transaction is a transaction start command, such as begin transaction. Based on this, the database operation device obtains the data executed by the application server Before the library operation command, the transaction start command in the target transaction sent by the application server to the database can be intercepted, and according to the transaction start command, it is known that the target transaction needs to be executed. Alternatively, before the database operation device obtains the database operation command executed by the application server, it may receive the transaction start command in the target transaction actively issued by the application server, and learn that the target transaction needs to be executed according to the transaction start command. . If the target transaction is an implicit transaction, the first command of the target transaction is the database operation command, excluding the step of obtaining the transaction start command.

After the database operation device obtains the database operation instructions executed by the application program server, it predictively executes the obtained database operation instructions, and returns the prediction execution results to the application program server for the application program server to determine A database operation command to execute. Among them, the predicted execution result determines the execution path of the target transaction, and the execution path here refers to the jump logic between database operation instructions. By returning the predictive execution result to the application server, the application server controls the execution logic of the entire target transaction.

In addition, the database operation device also needs to record the obtained database operation instruction locally, and when the database operation instruction belongs to the locked non-modified database instruction, record the predictive execution of the database operation instruction Generated forecast execution data. Described modification database class instruction refers to the database operation instruction that database has modification, such as the instruction beginning with update, insert etc.; Generally the instruction beginning with select does not belong to modification database class instruction (referred to as non-modification database class instructions). Sometimes, in order to ensure the consistency of the transaction and reduce the probability of return, some non-modifying database commands may need to be locked through the for update statement. For example, in the above account transfer example, the select command needs to be locked to form a select... ...for update database operation instruction, this instruction can be called as locked non-modification database instruction. The predicted execution data mainly refers to some data in the process of predicting and executing the database operation command, such as the data operated by the database operation command and the primary key ID (itemId) of the data and the version number information (version) corresponding to the primary key ID .

It is worth noting that, for the same database operation command, either it belongs to the database modification command, or it belongs to the locked non-modification database command, or it belongs to the unlocked non-modification database command. In this embodiment, the speculative execution data generated by speculative execution of all locked non-modifying database instructions in the target transaction will be recorded locally.

For example, the database operating device can create a memory bank locally, and store the obtained database operation instructions and predicted execution data into the memory bank. Further, if the predictive execution data does not include the predictive execution result, the predictive execution result may also be stored in the memory bank.

In order to realize the simulation of the locking scheme on the basis of the core idea of predictive execution, this embodiment records the predictive execution data generated by the predictive execution for the database operation instructions of the locked non-modified database instructions, but Predictive execution data corresponding to other database operation instructions are not recorded. It can be seen that in the case of implementing the locking scheme, the amount of recorded data is relatively small, so that in the subsequent process of actually executing the target transaction in the database based on locally recorded database operation instructions and predicted execution data, all The amount of data to be processed is relatively small, which is conducive to further improving execution efficiency and performance, and increasing transaction throughput.

More preferably, when the database operation instruction is a locked non-modifying database instruction, only the primary key ID and version number of the data operated by the database operation instruction can be recorded as the predictive execution data. In this way, in the process of actually executing the target transaction in the database based on the locally recorded database operation instructions and predicted execution data, the amount of data to be processed will be further reduced, which is conducive to further improving execution efficiency and performance, and increasing the number of transactions. throughput. In addition, sometimes the data may be modified many times during the transaction process, but eventually returns to the initial value, but the primary key ID and version number of the data will not change. Therefore, if the method of recording the data itself is used, the data may be different when the predicted execution is compared with the actual execution, and a reversal will occur, and the cost of the reversal is relatively high. However, in this embodiment, only the primary key ID and the version number are recorded, so the probability of inconsistency and the turnaround probability are relatively low when the predicted execution is compared with the actual execution.

Wherein, the operation of the database operation command on the database is mainly to access the data in the database, so the data environment of the database operation command can be simulated, and the database operation command can be predicted and executed based on the simulated data environment. Further, considering that the database operation instructions belonging to non-modifying database instructions will not cause changes in the database, so the database operation instructions can be directly predicted and executed in the database, which can save the operation of the simulated data environment, The implementation is relatively simple, which is conducive to saving resources and improving execution efficiency. However, as for the database operation command belonging to the database modification command, because it will cause the change of the database (mainly refers to the change of the data in the database), it is necessary to simulate the data environment and perform the operation in the simulated data environment. Predictive execution.

Based on the above analysis, a method of predictive execution of database operation instructions can be: identify the type of the acquired database operation instructions; The database simulates the data environment of the database operation command in the database, and predicts and executes the database operation command based on the simulated data environment; if the database operation command is a non-modifying database command, execute the database operation command in the database, to perform speculative execution on the database operation instruction.

Further, a data environment for simulating database operation instructions in a locally created memory bank, the implementation of predictive execution of database operation instructions based on the simulated data environment includes: splitting the obtained database operation instructions into A read command and a write command, run the read command on the real database, that is, execute the read command in the database to obtain the read data set (ReadSet), store the read data set in the local memory bank, and Simulate the data environment required by the database operation command. Then, the write command acts on the memory bank to realize the predictive execution of the database operation command, that is, execute the write command in the memory bank to modify the read data set, such as updating or querying the relevant data in the read data set processing etc. Wherein, executing the write command to modify the read data set can generate a result data set (affectRowInMemdb), and the result data set includes the predicted execution result.

When the database operation device obtains the transaction submission command in the target transaction (that is, the application program server executes the transaction submission command), it can control the corresponding application program server according to the locally recorded database operation command and predicted execution data. The database actually executes the target transaction.

An optional implementation manner of controlling the database corresponding to the application server to actually execute the target transaction according to the locally recorded database operation instructions and the predicted execution data includes: the database operation device sends the locally recorded database operation to the database command to instruct the database to execute the database operation command, and receive the actual execution result of the database operation command returned by the database; then, compare the actual execution result of the database operation command with the predicted execution result, if the actual execution result is the same as If the predicted execution result is the same, issue a transaction submission command to the database for the database to submit the target transaction; if the actual execution result is different from the predicted execution result, issue a transaction return command to the database for the database to return the target transaction.

In the above embodiments, considering the atomicity of the transaction, by comparing the actual execution result with the predicted execution result, errors in the transaction execution process can be avoided, which is beneficial to improve the success probability of the transaction execution.

Furthermore, in the prior art, since it is impossible to know in advance all the database operation instructions that the transaction needs to execute, the database operation instructions can only be executed sequentially according to the execution logic of the transaction, which will lead to frequent interaction between the application server and the database , which will consume a lot of network resources in remote application scenarios. However, in this embodiment, all database operation instructions that need to be executed by the target transaction, that is, locally recorded database operation instructions, have been obtained in advance through the predictive execution process, so the locally recorded database operation instructions are sent to the database to During the process of instructing the database to execute the database operation instruction, the database operation instructions recorded locally can be sent to the database at the same time (or together), which is beneficial to save network resources. In addition, for some parallel database operation instructions, the database can also be operated in parallel, which is conducive to further improving transaction execution efficiency and increasing transaction throughput.

Furthermore, all the database operation instructions recorded locally together with the predicted execution data recorded locally can also be issued to the database.

Further, in an optional implementation mode in which the database corresponding to the application server is controlled to actually execute the target transaction according to the locally recorded database operation instruction and the predicted execution data, the database operation device may use the locally recorded database operation instruction , to judge whether the target transaction is a stand-alone transaction; if the judgment result is yes, then according to the locally recorded database operation instructions and predicted execution data, the stand-alone transaction processing logic is used to control the database to actually execute the target transaction; if the judgment result is no, then according to The locally recorded database operation instructions and forecast execution data control the database to actually execute the target transaction with the distributed transaction processing logic.

Wherein, it can be identified whether the target transaction is a stand-alone transaction or a distributed transaction based on the database operation instructions obtained during the predictive execution process. For example, it can be judged whether the operation object of the locally recorded database operation command acts on the same physical device, if the judgment result is yes, it can be determined that the target transaction is a stand-alone transaction; if the judgment result is no, it can be determined that the target transaction for distributed transactions.

Because the processing logic of distributed transactions is different from that of stand-alone transactions, the processing logic of stand-alone transactions is relatively simple, for example, it does not involve issues such as read-write locks, so the resources consumed are relatively small. Based on this, in this embodiment, by identifying whether the target transaction is a stand-alone transaction, when it is identified as a stand-alone transaction, the processing logic of the stand-alone transaction is used for processing, instead of uniformly adopting the distributed transaction processing logic for processing as in the prior art, which is beneficial to improve the processing efficiency. efficiency and save processing costs.

In an optional implementation, if the database operation device obtains the transaction return command in the target transaction (that is, the application server executes the transaction return command), it can delete the locally recorded database operation command and predictive execution data, for example, the local memory bank can be cleared directly to realize the transfer operation. In this case, since the target transaction is not actually executed in the database, the database does not need to perform a rollback operation. It can be seen that in the case of transaction rollback, the method of this embodiment can also improve transaction execution efficiency.

It can be seen from the above that in the method provided by this embodiment, the database operation device and the application server cooperate with each other, and the predictive execution process is added to the process of executing the target transaction on the application server. During the predictive execution process, the target can be obtained in advance. The execution path of the transaction, that is, the database operation instructions that really need to be executed, and only record the predicted execution data generated by the predicted execution for the database operation instructions that belong to the locked non-modified database instructions, so as to adapt to the locking scheme. Provides conditions for the real execution of transactions, and then when the target transaction is actually executed according to the recorded database operation instructions and predicted execution data, for example, based on the predicted execution data, relevant data information can be obtained in advance and jumps between instructions can be reduced, etc., so It can improve execution efficiency and increase transaction throughput.

It should be noted that for the foregoing method embodiments, for the sake of simple description, they are expressed as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence. Because of the present invention, certain steps may be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification belong to preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

In the foregoing embodiments, the descriptions of each embodiment have their own emphases, and for the parts not described in detail in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

Fig. 5 is a schematic structural diagram of a database operation device provided by another embodiment of the present invention. As shown in FIG. 5 , the device includes: an acquisition module 51 , a prediction execution module 52 and a control execution module 53 .

The obtaining module 51 is used for sequentially obtaining the database operation instructions in the target transaction executed by the application server during the execution of the target transaction by the application server.

The predictive execution module 52 is used to predictively execute the database operation instructions, return the predictive execution results to the application server for the application server to determine the next database operation instruction to be executed, and record the database locally The operation instruction and the predictive execution data generated by the predictive execution are recorded when the database operation instruction belongs to one of the modified database type instruction and the locked non-modified database type instruction.

The control execution module 53 is used to control the database corresponding to the application server to actually execute the target transaction according to the locally recorded database operation instructions and predicted execution data when the transaction commit command in the target transaction is obtained.

In an optional implementation, the predictive execution module 52 is in the process of recording the predictive execution data generated by the predictive execution when the database operation instruction belongs to one of the modified database type instruction and the locked non-modified database type instruction, Specifically, it can be used for: when the database operation instruction belongs to the locked non-modifying database instruction, only the primary key ID and the data version number of the data operated by the database operation instruction are recorded as the predictive execution data.

In an optional implementation manner, the device further includes: a configuration module, configured to configure the database as a read-committed isolation level before the application server executes the target transaction.

In an optional implementation manner, the predictive execution module 52 is specifically used for: if the database operation instruction belongs to the instruction to modify the database, simulate the data environment of the database operation instruction in the locally created memory library, based on the simulated data The environment performs predictive execution on the database operation instruction; if the database operation instruction is a non-modifying database instruction, the database operation instruction is executed in the database to predictively execute the database operation instruction.

Further, the predictive execution module 52 simulates the data environment of the database operation instruction in the locally created memory bank when the database operation instruction belongs to the database modification instruction, and executes the database operation instruction based on the simulated data environment. During predictive execution, it is specifically used to: split database operation instructions into read instructions and write instructions; execute read instructions in the database to obtain the read data set, and store the read data set in the memory bank to simulate database operations The data context of the command; executes the write command in the memory bank to modify the read data set.

In an optional implementation, the control execution module 53 can be specifically configured to: issue locally recorded database operation instructions to the database to instruct the database to execute the database operation instructions, and receive the database operation instructions returned by the database The actual execution result; if the actual execution result is the same as the predicted execution result, issue a transaction submission command to the database for the database to submit the target transaction; if the actual execution result is different from the predicted execution result, issue a transaction submission command to the database Return command for the database to return the target transaction.

The database operation device provided in this embodiment cooperates with the application server, adds a predictive execution process in the process of the application server executing the target transaction, obtains and records in advance all the database operation instructions that the target transaction needs to execute, and only Speculative executions resulting from record speculative execution for database operation instructions that belong to the modifying database class instruction and the locked non-modifying database class instruction, or only for database operation instructions that belong to the locked non-modifying database class instruction Data, to adapt to the locking scheme under the read-committed isolation level, provides conditions for the actual execution of transactions, and then controls the database corresponding to the application server to actually execute the target according to the recorded database operation instructions and predicted execution data Transactions are conducive to improving execution efficiency and thereby increasing transaction throughput.

FIG. 6 is a schematic structural diagram of a database operating device provided by another embodiment of the present invention. As shown in FIG. 6 , the device includes: an acquisition module 61 , a prediction execution module 62 and a control execution module 63 .

The obtaining module 61 is used for sequentially obtaining the database operation instructions in the target transaction executed by the application server during the execution of the target transaction by the application server.

The predictive execution module 62 is used to predictively execute the database operation instructions, return the predictive execution results to the application server for the application server to determine the next database operation instruction to be executed, and record the database locally The operation instruction and the speculative execution data generated by the speculative execution are recorded when the database operation instruction belongs to the locked non-modifying database instruction.

The control execution module 63 is used to control the database corresponding to the application server to actually execute the target transaction according to the locally recorded database operation instructions and predicted execution data when the transaction commit command in the target transaction is obtained.

In an optional implementation, the predictive execution module 62 can be specifically configured to: when a database operation instruction belongs to a locked non-modified database instruction, only record the primary key ID and ID of the data operated by the database operation instruction. The version number is used as forecast execution data.

In an optional implementation manner, the device further includes: a configuration module, configured to configure the database as a read-committed isolation level before the application server executes the target transaction.

In an optional implementation manner, the predictive execution module 62 is specifically used to: if the database operation instruction belongs to the instruction to modify the database, simulate the data environment of the database operation instruction in the locally created memory library, based on the simulated data The environment performs predictive execution on the database operation instruction; if the database operation instruction is a non-modifying database instruction, the database operation instruction is executed in the database to predictively execute the database operation instruction.

Further, the predictive execution module 62 simulates the data environment of the database operation instruction in the locally created memory bank when the database operation instruction belongs to the database modification instruction, and executes the database operation instruction based on the simulated data environment. During predictive execution, it is specifically used to: split database operation instructions into read instructions and write instructions; execute read instructions in the database to obtain the read data set, and store the read data set in the memory bank to simulate database operations The data context of the command; executes the write command in the memory bank to modify the read data set.

In an optional implementation, the control execution module 63 can be specifically configured to: issue locally recorded database operation instructions to the database to instruct the database to execute the database operation instructions, and receive the database operation instructions returned by the database The actual execution result; if the actual execution result is the same as the predicted execution result, issue a transaction submission command to the database for the database to submit the target transaction; if the actual execution result is different from the predicted execution result, issue a transaction submission command to the database Return command for the database to return the target transaction.

The database operation device provided in this embodiment cooperates with the application server, adds a predictive execution process in the process of the application server executing the target transaction, obtains and records in advance all the database operation instructions that the target transaction needs to execute, and only For the database operation instructions that belong to the locked non-modified database instructions, record the prediction execution data generated by the prediction execution, so as to adapt to the locking scheme, provide conditions for the real execution of the transaction, and then according to the recorded database operation instructions and Predicting the execution data and controlling the database corresponding to the application server to actually execute the target transaction is beneficial to improve the execution efficiency and further increase the transaction throughput.

Those skilled in the art can clearly understand that for the convenience and brevity of the description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the foregoing method embodiment, which will not be repeated here.

In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or can be Integrate into another system, or some features may be ignored, or not implemented. In another point, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.

The unit described as a separate component may or may not be physically separated, and the component displayed as a unit may or may not be a physical unit, that is, it may be located in one place, or may also be distributed to multiple network units . Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, each unit may exist separately, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented not only in the form of hardware, but also in the form of hardware plus software functional units.

The above-mentioned integrated units realized in the form of software functional units can be stored in a computer-readable storage medium. The above-mentioned software functional units are stored in a storage medium, and include several instructions to enable a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) to execute the various embodiments of the present invention. part of the method. The aforementioned storage media include: USB flash drive, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk, etc., which can store programs. coded media.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: it can still be Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent replacements are made to some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims (10)

A database operation method, characterized in that it comprises: the database operation device sequentially obtains the database operation instructions in the target transaction executed by the application server during the process of executing the target transaction by the application server; the data The library operation device performs predictive execution on the database operation command, returns the predictive execution result to the application server for the application server to determine the next database operation command to be executed, and records the database operation locally Instructions and when the database operation instruction belongs to one of the modified database instruction and the locked non-modified database instruction, record the predictive execution data generated by the predictive execution; when the database operation device obtains the target transaction When submitting an instruction for a transaction, control the database corresponding to the application server to actually execute the target transaction according to the locally recorded database operation instruction and the predicted execution data generated by using the recorded predicted execution, including the following steps: issuing the database operation instruction recorded locally to the database to instruct the database to execute the database operation instruction, and receiving the actual execution result of the database operation instruction returned by the database; if the actual execution result is the same as If the predicted execution result is the same, issue the transaction submission command to the database for the database to submit the target transaction; if the actual execution result is different from the predicted execution result, issue a transaction return ( rollback) command for the database to roll back the target transaction. According to the method of item 1 of the scope of patent application, wherein, when the database operation instruction belongs to one of the modified database type instruction and the locked non-modified database type instruction, the predicted execution data generated by the predicted execution is recorded, including : When the database operation instruction is a locked non-modifying database instruction, only record the primary key ID and version number of the data operated by the database operation instruction as the predictive execution data. The method according to item 1 of the scope of the patent application further includes: before the application server executes the target transaction, configuring the database as a read-committed isolation level. According to the method of item 1 of the scope of the patent application, wherein the predictive execution of the database operation instruction includes: if the database operation instruction belongs to the instruction of modifying the database, simulating the database in the locally created memory bank The data environment of the operation command, based on the simulated data environment, the database operation command is predicted and executed; if the database operation command belongs to the non-modified database type command, the database operation command is executed in the database to The database operation instruction performs speculative execution. According to the method of item 4 of the scope of patent application, wherein, the data environment required by the database operation instruction is simulated in the locally created memory library, and the database operation instruction is predicted and executed based on the simulated data environment, including : Split the database operation command into a read command and a write command; Execute the read command in the database to obtain a read data set, store the read data set in the memory bank to simulate the data environment of the database operation command; execute the write command in the memory bank to modify It's time to read the dataset. A database operation device, characterized in that it includes: an acquisition module, used to sequentially acquire database operation instructions in the target transaction executed by the application server during the execution of the target transaction by the application server; The predictive execution module is used to predictively execute the database operation command, return the predictive execution result to the application program server for the application program server to determine the next database operation command to be executed, and record it locally The database operation instruction and when the database operation instruction belongs to one of the database modification instruction and the locked non-modification database instruction, record the prediction execution data generated by the prediction execution; the control execution module is used for when When the transaction commit command in the target transaction is obtained, control the database corresponding to the application server to actually execute the target according to the locally recorded database operation command and the predicted execution data generated by using the recorded predicted execution transaction, wherein the control execution module is specifically used to: issue the database operation instruction recorded locally to the database, instruct the database to execute the database operation instruction, and receive the database operation instruction returned by the database The actual execution result of the operation instruction; if the actual execution result is the same as the predicted execution result, issue the transaction submission instruction to the database for the database to submit the target transaction; If the actual execution result is different from the predicted execution result, issue a transaction rollback (rollback) command to the database for the database to roll back the target transaction. The device according to item 6 of the scope of the patent application, wherein the predictive execution module is specifically used for: when the database operation instruction belongs to a locked non-modified database instruction, only record the database operation instruction operated The primary key ID and version number of the data are used as the forecast execution data. The device according to item 6 of the scope of the patent application further includes: a configuration module configured to configure the database as a read-committed isolation level before the application server executes the target transaction. According to the device according to item 6 of the scope of the patent application, the predictive execution module is specifically used for: if the database operation instruction belongs to the modification database instruction, simulate the data of the database operation instruction in the locally created memory bank Environment, based on the simulated data environment, predictively execute the database operation instruction; if the database operation instruction is a non-modified database instruction, execute the database operation instruction in the database to operate the database instructions are speculatively executed. The device according to item 9 of the scope of the patent application, wherein the predictive execution module is specifically used to: split the database operation instruction into a read instruction and a write instruction; execute the read instruction in the database to obtain a read data set , which will read The data set is stored in the memory bank to simulate the data environment of the database operation command; the write command is executed in the memory bank to modify the read data set.

Images