WO2013026287A1

WO2013026287A1 - Data control method and system based on distributed database system

Info

Publication number: WO2013026287A1
Application number: PCT/CN2012/074552
Authority: WO
Inventors: 徐宜良; 常二鹏
Original assignee: 中兴通讯股份有限公司
Priority date: 2011-08-25
Filing date: 2012-04-23
Publication date: 2013-02-28
Also published as: CN102955801A; CN102955801B

Abstract

Disclosed are a data control method and system based on a distributed database system. The method comprises: a master server node receiving a storage procedure and function execution command, and sending, according to the storage procedure and function execution command, a transaction start notification to a non-master server node, and the non-master server node creating a new transaction; when executing a processing statement, the master server node sending the processing statement to the non-master server node; the non-master server node executing the processing statement, and returning an obtained processing result to the master server node; the master server node sending a transaction submitting notification to any one or multiple non-master server nodes, and the non-master server node receiving the transaction submitting notification and submitting the created new transaction. By means of the present invention, the storage procedure and function can access and operate data stored in hard disks of multiple server nodes, thereby improving the utilization ratio of resources and execution efficiency of the system.

Description

Data control method and system based on distributed database system

The present invention relates to the field of communications, and in particular, to a data control method and system based on a distributed database system. Background technique

With the maturity of traditional database technologies and the rapid development of Internet applications, database applications have generally appeared on the Internet. The traditional centralized database system also shows its shortcomings: Centralized management, it is bound to cause performance bottlenecks, a single improvement in hardware performance can not meet the current large data volume of the Internet, high concurrent query write requirements, A lot of hardware maintenance and upgrade costs are incurred. The centralized management of the database often results in multiple database associations, and once a database crashes, the entire system will crash. Therefore, distributed databases will gradually replace traditional centralized database systems.

In a distributed database system, multiple server nodes form a server cluster, and data is distributed on multiple server nodes, and data is processed synchronously by using a dedicated central processing unit (CPU) and a hard disk under each node. Each server node has equal status, including: a structured query language (SQL, STRUCTURED QUERY LANGUAGE) parsing module, an execution module, an inter-node communication module, a data storage module, and the like. The client selects a server node to access the distributed database system through a certain consistency algorithm (such as: HASH algorithm).

For the traditional centralized database system stored procedures and functions, because of its centralized management features, the execution of stored procedures and functions can only use the CPU of one server node, the table or view in the database instance under the same server node. In a distributed database system, due to the multi-server cluster environment, the data tables are distributed on multiple server nodes. The execution mode of stored procedures and functions in the traditional centralized database system cannot meet the distributed data requirements. In view of the above prior art, since the stored procedures and functions cannot be executed on the distributed database system multiple server nodes, the CPU, the hard disk, and the like of all the server nodes cannot be fully utilized, and the server execution efficiency is reduced. s solution. Summary of the invention

The main object of the present invention is to provide a data control method and system based on a distributed database system, so as to solve the problem in the prior art that stored procedures and functions cannot be executed on multiple server nodes of a distributed database system, resulting in failure to fully utilize all of them. The server node's CPU, hard disk and other resources reduce the server's execution efficiency.

In order to achieve the above object, according to an aspect of the present invention, a data control method based on a distributed database system is provided.

The data control method based on the distributed database system of the present invention comprises: the main control server node receives the stored procedure and the function execution command, and sends a transaction start notification to any one or more non-master server nodes according to the stored procedure and the function execution command, and receives The non-master server to the transaction start notification creates a new transaction; executes the processing statements in the stored procedure and function at the master server node, and sends the processing statement to any one or more non-master server nodes; the non-master server The node executes the processing statement, and returns the obtained processing result to the master server node; the master server node executes all processing statements of the stored procedure and the function according to the processing result, and sends a transaction commit notification to any one or more The non-master server node, the non-master server node receives the transaction commit notification and submits the created new transaction.

Further, before the master server node receives the stored procedure and the function execution command, the method further includes: the master server receives a create command for creating a stored procedure and a function; and parses the vocabulary and syntax of the stored procedure and the function according to the create command, and Record the contents of each statement in the stored procedure and function.

Further, before the master server node executes the processing statement in the stored procedure and the function, the method further includes: the master server node replaces the variable name in the processing statement with a specific variable value Change.

Further, after the master server node sends a transaction start notification to any one or more non-master server nodes according to the stored procedure and the function execution command, the method further includes: determining whether the master server node sends the transaction start notification successfully, In the case of a failed transmission, the system ends the stored procedure and function execution.

Further, after the master server node sends a transaction start notification to any one or more non-master server nodes according to the stored procedure and the function execution command, after the non-master server that receives the transaction start notification creates a new transaction, the method further The method includes: returning, by the non-master server node, a result of processing the transaction start notification to the master server node; determining whether the time at which the master server node receives the processing transaction start notification result exceeds a predetermined time, wherein, in a case that the predetermined time is exceeded, Execution of the stored procedure and function command fails; if the non-master server node received by the master server node processes the transaction start notification result in the case where the predetermined time is not exceeded, the start transaction operation of any of the nodes fails, and then execution is performed. Stored procedures and functions fail.

Further, after the master server node executes the processing statement in the stored procedure and the function, and sends the processing statement to any one or more non-master server nodes, the method further includes: determining whether the processing node sends the processing statement Failed, in the case of a failed delivery, the execution of the stored procedure and function failed this time.

Further, after the non-master server node executes the processing statement and returns the obtained processing result to the master server node, the method further includes: executing a processing statement on all non-master server nodes received by the master server node In the result, if any non-master server node fails to execute the processing statement, the execution of the stored procedure and the function fails; or the time that the master server node waits for any non-master server node to execute the processing statement exceeds the time. After a certain time limit, the stored procedure and function execution failed.

In order to achieve the above object, according to another aspect of the present invention, a data control system based on a distributed database system is provided. The data control system based on the distributed database system of the present invention comprises: a master control server and one or more non-master control servers, wherein

The master server is set to receive the stored procedure and function execution commands issued by the client, send a transaction start notification according to the stored procedure and the function execution command, execute the processing statement in the stored procedure and the function, and execute all processing of the stored procedure and the function. After the statement, send a transaction commit notification;

One or more non-master servers, set to create a new transaction after receiving a transaction start notification, execute the received processing statement, and submit the created new transaction after receiving the transaction commit notification.

Further, the master server further includes: a creating unit, configured to receive a creation command issued by the client to create a stored procedure and a function, and parse the vocabulary and syntax of the stored procedure and the function according to the create command, and record the stored procedure and function in the record Each sentence handles the contents of the statement.

Further, the main control server further includes: a first judging module, configured to determine whether the main control server node sends a transaction start notification is successful, in the case that the sending fails, the system ends the storing process and the function performing the operation; the second determining module, Set to determine whether the time when the master server node receives the result of processing the transaction start notification exceeds a predetermined time, wherein, in the case that the predetermined time is exceeded, the execution of the stored procedure and the function command fails; if the predetermined time is not exceeded, if the master If all the non-master server nodes received by the control server node process the transaction start notification result, if the start transaction operation of any node fails, the execution of the stored procedure and the function fails; the third judgment module is set to determine that the master server node sends Whether the processing statement fails or not, in the case of a failed transmission, the execution of the stored procedure and function fails this time.

Through the invention, the master server node is used to receive the stored procedure and the function execution command; the master server node sends a transaction start notification to any one or more non-master server nodes according to the stored procedure and the function execution command, and receives the transaction start notification. The non-master server creates a new transaction; executes the processing statements in the stored procedure and function at the master server node, and processes The statement is sent to any one or more non-master server nodes; the non-master server node executes the processing statement, and returns the obtained processing result to the master server node; the master server node executes the stored procedure according to the processing result and All processing statements of the function, and send a transaction commit notification to any one or more non-master server nodes. The non-master server node receives the transaction commit notification and submits the created new transaction, which solves the prior art due to stored procedures and functions. Cannot be executed on a distributed database system multi-server node, which can not fully utilize the CPU, hard disk and other resources of all server nodes, which reduces the problem of server execution efficiency, and thus achieves that stored procedures and functions can store data on multiple server node hard disks. Access and operations improve the utilization of resources and the efficiency of system execution. DRAWINGS

The drawings are intended to provide a further understanding of the present invention, and are intended to be a part of the invention. In the drawing:

1 is a schematic structural diagram of a data control system based on a distributed database system according to an embodiment of the present invention;

2 is a flow chart of a distributed database method according to an embodiment of the present invention; FIG. 3 is a flowchart of a stored procedure and function creation process according to an embodiment of the present invention; FIG. 4 is a distributed database method according to an embodiment of the present invention; Detailed process diagram. detailed description

The present invention will be further described in detail below with reference to the accompanying drawings and embodiments in order to make the present invention. It is understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The invention provides a data control method and system based on a distributed database system. 1 is a data control system based on a distributed database system according to an embodiment of the present invention. Schematic.

As shown in FIG. 1, the system includes: a client 10, a master server node 30, and a non-master server node 50.

The master server node 30 is configured to receive a stored procedure and a function execution command issued by the client 10, and send a transaction start notification according to the stored procedure and the function execution command, and execute the processing statement in the stored procedure and the function at the same time. After all the processing statements of the stored procedure and the function are sent, a transaction commit notification is sent; and one or more non-master server nodes 50 are configured to create a new transaction after receiving the transaction start notification, and simultaneously execute the received processing statement, and Submit the created new transaction after receiving the transaction commit notification. Preferably, the processing statement in the above embodiment of the present invention is an SQL statement.

The foregoing embodiment of the present invention implements a method for implementing a stored procedure and a function execution in a distributed database multi-server cluster. The distributed database system is composed of a plurality of server nodes, and the client 10 adopts a certain consistency algorithm (for example: HASH algorithm). Selecting a server node to access the distributed database system, that is, the server node with the TCP link linked to the client 10 is called the master server node, and the server node without the link link is called the non-master server node 50. . The above embodiment solves the problem that the stored procedure and the function cannot be executed on the multi-server node of the distributed database system by establishing the communication data feedback mechanism, so that the stored procedure and the function can store the data on the multi-server node hard disk. By performing access and operations, the stored procedures and functions can be executed concurrently on multiple server nodes of the distributed database, and the CPU, hard disk and other resources of each server node can be fully utilized, thereby greatly improving the execution efficiency.

The master server in the foregoing embodiment may further include: a creating unit, configured to receive a create command for creating a stored procedure and a function issued by the client 10, and parse the vocabulary and syntax of the stored procedure and the function according to the create command, and simultaneously save and save the record The contents of each statement in the stored procedure and function. Preferably, the master control server may further include: a first determining module, configured to determine whether the master server node sends a transaction start notification is successful, and in case the sending fails, the system ends the storing process and the function performing the operation; a module, configured to determine whether the time when the master server node receives the result of processing the transaction start notification exceeds a predetermined time, wherein, in the case that the predetermined time is exceeded, the execution of the stored procedure and the function command fails; if the predetermined time is not exceeded, If all of the non-master server nodes 50 received by the master server node process the transaction start notification result, and the start transaction operation of any of the nodes fails, the execution of the stored procedure and the function fails. And a third judging module, configured to determine whether the main control server node fails to send the processing statement, and if the sending fails, executing the stored procedure and the function fails to perform this time.

2 is a flow chart of a distributed database based method in accordance with an embodiment of the present invention. As shown in Figure 2, the method includes the following steps:

Step S102, the master server node 30 receives the stored procedure and the function execution command.

Step S104, the master server node 30 sends a transaction start notification to any one or more non-master server nodes 50 according to the stored procedure and the function execution command, and the non-master server node that receives the transaction start notification creates a new transaction.

Step S106, executing a processing statement in the stored procedure and the function at the master server node, and sending the processing statement to any one or more non-master server nodes 50.

Step S108: The non-master server node 50 executes the processing statement, and returns the obtained processing result to the master server node.

In step S110, the master server node 30 executes all the processing statements of the stored procedure and the function according to the processing result, and sends a transaction commit notification to any one or more non-master server nodes 50, and the non-master server node 50 receives the transaction commit notification. And submit the new transaction created.

The foregoing embodiment implements a storage procedure and a function execution implementation method in a server cluster environment composed of multiple server nodes, and the processing statement involved may be a SQL statement, which solves the problem that the stored procedure and the function cannot be in the distributed database system. Problems executed on the server node, Enables stored procedures and functions to access and manipulate data stored on multiple server nodes. It allows concurrent execution of stored procedures and functions on multiple server nodes in a distributed database, enabling full utilization of CPU, hard disk, etc. of each server node. Resources have greatly improved the efficiency of execution.

In the above embodiment, before the master server node receives the stored procedure and the function execution command, the method may further include: the master server receiving the create command for creating the stored procedure and the function; and parsing the vocabulary of the stored procedure and the function according to the create command Grammar, and records the contents of each statement in the stored procedure and function.

3 is a flow chart of a stored procedure and function creation process, as shown in FIG. 3, in accordance with an embodiment of the present invention. The creation process includes the following steps:

Step 201, the client 10 sends a create stored procedure and a function command to the master server node.

30.

Step 202: The distributed database system lexically and syntaxally parses the stored procedures and functions, and records and saves each processing statement, such as an SQL statement.

Specifically, the foregoing embodiment may be implemented when a stored procedure and a function are created in a distributed database system, that is, when a lexical and syntax analysis is performed on a stored procedure and a function in a structured query language (STRUCTURED QUERY LANGUAGE, SQL for short). Record the contents of each SQL statement in the stored procedure and function.

Preferably, before the master server node 30 executes the processing statements in the stored procedure and the function, the method may further include: the master server node 30 replacing the variable name in the processing statement with a specific variable value.

In the foregoing embodiment of the present invention, after the master server node 30 sends a transaction start notification to any one or more non-master server nodes 50 according to the stored procedure and the function execution command, the method may further include: determining that the master server node 30 sends The transaction start notification is successful. In the case of a failed transmission, the system ends the stored procedure and function execution. In the above embodiment of the present invention, the master server node 30 transmits a transaction start notification to any one or more non-master server nodes 50 according to the stored procedure and the function execution command, and the non-master server node 50 that has received the transaction start notification is created. After a new transaction, the method may further include: the non-master server node 50 returns a result of processing the transaction start notification to the master server node; determining whether the time at which the master server node 30 receives the processing transaction start notification result exceeds a predetermined time, wherein In the case where the predetermined time is exceeded, the execution of the stored procedure and the function command fails; if the predetermined time is not exceeded, if all the non-master server nodes 50 received by the master server node process the transaction start notification result, If a node's start transaction operation fails, the stored procedure and function fail.

In the above embodiment of the present invention, after the master server node 30 executes the processing statement in the stored procedure and the function, and sends the processing statement to any one or more non-master server nodes 50, the method may further include: determining the master Whether the server node 30 sends a processing statement fails, and if the transmission fails, the execution of the stored procedure and the function fails this time.

In the above embodiment of the present invention, after the non-master server node 50 executes the processing statement and returns the obtained processing result to the master server node, the method further includes: all non-masterings received at the master server node 30 In the result of executing the processing statement by the server node 50, if any non-master server node 50 fails to execute the processing statement, the execution of the stored procedure and the function fails; or the host server node waits for any non-master server node 50. When the execution of the result of the processing statement exceeds a certain time limit, the stored procedure and function execution fails.

4 is a detailed flow diagram of a method based on a distributed database according to an embodiment of the present invention. As shown in FIG. 4, a stored procedure and a function are executed on a multi-server node of a distributed database system, and a multi-server node hard disk can be executed during execution. The operation of storing data includes the following steps: Step 301: The client 10 sends a stored procedure and a function execution command to the master server node.

30;

Step 302, the master server node 30 sends a transaction start notification to other non-master servers. Node 50. Specifically, the step may be implemented. When the master server node 30 of the distributed database system executes the stored procedure and function, the master server node 30 first sends a stored procedure and a function transaction start notification to the other non-master server node 50, and other non- The master server node 50 creates a new transaction after receiving the transaction start notification, and sends the processing result to the master server node 30.

Preferably, during the execution of the step, if the master server node 30 sends a transaction start notification to any non-master server node 50, it is considered that the stored procedure and function send an error during the execution, and end the stored procedure. And the function performs the operation.

Meanwhile, after the non-master server node 50 receives the transaction start notification transmitted from the master server node 30, the non-master server node 50 starts creating a new transaction and transmits the processing result to the master server node.

The master server node 30 waits for any non-master server node 50 to process the transaction start notification. After the time exceeds a certain time limit, the stored procedure and function execution is considered to have failed, and the stored procedure and function execution operation are ended.

All non-master server nodes 50 received by the master server node 30 process the transaction start notification result. If the start transaction operation of any node fails, the stored procedure and the function execution are considered to be failed, and the stored procedure and the function execution operation are ended. .

Step 303: The master server node 30 replaces the variable name with a specific variable value before executing each SQL statement. Specifically, the step can implement the master server of the distributed database system. Before each SQL statement in the stored procedure and function is executed, the variable value in the content of the SQL statement is replaced by the actual value of the variable.

In step 304, the master server node 30 sends the SQL statement to the other non-master server node 50. Specifically, the step may be that the master server node of the distributed database system executes the SQL statement on the local machine. If the SQL statement needs data stored by the hard disk of the other non-master server node 50, the master server node 30 The SQL statement is sent to the corresponding non-master server node 50 through a network protocol, and the non-master server node 50 receives and executes the SQL statement. And send back the execution result.

During the execution of this step, if the master server node 30 fails to send a SQL statement to any of the non-master server nodes 50, it is considered that the stored procedure and function send an error during the execution, and end the stored procedure and function execution. operating.

All the non-master server nodes 50 received by the master server node 30 execute the SQL statement result. If any node fails during the execution of the SQL statement, the stored procedure and the function execution are considered to be invalid, and the stored procedure and the function execution operation are ended. .

The master server node 30 waits for any non-master server node 50 to execute the SQL statement result for more than a certain time limit, then considers the stored procedure and function execution to fail, and ends the stored procedure and function execution operation.

In the above embodiment, when the master server node 30 determines that the storage process and the function execution fail during the execution, the master server node 30 needs to notify the non-master server node 50 to process the transaction rollback before ending the stored procedure and the function execution operation. Specifically, the master server node 30 may send a transaction rollback notification to the other non-master server node 50, and the other non-master server node 50 performs the transaction rollback process after receiving the notification, and sends back the processing result.

At the same time, the SELECT statement, the DELETE statement, and the UPDATE statement in the above process are sent to other non-master server nodes 50. The INSERT statement calculates the specific server node to be sent according to the consistency algorithm of the data distribution.

Step 305, the non-master server node 50 receives the SQL statement, and executes the SQL statement, and sends the processing result to the master server node 30;

Step 306, after executing all the SQL statements of the stored procedure and the function, the master server node 30 sends a transaction commit notification to the other non-master server node 50, and the non-master server node 50 receives the transaction commit notification and submits the transaction, and sends the processing result. Send to the master server node 30.

Specifically, the foregoing two steps may be implemented after the main control server node of the distributed database system executes the SQL statement in the stored procedure and the function, and sends a commit transaction notification to the other non-master control. The server node 50, the other non-master server node 50 receives the commit transaction notification and submits the transaction, and sends the processing result to the master server node.

If the master server node 30 sends a transaction commit notification to any of the non-master server nodes 50, it is considered that the stored procedure and function send an error during this execution and ends the stored procedure and function execution.

After receiving the transaction commit notification sent by the master server node for more than a certain time limit, the non-master server node 50 considers that the stored procedure and function execution failed, and performs a transaction rollback processing operation.

It should be noted that the steps shown in the flowchart of the accompanying drawings of the embodiments of the present invention may be executed in a computer system such as a set of computer executable instructions, and although a logical sequence is shown in the flowchart, In some cases, the steps shown or described may be performed in an order different than that described herein.

From the description of the above embodiments, it can be seen that the present invention achieves the following technical effects: The stored procedure and the function execution method of the present invention solve the problem that the stored procedure and function are executed on the distributed server system multiple server nodes, which may be more The server node hard disk stores data for operation; the stored procedure and function execution method of the present invention can enable the stored procedure and function to be executed concurrently on multiple server nodes of the distributed database, thereby greatly improving the execution efficiency.

Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general-purpose computing device, which can be concentrated on a single computing device or distributed over a network composed of multiple computing devices. Alternatively, they may be implemented by program code executable by the computing device, such that they may be stored in the storage device by the computing device, or they may be separately fabricated into a plurality of integrated circuit modules, or they may be Multiple modules or steps in the fabrication are implemented as a single integrated circuit module. Thus, the invention is not limited to any specific combination of hardware and software.

The above description shows and describes a preferred embodiment of the invention, but as mentioned above, It is to be understood that the invention is not to be construed as being limited to the details of the embodiments disclosed herein. Or changes in the technology or knowledge of the relevant field. All changes and modifications made by those skilled in the art are intended to be within the scope of the appended claims.

Claims

Claim

A data control method based on a distributed database system, wherein the method comprises: a master server node receiving a stored procedure and a function execution command, and transmitting a transaction start notification to any one or more according to the stored procedure and a function execution command a non-master server node, the non-master server node receiving the transaction start notification creates a new transaction;

Performing, in the master server node, the processing statement in the stored procedure and the function, and sending the processing statement to any one or more of the non-master server nodes;

The non-master server node executes the processing statement, and returns the obtained processing result to the main control server node;

The master server node executes all processing statements of the stored procedure and the function according to the processing result, and sends a transaction commit notification to any one or more of the non-master server nodes, and the non-master server node receives the The transaction commits the notification and submits the new transaction created.

2. The data control method according to claim 1, wherein before the master server node receives the stored procedure and the function execution command, the method further includes:

The master server node receives a create command to create the stored procedure and function; parses the vocabulary and syntax of the stored procedure and function according to the create command, and records and saves each sentence in the stored procedure and function The content of the statement.

The data control method according to claim 2, wherein before the execution of the processing statement in the stored procedure and the function, the method further includes:

The master server node replaces the variable name in the processing statement with a specific variable value.

The data control method according to any one of claims 1 to 3, wherein, at the master server node, a transaction start notification is sent to any one or more non-masters according to the stored procedure and a function execution command. After the server node, the method further includes:

Determining whether the master server node sends the transaction start notification is successful, and in case the transmission fails, the system ends the stored procedure and the function execution operation.

The data control method according to claim 4, wherein the master server node sends a transaction start notification to any one or more non-master server nodes according to the stored procedure and a function execution command, and receives the After the non-master server node that the transaction start notification creates a new transaction, the method further includes:

The non-master server node returns a result of processing a transaction start notification to the master server node;

Determining whether the time when the master server node receives the processing transaction start notification result exceeds a predetermined time, where

In the case that the predetermined time is exceeded, the execution of the stored procedure and the function command fails; if the predetermined time is not exceeded, if all the non-master server nodes received by the master server node process the transaction start In the result of the notification, if the start transaction operation of any of the nodes fails, the execution of the stored procedure and the function fails.

The data control method according to any one of claims 1 to 3, wherein a processing statement in the stored procedure and function is executed at the master server node, and the processing statement is sent to any one of After the plurality of non-master server nodes, the method further includes: determining whether the master server node sends the processing statement fails, and if the sending fails, performing the storing process and the function fails to perform this time .

The method of claim 6, wherein after the non-master server node executes the processing statement and returns the obtained processing result to the master server node, the method further includes:

In the execution statement result of all non-master server nodes received by the master server node, if any non-master server node fails to execute the processing statement, the execution of the stored procedure and function fails;

The stored procedure and function execution fails after the master server node waits for any non-master server node to execute the processing statement result for more than a certain time limit.

8. A data control system based on a distributed database system, wherein the system comprises: a master server and one or more non-master servers, wherein

a master server, configured to receive a stored procedure and a function execution command, send a transaction start notification according to the stored procedure and a function execution command, execute a processing statement in the stored procedure and the function, and execute all processing of the stored procedure and the function After the statement, a transaction commit notification is sent; one or more non-master servers are set to create a new transaction after receiving the transaction start notification, execute the received processing statement, and receive the transaction commit Submit the new transaction created after the notification.

9. The data control system according to claim 8, wherein the master server further comprises:

a creating unit, configured to receive a create command issued by the client to create the stored procedure and function, and parse the vocabulary and syntax of the stored procedure and function according to the create command, and record the save in the stored procedure and function The content of each sentence is processed.

The data control system according to claim 8, wherein the master server further comprises:

a first determining module, configured to determine whether the master server node sends the transaction start notification is successful, in the case that the sending fails, the system ends the storing process and the function performing the operation; the second determining module is configured to determine the Whether the time when the master server node receives the processing transaction start notification result exceeds a predetermined time, wherein, in the case that the predetermined time is exceeded, the execution of the stored procedure and the function command fails; if the predetermined time is not exceeded In the case that if all the non-master server nodes received by the master server node process the transaction start notification result, if any node's start transaction operation fails, executing the stored procedure and function fails;

The third judging module is configured to determine whether the processing of the processing statement by the main control server node fails. If the sending fails, the execution of the stored procedure and the function fails.