C++ operation MySQL large data insertion inefficient solution

Generally speaking, when C++ operates MySQL, it is very slow to insert 10,000 pieces of simple data into MySQL. It takes about 5 minutes.
And open the transaction, less than a second to be done!

The specific implementation code is as follows:

#include <iostream>
#include <winsock2.h>
#include <string>

#include "mysql.h"

#pragma comment(lib, "libmysql.lib");

using namespace std;

int main()
{
 MYSQL mysql;
 mysql_init(&mysql); //Initialize the

 MYSQL *ConnStatus = mysql_real_connect(&mysql,"localhost","root","","sky",3306,0,0);
 if (ConnStatus == NULL)
 {
 //The connection fails
 int i = mysql_errno(&mysql);
 string strError= mysql_error(&mysql);
 cout <<"Error info: "<<strError<<endl;

 return 0;
 }


 cout<<"Mysql Connected..."<<endl;
 
 string strsql;
 MYSQL_RES *result=NULL; //Data result set

 //The insert
 strsql = "insert into t1 values(2,'lyb')";

 mysql_query(&mysql,"START TRANSACTION"); //Open transactions, if not open transactions, then the efficiency is very low!

 for (int i=0; i<10000; i++)
 {
 mysql_query(&mysql,strsql.c_str());
 }

 mysql_query(&mysql,"COMMIT");   //Commit the transaction

 cout<<"insert end"<<endl;

 
 //Releasing the result set closes the database
 mysql_free_result(result);
 mysql_close(&mysql);
 mysql_library_end();

 return 0;
}

Conclusion:

In the case of a large number of data inserts, changes, etc., the transaction should be started, after the completion of a series of operations, then commit the transaction, can improve the execution efficiency of the program.