测试mysql的sql语句预编译效果

玩Oracle的都比较关注shared pool，特别是library cache，在使用了绑定变量（预编译sql）之后确实能得到很大的性能提升。现在在转Mysql之后特别是innodb很多东西都还能和Oracle对得上号的，就像innodb_buffer_pool_size类似于Oracle的database buffer cache，innodb_log_buffer_size类似于redo log buffer，但是innodb_additional_mem_pool_size仅仅类似于shared pool的Data dictionary cache，似乎还缺少和library cache相对应的东西。那就有一个问题了，在Mysql里面使用预编译的sql还会有性能提升吗？

这里我用Java的jdbc做了一下测试，分别用Statement和PreparedStatement执行1000个sql，并运行10次

1.使用Statement做硬解析：

 package exmysql;

 import java.sql.Connection;
 import java.sql.DriverManager;
 import java.sql.ResultSet;
 import java.sql.SQLException;
 import java.sql.Statement;
 import java.util.Date;

 public class adddata {

     private static long worker(){
         Date begin = new Date(); 

         String driver="com.mysql.jdbc.Driver";

         String url="jdbc:mysql://172.16.2.7:3306/testdb";

         Connection conn=null;
         Statement stmt=null;
         ResultSet rs=null;

         try{
             Class.forName(driver);
             conn=DriverManager.getConnection(url,"dbaadmin","123456");
             stmt=conn.createStatement();
             String sql;
             for (int i=1;i<=5000;i++){
                 sql="select * from test1 where id="+i;
                 rs=stmt.executeQuery(sql);
             }
         }
         catch(SQLException | ClassNotFoundException e){
             e.printStackTrace();
         }

         if(stmt!=null){
             try{
                 stmt.close();
             }
             catch(SQLException e){
                 e.printStackTrace();
             }
         }

         if(conn!=null){
             try{
                 conn.close();
             }
             catch(SQLException e){
                 e.printStackTrace();
             }
         }

         Date end = new Date();
         return end.getTime()-begin.getTime();
     }

     public static void main(String[] args) {
         // TODO Auto-generated method stub

         long elapsed,average;
         average=0;
         for (int i=1;i<=10;i++){
             elapsed=worker();
             System.out.println("elapsed time(ms):"+elapsed);
             average=average+elapsed;
         }
         System.out.println("average time(ms):"+average/10);
     }

 }

结果如下：

elapsed time(ms):24652
elapsed time(ms):13380
elapsed time(ms):13250
elapsed time(ms):13877
elapsed time(ms):13275
elapsed time(ms):13193
elapsed time(ms):19022
elapsed time(ms):13558
elapsed time(ms):14138
elapsed time(ms):13364
average time(ms):15170

2.同样的sql用PreparedStatement预编译执行

 package exmysql;

 import java.sql.*;
 import java.util.Date;

 public class insert_data {

     private static long worker(){
         Date begin = new Date(); 

         String driver="com.mysql.jdbc.Driver";

         String url="jdbc:mysql://172.16.2.7:3306/testdb";

         Connection conn=null;
         PreparedStatement pstm=null;
         ResultSet rs=null;

         try{
             Class.forName(driver);
             conn=DriverManager.getConnection(url,"dbaadmin","123456");
             String sql="select * from test1 where id=?";
             pstm=conn.prepareStatement(sql);
             for(int i=1;i<=5000;i++){
                 pstm.setInt(1, i);
                 rs=pstm.executeQuery();
             }

         }
         catch(SQLException | ClassNotFoundException e){
             e.printStackTrace();
         }

         if(pstm!=null){
             try{
                 pstm.close();
             }
             catch(SQLException e){
                 e.printStackTrace();
             }
         }

         if(conn!=null){
             try{
                 conn.close();
             }
             catch(SQLException e){
                 e.printStackTrace();
             }
         }

         Date end = new Date();
         return end.getTime()-begin.getTime();
     }

     public static void main(String[] args) {
         // TODO Auto-generated method stub

         long elapsed,average;
         average=0;
         for (int i=1;i<=10;i++){
             elapsed=worker();
             System.out.println("elapsed time(ms):"+elapsed);
             average=average+elapsed;
         }
         System.out.println("average time(ms):"+average/10);
     }

 }

结果如下：

elapsed time(ms):14773
elapsed time(ms):16352
elapsed time(ms):14797
elapsed time(ms):15800
elapsed time(ms):12069
elapsed time(ms):14953
elapsed time(ms):13238
elapsed time(ms):12366
elapsed time(ms):15263
elapsed time(ms):13089
average time(ms):

可以看出两种方式执行的结果几乎相同，不像Oracle差距那么大。而且就算是用PreparedStatement的方式，在Mysql数据库端抓出来的sql语句也不是以变量id=?的形式出现的，而是实际的数值。后来在网上看到在连接字符串上加上useServerPrepStmts=true可以实现真正的预编译

        String url="jdbc:mysql://172.16.2.7:3306/testdb";
        url=url+"?useServerPrepStmts=true";

加上这段后可以在数据库端可以看到明确的结果：

mysql> show global status like 'Com_stmt_prepare';
+------------------+-------+
| Variable_name | Value |
+------------------+-------+
| Com_stmt_prepare | 11 |
+------------------+-------+
1 row in set (0.00 sec)

但是实际的运行结果和上面几乎一样，性能上也没有任何的提升。由此可以推断出Mysql由于缺少类似于Oracle的library cache的部件，因此采用预编译方式执行sql是没有性能上的提升的。