Oracle vs PostgreSQL Develop(16) - Prepared Statement
对于除绑定变量外其余相同的SQL语句,PostgreSQL提供了Prepared Statement用于缓存Plan,以达到Oracle中cursor_sharing=force的目的.
PSQL
通过prepare语句,可为SQL生成Prepared Statement,减少Plan的时间
[local]:5432 pg12@testdb=# explain (analyze,verbose) select * from t_prewarm where id = 1; QUERY PLAN -------------------------------------------------------------------------------------------- Index Scan using idx_t_prewarm_id on public.t_prewarm (cost=0.42..8.44 rows=1 width=13) (actual time=0.125..0.127 rows=1 loops=1) Output: id, c1 Index Cond: (t_prewarm.id = 1) Planning Time: 0.613 ms Execution Time: 0.181 ms(5 rows)Time: 2.021 ms[local]:5432 pg12@testdb=# explain (analyze,verbose) select * from t_prewarm where id = 1; QUERY PLAN -------------------------------------------------------------------------------------------- Index Scan using idx_t_prewarm_id on public.t_prewarm (cost=0.42..8.44 rows=1 width=13) (actual time=0.184..0.193 rows=1 loops=1) Output: id, c1 Index Cond: (t_prewarm.id = 1) Planning Time: 0.520 ms Execution Time: 0.276 ms(5 rows)
不使用prepare,可看到每次的Planning时间比Execution时间还要长
[local]:5432 pg12@testdb=# prepare p(int) as select * from t_prewarm where id=$1;PREPARETime: 1.000 ms[local]:5432 pg12@testdb=# explain (analyze,verbose) execute p(2); QUERY PLAN -------------------------------------------------------------------------------------------- Index Scan using idx_t_prewarm_id on public.t_prewarm (cost=0.42..8.44 rows=1 width=13) (actual time=0.037..0.039 rows=1 loops=1) Output: id, c1 Index Cond: (t_prewarm.id = 2) Planning Time: 0.323 ms Execution Time: 0.076 ms(5 rows)Time: 1.223 ms[local]:5432 pg12@testdb=# explain (analyze,verbose) execute p(3); QUERY PLAN ------------------------------------------------------------------------------------------------------------------------------------ Index Scan using idx_t_prewarm_id on public.t_prewarm (cost=0.42..8.44 rows=1 width=13) (actual time=0.077..0.081 rows=1 loops=1) Output: id, c1 Index Cond: (t_prewarm.id = $1) Planning Time: 0.042 ms Execution Time: 0.174 ms(5 rows)Time: 1.711 ms[local]:5432 pg12@testdb=# explain (analyze,verbose) execute p(4); QUERY PLAN ------------------------------------------------------------------------------------------------------------------------------------ Index Scan using idx_t_prewarm_id on public.t_prewarm (cost=0.42..8.44 rows=1 width=13) (actual time=0.042..0.044 rows=1 loops=1) Output: id, c1 Index Cond: (t_prewarm.id = $1) Planning Time: 0.019 ms Execution Time: 0.084 ms(5 rows)
使用prepare,可看到Planning时间明显降低
JDBC Driver
下面是测试代码
/* * TestPlanCache * * Copyright (C) 2004-2016, Denis Lussier * Copyright (C) 2016, Jan Wieck * */package testPG;import java.sql.Connection;import java.sql.DriverManager;import java.sql.PreparedStatement;import java.sql.ResultSet;import java.sql.SQLException;import java.sql.Statement;import java.util.Properties;public class TestPGPlanCache { public static void main(String[] args) { Connection conn = null; Statement stmt = null; String rLine = null; StringBuffer sql = new StringBuffer(); try { Properties ini = new Properties(); // ini.load(new FileInputStream(System.getProperty("prop"))); // Register jdbcDriver Class.forName("org.postgresql.Driver"); // make connection conn = DriverManager.getConnection("jdbc:postgresql://192.168.26.28:5432/testdb", "pg12", "pg12"); conn.setAutoCommit(true); PreparedStatement pstmt = conn.prepareStatement("SELECT * from t_prewarm where id = ?"); // cast to the pg extension interface org.postgresql.PGStatement pgstmt = pstmt.unwrap(org.postgresql.PGStatement.class); // on the third execution start using server side statements // pgstmt.setPrepareThreshold(3); for (int i = 1; i <= 10; i++) { pstmt.setInt(1, i); boolean usingServerPrepare = pgstmt.isUseServerPrepare(); ResultSet rs = pstmt.executeQuery(); rs.next(); System.out.println( "Execution: " + i + ", Used server side: " + usingServerPrepare + ", Result: " + rs.getInt(1)); rs.close(); } pstmt.close(); } catch (SQLException se) { System.out.println(se.getMessage()); } catch (Exception e) { e.printStackTrace(); // exit Cleanly } finally { try { if (conn != null) conn.close(); } catch (SQLException se) { se.printStackTrace(); } // end finally } // end try } // end main} // end ExecJDBC Class
输出为
Execution: 1, Used server side: false, Result: 1Execution: 2, Used server side: false, Result: 2Execution: 3, Used server side: false, Result: 3Execution: 4, Used server side: false, Result: 4Execution: 5, Used server side: true, Result: 5Execution: 6, Used server side: true, Result: 6Execution: 7, Used server side: true, Result: 7Execution: 8, Used server side: true, Result: 8Execution: 9, Used server side: true, Result: 9Execution: 10, Used server side: true, Result: 10
5次后开始使用服务器端的Prepared Statement.
参考资料
Server Prepared Statements
声明:本站所有文章资源内容,如无特殊说明或标注,均为采集网络资源。如若本站内容侵犯了原著者的合法权益,可联系本站删除。