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.
为察布查尔锡伯等地区用户提供了全套网页设计制作服务，及察布查尔锡伯网站建设行业解决方案。主营业务为网站设计制作、成都网站建设、察布查尔锡伯网站设计，以传统方式定制建设网站，并提供域名空间备案等一条龙服务，秉承以专业、用心的态度为用户提供真诚的服务。我们深信只要达到每一位用户的要求，就会得到认可，从而选择与我们长期合作。这样，我们也可以走得更远！
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) (a
ctual 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) (a
ctual 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;
PREPARE
Time: 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) (a
ctual 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) (a
ctual 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) (a
ctual 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: 1
Execution: 2, Used server side: false, Result: 2
Execution: 3, Used server side: false, Result: 3
Execution: 4, Used server side: false, Result: 4
Execution: 5, Used server side: true, Result: 5
Execution: 6, Used server side: true, Result: 6
Execution: 7, Used server side: true, Result: 7
Execution: 8, Used server side: true, Result: 8
Execution: 9, Used server side: true, Result: 9
Execution: 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
