【YashanDB知识库】从PostgreSQL迁移到YashanDB如何进行数据行数比对

本文内容来自YashanDB官网，原文内容请见https://www.yashandb.com/newsinfo/7863054.html?templateId=1718516

SQL语句的性能数据可以从视图v$sql中获取，针对可能有性能问题的SQL语句，可以根据该语句对应的plan_hash_value从视图v$sql_plan获取到该语句的执行计划。

创建UDF - REPEAT，后面SQL中有用到，格式化输出用

CREATE OR REPLACE FUNCTION repeat(p_str VARCHAR, p_repeat INT) RETURN VARCHAR

IS v_str VARCHAR(1000);

BEGIN

  v_str := '';

  IF p_repeat > 0 THEN

    FOR i in 1..p_repeat

    LOOP

      v_str := v_str || p_str;

    END LOOP;

  END IF;

  RETURN v_str;

END;

/

        

          
        
        
        

          

          AI 代码解读
        
      
      

根据plan_hash_value获取SQL文本（以2609204012为例）

SELECT sql_fulltext FROM v$sql WHERE plan_hash_value = 2609204012;

        

          
        
        
        

          

          AI 代码解读
        
      
      

SQL文本输出示例：

SELECT count(*) AS low_stock FROM (    SELECT s_w_id, s_i_id, s_quantity         FROM bmsql_stock s         WHERE s_w_id = ? AND s_quantity < ? AND s_i_id IN (            SELECT ol_i_id                 FROM bmsql_district                 JOIN bmsql_order_line ON ol_w_id = d_w_id                  AND ol_d_id = d_id                  AND ol_o_id >= d_next_o_id - 20                  AND ol_o_id < d_next_o_id                 WHERE d_w_id = ? AND d_id = ?         )     )

        

          
        
        
        

          

          AI 代码解读
        
      
      

获取内存中SQL语句的执行计划

SELECT id, repeat('||', depth) || operation as operation, object_name FROM v$sql_plan WHERE plan_hash_value = 2609204012 AND id IS NOT NULL ORDER BY id;

        

          
        
        
        

          

          AI 代码解读
        
      
      

输出示例：