Question

我的问题很简单。我试图将UPDATE语句转换为MERGE语句以获得更好的性能，因为更新计数可能在10K到1M之间。这是我的更新代码：

UPDATE table_1 t1
   SET t1.col_1 =
          (SELECT t2.name
             FROM table_2 t2
            WHERE t2.fn = t1.col_1),
       t1.Col_3 =
          (SELECT t2.name
             FROM table_2 t2
            WHERE t2.fn = t1.Col_3),
       t1.Col_5 =
          (SELECT t2.name
             FROM table_2 t2
            WHERE t2.fn = t1.Col_5);

我尝试换成MERGE：

MERGE INTO Table_1
     USING (SELECT NAME FROM table_2) A
        ON (1 = 1)
WHEN MATCHED
THEN
   UPDATE SET Table_1.col_1 = A.NAME,
              Table_1.col_3 = A.NAME,
              Table_1.col_5 = A.NAME;

上面的代码抛出错误，请帮我解决问题。提前谢谢。

Answer 1

使用

WHEN NOT MATCHED THEN
    INSERT (table_1.col1)
    VALUES (a.name);

试图将UPDATE转换为MERGE

1 个答案: