Question

我是SQL Server的新手，最后让我的代码运行，但速度非常慢。我的数据库有点偏大（1.22千米行，3列），我有一些while循环，我知道这是不理想但我无法找到解决方法。

任何帮助都将非常感谢!!!

declare @b float
declare @c float
declare @dateloopfora datetime
declare @dateloopforc datetime

BEGIN 
    SET @MYCURSOR = CURSOR FOR  
        SELECT DISTINCT [STOCKS] FROM [tsxvPrices].[dbo].[3coldata] 

    OPEN @MYCURSOR

    FETCH NEXT FROM @MYCURSOR INTO @STOCK

    set @end = '12/30/2012'

    WHILE @@FETCH_STATUS = 0
    BEGIN 
        print ('Stockname restart: ' + @stock)
        SET @dateloop = '01/01/2012'

        WHILE (@dateloop <= @end)
        BEGIN
            set @datedelta = 0
            SET @dateloop = dateadd(day,1,@dateloop)
            set @b = (SELECT [val] FROM [tsxvPrices].[dbo].[3coldata] WHERE( [dates] = @dateloop and [stocks] = @stock))

            WHILE (@datedelta < 4) 
            BEGIN
                set @datedelta = @datedelta + 1

                if (@b is null)
                begin
                    SET @dateloop = dateadd(day,1,@dateloop)
                    set @b = (SELECT [val] FROM [tsxvPrices].[dbo].[3coldata] WHERE( [dates] = @dateloop and [stocks] = @STOCK))
                    /*print ('b= ' + cast( @b as varchar(10)) + ' dateadd1=' + cast( @datedelta as varchar(10))) */
                end
            END

        SET @datedelta = 0
        set @a = (SELECT [val] FROM [tsxvPrices].[dbo].[3coldata] WHERE( [dates] = dateadd(day,-1,@dateloop) and [stocks] = @stock))
        set @c = (SELECT [val] FROM [tsxvPrices].[dbo].[3coldata] WHERE( [dates] = dateadd(day,+96,@dateloop) and [stocks] = @stock))

        if @b/@a < 0.8 
        begin
            WHILE (@datedelta < 4)
            BEGIN
                set @datedelta = @datedelta + 1

                if (@a is null)
                begin
                    SET @dateloopfora = dateadd(day,-@datedelta,@dateloop)
                    set @a = (SELECT [val] FROM [tsxvPrices].[dbo].[3coldata] WHERE( [dates] = @dateloopfora and [stocks] = @STOCK))
                    /*print ('b= ' + cast( @b as varchar(10)) + ' dateadd1=' + cast( @datedelta as varchar(10))) */
                end

                if (@c is null)
                begin
                    SET @dateloopforc = dateadd(day,96+@datedelta,@dateloop)
                    set @c = (SELECT [val] FROM [tsxvPrices].[dbo].[3coldata] WHERE( [dates] = @dateloopforc and [stocks] = @STOCK))
                    /*print ('b= ' + cast( @b as varchar(10)) + ' dateadd1=' + cast( @datedelta as varchar(10))) */
                end
            END

            /*print ('Stockname: ' + @stock)
            print @dateloop
            print('daily')
            print @b/@a 
            print ('quarterly')*/
            print @c/@b
            /*print ('======================')*/
        end
    END

    FETCH NEXT FROM @MYCURSOR INTO @STOCK
END

CLOSE @MYCURSOR
DEALLOCATE @MYCURSOR

Answer 1

我可以立即看到一个简单的问题，你正在使用Cursor。这可能是处理SQL的最糟糕的方法之一，它非常慢，并且它迫使SQL引擎效率非常低。

你是SQL的新手，所以我会放弃技术解释，但请看看其他StackOverflow问题：

Why is it considered bad practice to use cursors in SQL Server?

我或许可以提供进一步的建议，您使用的是什么版本的SQL？

获取SQL服务器以更快地查询运行

1 个答案: