如何有效地从 INNER JOINed 表中选择聚合 SUM 值？

Question

我有一个带有状态列的子表。当我从父表中选择列时，我还想报告基于子表中所有相关行的计算状态。状态的计算非常简单：如果相关子表行中的所有状态都为 0，则查询中的状态将被设置为 0。否则，状态将被设置为非零值（并不重要）实际值为）。

我的查询实现（如下所示）在相关子行上使用 SUM()。这种方法效果很好。但是，由于 SUM()，我需要对从其他两个表中选择的所有列进行 GROUP BY。

有没有更好的方法来做到这一点？按这么多列分组时我应该担心性能吗？

顺便说一句，示例中省略了键、索引和关系。

示例表和存储过程：

CREATE DATABASE Test
GO

USE [Test]
GO

CREATE TABLE [dbo].[Customer]
(
    [CustomerId] [bigint] IDENTITY(1,1) NOT NULL,
    [Name] [nvarchar](128) NOT NULL,
    [PhoneNumber] [varchar](50) NOT NULL
)
GO

CREATE TABLE [dbo].[Order]
(
    [OrderId] [bigint] IDENTITY(1,1) NOT NULL,
    [CustomerId] [bigint] NOT NULL,
    [Description1] [nvarchar](256) NOT NULL,
    [Description2] [nvarchar](26) NOT NULL,
    [Description3] [nvarchar](256) NOT NULL,
    [Date] [datetime] NOT NULL,
    [Amount] [decimal](9, 2) NOT NULL,
    [RecommendedBy] [nvarchar](50) NOT NULL
)
GO

CREATE TABLE [dbo].OrderItem
(
    [OrderId] [bigint] NOT NULL,
    [Description] [nvarchar](50) NOT NULL,
    [Status] [tinyint] NOT NULL
) 
GO

CREATE PROCEDURE [dbo].[uspSelectInvoiceSubmissionRecords]
    @CustomerId bigint
AS
BEGIN
    SET NOCOUNT ON;

    SELECT
        c.Name
        , c.PhoneNumber
        , o.OrderId
        , o.Description1
        , o.Description2
        , o.Description3
        , o.[Date]
        , o.Amount
        , o.[RecommendedBy]
        , SUM(i.Status) AS [ItemStatus]
    FROM [Order] o
    INNER JOIN OrderItem i ON i.OrderId = o.OrderId
    INNER JOIN [Customer] c ON c.CustomerId = o.CustomerId
    WHERE o.CustomerId = @CustomerId
    GROUP BY c.Name, c.PhoneNumber, o.OrderId, o.Description1, o.Description2, o.Description3, o.Date, o.Amount, o.RecommendedBy;
END

Answer 1

您可能会发现使用相关子查询可以提供更好的性能：

SELECT c.Name, c.PhoneNumber, o.OrderId, o.Description1, o.Description2, o.Description3,
       o.[Date], o.Amount, o.[RecommendedBy]
       (SELECT SUM(i.Status)
        FROM OrderItem i
        WHERE i.OrderId = o.OrderId
       ) AS [ItemStatus]
FROM [Order] o JOIN
     [Customer] c
     ON c.CustomerId = o.CustomerId     
WHERE o.CustomerId = @CustomerId;

对于如此少量的数据（单个客户），这可能对性能影响不大，但查询更容易维护。