查询性能 - select 仅当存在时

Question

我正在努力提高查询效率。这是逐行检查，很多时候由于没有匹配的记录，查询不会到达更新点。我 (EDIT) ~~know~~ 认为最好的方法是先检查是否存在匹配项，然后再执行查询。以下面的真实案例为例，我试图将性能成本降至最低：

/* LOOPING FROM MIN(ID) TILL MAX(ID) */
IF EXISTS(SELECT *
            FROM InvoicesHC IHC
                INNER JOIN ContractSpecifications CS ON IHC.ContractSpecificationID = CS.ContractSpecificationID
                INNER JOIN InstalledProductGroups IPG ON IPG.InstalledProductGroupID = CS.InstalledProductGroupID
                INNER JOIN ProductGroups PG ON PG.ProductGroupID = IPG.ProductGroupID
                INNER JOIN ContractFinYears CFY ON IHC.ContractFinYearID = CFY.ContractFinYearID
            WHERE PG.ProductGroupID = 2 AND IHC.ContractNr = @CCODE AND CFY.ContractFinYear = @CYEAR)
    BEGIN
            UPDATE InvoicesHC SET InvoicePrice = @IGAS, PONumber = @POCO
            FROM InvoicesHC IHC
                INNER JOIN ContractSpecifications CS ON IHC.ContractSpecificationID = CS.ContractSpecificationID
                INNER JOIN InstalledProductGroups IPG ON IPG.InstalledProductGroupID = CS.InstalledProductGroupID
                INNER JOIN ProductGroups PG ON PG.ProductGroupID = IPG.ProductGroupID
                INNER JOIN ContractFinYears CFY ON IHC.ContractFinYearID = CFY.ContractFinYearID
            WHERE PG.ProductGroupID = 2 AND IHC.ContractNr = @CCODE AND CFY.ContractFinYear = @CYEAR
    END

所以我首先检查它是否存在，然后我实际执行它。这应该使查询运行更快，但我不喜欢的事实是我实际上再次复制粘贴相同的查询 - 并且在 T-SQL 中进行 6 次这样的检查，它变得非常大 - 比我认为实际应该看起来的要大。我关心的另一个问题是性能：我已经将表连接在一起（现在只有 4 个内部连接，但如果有更多会怎样）来查找记录是否匹配，然后再次将它们连接在一起以执行更新查询。在我使用的小型数据集中，这没什么大不了的。但是，如果这种逐行检查应该执行数百万次怎么办？

我相信应该有更有效的方法来创建此语句？有人有什么想法吗？

谢谢。

Answer 1

but what I dislike is the fact that I actually copy-paste the same query over again - and within a T-SQL with 6 times a check like this, it becomes very big

我认为您应该再次使用 Merge.Using 合并减少那些重复查询，并且 again.And 性能方面至少会比以前的代码更好。

这只是演示。

MERGE InvoicesHC AS TRG
USING (
SELECT InvoicesHC
            FROM InvoicesHC IHC
                INNER JOIN ContractSpecifications CS ON IHC.ContractSpecificationID = CS.ContractSpecificationID
                INNER JOIN InstalledProductGroups IPG ON IPG.InstalledProductGroupID = CS.InstalledProductGroupID
                INNER JOIN ProductGroups PG ON PG.ProductGroupID = IPG.ProductGroupID
                INNER JOIN ContractFinYears CFY ON IHC.ContractFinYearID = CFY.ContractFinYearID
            WHERE PG.ProductGroupID = 2 AND IHC.ContractNr = @CCODE AND CFY.ContractFinYear = @CYEAR
)as SRC
on trg.ContractNr=src.ContractNr
When matched then
    update 
        SET InvoicePrice = @IGAS, PONumber = @POCO;

Answer 2

这个应该没问题：

        UPDATE InvoicesHC SET InvoicePrice = @IGAS, PONumber = @POCO
        FROM InvoicesHC IHC
            INNER JOIN ContractSpecifications CS ON IHC.ContractSpecificationID = CS.ContractSpecificationID
            INNER JOIN InstalledProductGroups IPG ON IPG.InstalledProductGroupID = CS.InstalledProductGroupID
            INNER JOIN ProductGroups PG ON PG.ProductGroupID = IPG.ProductGroupID
            INNER JOIN ContractFinYears CFY ON IHC.ContractFinYearID = CFY.ContractFinYearID
        WHERE PG.ProductGroupID = 2 AND IHC.ContractNr = @CCODE AND CFY.ContractFinYear = @CYEAR

查询性能 - select 仅当存在时

Query performance - select only if exists

sql

sql-server

query-performance