我有这张桌子。
我不知道为什么这张桌子太大了。
这是表结构。
CREATE TABLE [dbo].[ACI_HISCLI](
[TER_CODI] [varchar](13) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[HIS_NUME] [int] NOT NULL,
[HIS_FECH] [datetime] NOT NULL,
[HIS_ESTA] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[HIS_HORA] [datetime] NOT NULL,
[MED_CODI] [decimal](13, 0) NOT NULL,
[HIS_HOEN] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[HIS_HODC] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[HIS_MOTI] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_AODS] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_AODC] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_AOSS] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_AOSC] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_ODQ1] [decimal](4, 2) NULL,
[HIS_ODQ2] [int] NULL,
[HIS_ODQ3] [decimal](4, 2) NULL,
[HIS_ODQ4] [int] NULL,
[HIS_OSQ1] [decimal](4, 2) NULL,
[HIS_OSQ2] [int] NULL,
[HIS_OSQ3] [decimal](4, 2) NULL,
[HIS_OSQ4] [int] NULL,
[HIS_ODR1] [decimal](4, 2) NULL,
[HIS_ODR2] [decimal](4, 2) NULL,
[HIS_ODR3] [int] NULL,
[HIS_OSR1] [decimal](4, 2) NULL,
[HIS_OSR2] [decimal](4, 2) NULL,
[HIS_OSR3] [int] NULL,
[HIS_ODS1] [decimal](4, 2) NULL,
[HIS_ODS2] [decimal](4, 2) NULL,
[HIS_ODS3] [int] NULL,
[HIS_ODS4] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_OSS1] [decimal](4, 2) NULL,
[HIS_OSS2] [decimal](4, 2) NULL,
[HIS_OSS3] [int] NULL,
[HIS_OSS4] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_ADI1] [decimal](3, 2) NULL,
[HIS_ADI2] [varchar](2) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_ADI3] [int] NULL,
[HIS_TON1] [datetime] NULL,
[HIS_TOOD] [int] NULL,
[HIS_TOOS] [int] NULL,
[HIS_MOVI] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_BIOM] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_FOJO] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_OTRO] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[DIA_COD1] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[DIA_COD2] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[DIA_COD3] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_COND] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_OBSE] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_IMA1] [image] NULL,
[HIS_IMN1] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_IMA2] [image] NULL,
[HIS_IMN2] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_CONS] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_ORIL] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_ORIT] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_ORIP] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_INGS] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_INGA] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_INGC] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_INGE] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_INGL] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_TAIN] [varchar](7) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_PUIN] [int] NULL,
[HIS_TEIN] [decimal](6, 2) NULL,
[HIS_TAEG] [varchar](7) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_PUEG] [int] NULL,
[HIS_TEEG] [decimal](6, 2) NULL,
[HIS_OBHE] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[USU_CREA] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[HIS_ASPR] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[PRO_CONS] [int] NULL,
[MED_CODC] [decimal](13, 0) NULL,
[ANE_CODI] [decimal](13, 0) NULL,
[AYU_CODI] [decimal](13, 0) NULL,
[INS_CODI] [decimal](13, 0) NULL,
[HIS_DIPR] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DIPO] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DECI] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_GODV] [int] NULL,
[HIS_GODT] [varchar](200) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_GOIV] [int] NULL,
[HIS_GOIT] [varchar](200) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_BIFO] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_BIOT] [varchar](51) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_OBSD] [varchar](100) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_OBSS] [varchar](100) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_ODLA] [image] NULL,
[HIS_ODFR] [image] NULL,
[HIS_OILA] [image] NULL,
[HIS_OIFR] [image] NULL,
[HIS_ODSE] [image] NULL,
[HIS_OISE] [image] NULL,
[HIS_GOD1] [decimal](3, 2) NULL,
[HIS_GOD2] [decimal](3, 2) NULL,
[HIS_GOD3] [decimal](3, 2) NULL,
[HIS_GOD4] [decimal](3, 2) NULL,
[HIS_GOD5] [decimal](3, 2) NULL,
[HIS_GOD6] [decimal](3, 2) NULL,
[HIS_GOD7] [decimal](3, 2) NULL,
[HIS_GOD8] [decimal](3, 2) NULL,
[HIS_GOI1] [decimal](3, 2) NULL,
[HIS_GOI2] [decimal](3, 2) NULL,
[HIS_GOI3] [decimal](3, 2) NULL,
[HIS_GOI4] [decimal](3, 2) NULL,
[HIS_GOI5] [decimal](3, 2) NULL,
[HIS_GOI6] [decimal](3, 2) NULL,
[HIS_GOI7] [decimal](3, 2) NULL,
[HIS_GOI8] [decimal](3, 2) NULL,
[HIS_CONT] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_RESU] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_GLIC] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_HEMA] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_PAQU] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_CREA] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_ELEC] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_BIO1] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_BIO2] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_BIO3] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_BIO4] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_BIO5] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_OBEX] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_OBST] [varchar](150) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DIT1] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DIT2] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DIT3] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_TANE] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[ENT_CODI] [numeric](13, 0) NULL,
[HIS_DIDR] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DIDO] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DIPI] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DISA] [text] COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DIO1] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DIO2] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_DIO3] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[HIS_HOAN] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_NOAC] [varchar](120) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_TACO] [varchar](25) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_PACO] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_DACO] [varchar](70) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_IACO] [varchar](15) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_NORE] [varchar](120) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_TRES] [varchar](25) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_PRES] [varchar](50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_DRES] [varchar](70) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[TER_IRES] [varchar](15) COLLATE SQL_Latin1_General_CP1_CI_AS NULL,
[USU_AUDI] [varchar](10) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[FEC_AUDI] [datetime] NOT NULL,
[TIP_AUDI] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
[MOD_ORIG] [char](1) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL
) ON [PRIMARY] TEXTIMAGE_ON [PRIMARY]
sp_spaceued ACI_HISCLI
这是结果： 类型图像的字段，几乎为空。
这个表有很多用户的交易，但只有大小为60 GB的文本蚂蚁!!
我不知道为什么这张桌子太大了！
如何优化此表或缩小此表。
我试着用： 并重建索引
感谢。name: ACI_HISCLI
rows: 270719
reserved: 64521288 KB
data: 64377992 KB
index_size: 152 Kb
unused: 143144 KB
DBCC CLEANTABLE
答案 0 :(得分：3)
所有这些TEXT和IMAGE字段每个占用16个字节，因为它们是数据实际所在位置的指针（当数据存在时）。但我几乎可以肯定，就像固定宽度数据类型（INT，BIGINT，DATETIME等）一样，即使是NULL，它们也占用了那个空间。
此外，该表确实有太多字段无法管理。您应该考虑将其分解为两个（或更多）具有相同PK，FKed并且具有1对1关系的表。未使用的字段通常应移动到辅助表。
编辑：
使用的数据空间也受到聚集索引上的FILLFACTOR设置的影响。将其设置得太低，尤其是如此大的行大小，将导致很少的行适合8k数据页。
此外，有时删除行时不会释放空间。在Clustered Index上执行索引REBUILD将释放该空间（如果大小下降了很多，您将知道在执行REBUILD后是否有未释放的空间。）
EDIT2：
任何使用SQL Server 2005及更高版本的人都应该使用旧的blob类型（现在已弃用）将字段转换为更新的等价物：