SQL server 2005的表分区

来源：互联网发布：远程控制桌面软件编辑：程序博客网时间：2024/05/29 17:33

下面来说下，在SQL SERVER 2005的表分区里，如何对已经存在的有数据的表进行分区，其实道理和之前在http://www.cnblogs.com/jackyrong/archive/2006/11/13/559354.html说到一样，只不过交换下顺序而已，下面依然用例子说明：
依然在c盘的data2目录下建立4个文件夹，用来做4个文件组，然后建立数据库

use masterIF  EXISTS (SELECT name FROM sys.databases WHERE name = N'Data Partition DB3')DROP DATABASE [Data Partition DB3]GOCREATE DATABASE [Data Partition DB3]ON PRIMARY(NAME='Data Partition DB Primary FG3',FILENAME='C:\Data2\Primary\Data Partition DB Primary FG3.mdf',SIZE=5,MAXSIZE=500,FILEGROWTH=1 ),FILEGROUP [Data Partition DB3 FG1](NAME = 'Data Partition DB3 FG1',FILENAME ='C:\Data2\FG1\Data Partition DB3 FG1.ndf',SIZE = 5MB,MAXSIZE=500,FILEGROWTH=1 ),FILEGROUP [Data Partition DB3 FG2](NAME = 'Data Partition DB3 FG2',FILENAME ='C:\Data2\FG2\Data Partition DB3 FG2.ndf',SIZE = 5MB,MAXSIZE=500,FILEGROWTH=1 ),FILEGROUP [Data Partition DB3 FG3](NAME = 'Data Partition DB3 FG3',FILENAME ='C:\Data2\FG3\Data Partition DB3 FG3.ndf',SIZE = 5MB,MAXSIZE=500,FILEGROWTH=1 ),FILEGROUP [Data Partition DB3 FG4](NAME = 'Data Partition DB3 FG4',FILENAME ='C:\Data2\FG4\Data Partition DB3 FG4.ndf',SIZE = 5MB,MAXSIZE=500,FILEGROWTH=1 )然后建立一个数据表：
USE [Data Partition DB3]goCREATE TABLE MyTable(ID INT NOT NULL,Date DATETIME,Cost money ) on [primary]并建立一个索引
USE [Data Partition DB3]goCREATE UNIQUE CLUSTERED INDEX MyTable_IXCON MyTable(ID) on [PRIMARY]接下来往表里增加数据
USE [Data Partition DB3]godeclare @count intset @count =-25while @count <=100begininsert into MyTable select @count,getdate(),100.00set @count=@count+1endset @count =101while @count <=200begininsert into MyTable select @count,getdate(),200.00set @count=@count+1endset @count =201while @count <=300begininsert into MyTable select @count,getdate(),300.00set @count=@count+1endset @count =301while @count <=400begininsert into MyTable select @count,getdate(),400.00set @count=@count+1endset @count =401while @count <=800begininsert into MyTable select @count,getdate(),500.00set @count=@count+1end此时查询一下，可以看到数据都在一个表里
select * from sys.partitions where object_name(object_id)='MyTable'
 
 我们再建立表分区函数
use [Data Partition DB3]GOCREATE PARTITION FUNCTION [Data Partition Range](int)AS RANGE LEFT FOR VALUES (100,200,300)这里表明分区的原则是四个分区，从负数到100，101-200，201-300，大于300
当然，如果用right for values的话，就是从负数到99，100到199，200-299，和大于300
 
最后，把表分区函数应用到文件组里
USE [Data Partition DB3]goCREATE PARTITION SCHEME [Data Partition Scheme]AS PARTITION [Data Partition Range]TO ([Data Partition DB3 FG1], [Data Partition DB3 FG2], [Data Partition DB3 FG3],[Data Partition DB3 FG4]);把原来建立好的表，移动到这个表分区里
Drop index MyTable_IXC on MyTable with (Move To [Data Partition Scheme] (ID) )
 
最后看一看
select * from sys.partitions where object_name(object_id)='MyTable'
 
可以看到，原来的表的数据被正确分拆到四个文件组里去了，实现了表分区