SQLServer 分区

 

-- 进行演示操作前, 先备份, 以便可以在演示完成后, 恢复到原始状态
USE master
-- 备份
BACKUP DATABASE AdventureWorks
    TO DISK = 'AdventureWorks.bak'
WITH FORMAT

---- 恢复
--RESTORE DATABASE AdventureWorks
--    FROM DISK = 'AdventureWorks.bak'
--    WITH REPLACE
GO

--=========================================
-- 转换为分区表
--=========================================
USE AdventureWorks
GO

-- 1. 创建分区函数
--    a. 适用于存储历史存档记录的分区表的分区函数
DECLARE @dt datetime
SET @dt = '20020101'
CREATE PARTITION FUNCTION PF_HistoryArchive(datetime)
AS RANGE RIGHT
FOR VALUES(
    @dt,
    DATEADD(Year, 1, @dt))

--    b. 适用于存储历史记录的分区表的分区函数
--DECLARE @dt datetime
SET @dt = '20030901'
CREATE PARTITION FUNCTION PF_History(datetime)
AS RANGE RIGHT
FOR VALUES(
    @dt,
    DATEADD(Month, 1, @dt), DATEADD(Month, 2, @dt), DATEADD(Month, 3, @dt),
    DATEADD(Month, 4, @dt), DATEADD(Month, 5, @dt), DATEADD(Month, 6, @dt),
    DATEADD(Month, 7, @dt), DATEADD(Month, 8, @dt), DATEADD(Month, 9, @dt),
    DATEADD(Month, 10, @dt), DATEADD(Month, 11, @dt), DATEADD(Month, 12, @dt))
GO

-- 2. 创建分区架构
--    a. 适用于存储历史存档记录的分区表的分区架构
CREATE PARTITION SCHEME PS_HistoryArchive
AS PARTITION PF_HistoryArchive
TO([PRIMARY], [PRIMARY], [PRIMARY])

--    b. 适用于存储历史记录的分区表的分区架构
CREATE PARTITION SCHEME PS_History
AS PARTITION PF_History
TO([PRIMARY], [PRIMARY],
    [PRIMARY], [PRIMARY], [PRIMARY],
    [PRIMARY], [PRIMARY], [PRIMARY],
    [PRIMARY], [PRIMARY], [PRIMARY],
    [PRIMARY], [PRIMARY], [PRIMARY])
GO

-- 3. 删除索引
--    a. 删除存储历史存档记录的表中的索引
DROP INDEX Production.TransactionHistoryArchive.IX_TransactionHistoryArchive_ProductID
DROP INDEX Production.TransactionHistoryArchive.IX_TransactionHistoryArchive_ReferenceOrderID_ReferenceOrderLineID

--    b. 删除存储历史记录的表中的索引
DROP INDEX Production.TransactionHistory.IX_TransactionHistory_ProductID
DROP INDEX Production.TransactionHistory.IX_TransactionHistory_ReferenceOrderID_ReferenceOrderLineID
GO

-- 4. 转换为分区表
--    a. 将存储历史存档记录的表转换为分区表
ALTER TABLE Production.TransactionHistoryArchive
    DROP CONSTRAINT PK_TransactionHistoryArchive_TransactionID
    WITH(
        MOVE TO PS_HistoryArchive(TransactionDate))

--    b.将存储历史记录的表转换为分区表
ALTER TABLE Production.TransactionHistory
    DROP CONSTRAINT PK_TransactionHistory_TransactionID
    WITH(
        MOVE TO PS_History(TransactionDate))
GO

-- 5. 恢复主键
--    a. 恢复存储历史存档记录的分区表的主键
ALTER TABLE Production.TransactionHistoryArchive
    ADD CONSTRAINT PK_TransactionHistoryArchive_TransactionID
        PRIMARY KEY CLUSTERED(
            TransactionID,
            TransactionDate)

--    b. 恢复存储历史记录的分区表的主键
ALTER TABLE Production.TransactionHistory
    ADD CONSTRAINT PK_TransactionHistory_TransactionID
        PRIMARY KEY CLUSTERED(
            TransactionID,
            TransactionDate)
GO

-- 6. 恢复索引
--    a. 恢复存储历史存档记录的分区表的索引
CREATE INDEX IX_TransactionHistoryArchive_ProductID
    ON Production.TransactionHistoryArchive(
        ProductID)

CREATE INDEX IX_TransactionHistoryArchive_ReferenceOrderID_ReferenceOrderLineID
    ON Production.TransactionHistoryArchive(
        ReferenceOrderID,
        ReferenceOrderLineID)

--    b. 恢复存储历史记录的分区表的索引
CREATE INDEX IX_TransactionHistory_ProductID
    ON Production.TransactionHistory(
        ProductID)

CREATE INDEX IX_TransactionHistory_ReferenceOrderID_ReferenceOrderLineID
    ON Production.TransactionHistory(
        ReferenceOrderID,
        ReferenceOrderLineID)
GO

-- 7. 查看分区表的相关信息
SELECT
    SchemaName = S.name,
    TableName = TB.name,
    PartitionScheme = PS.name,
    PartitionFunction = PF.name,
    PartitionFunctionRangeType = CASE
WHEN boundary_value_on_right = 0 THEN 'LEFT'
ELSE 'RIGHT' END,
    PartitionFunctionFanout = PF.fanout,
    SchemaID = S.schema_id,
    ObjectID = TB.object_id,
    PartitionSchemeID = PS.data_space_id,
    PartitionFunctionID = PS.function_id
FROM sys.schemas S
    INNER JOIN sys.tables TB
        ON S.schema_id = TB.schema_id
    INNER JOIN sys.indexes IDX
        on TB.object_id = IDX.object_id
AND IDX.index_id < 2
INNER JOIN sys.partition_schemes PS
        ON PS.data_space_id = IDX.data_space_id
    INNER JOIN sys.partition_functions PF
        ON PS.function_id = PF.function_id
GO


--=========================================
-- 移动分区表数据
--=========================================
-- 1. 为存储历史存档记录的分区表增加分区, 并接受从历史记录分区表移动过来的数据
--    a. 修改分区架构, 增加用以接受新分区的文件组
ALTER PARTITION SCHEME PS_HistoryArchive
NEXT USED [PRIMARY]

--    b. 修改分区函数, 增加分区用以接受从历史记录分区表移动过来的数据
DECLARE @dt datetime
SET @dt = '20030901'
ALTER PARTITION FUNCTION PF_HistoryArchive()
SPLIT RANGE(@dt)

--    c. 将历史记录表中的过期数据移动到历史存档记录表中
ALTER TABLE Production.TransactionHistory
    SWITCH PARTITION 2
        TO Production.TransactionHistoryArchive PARTITION $PARTITION.PF_HistoryArchive(@dt)

--    d. 将接受到的数据与原来的分区合并
ALTER PARTITION FUNCTION PF_HistoryArchive()
MERGE RANGE(@dt)
GO

-- 2. 将存储历史记录的分区表中不包含数据的分区删除, 并增加新的分区以接受新数据
--    a. 合并不包含数据的分区
DECLARE @dt datetime
SET @dt = '20030901'
ALTER PARTITION FUNCTION PF_History()
MERGE RANGE(@dt)

--    b.  修改分区架构, 增加用以接受新分区的文件组
ALTER PARTITION SCHEME PS_History
NEXT USED [PRIMARY]

--    c. 修改分区函数, 增加分区用以接受新数据
SET @dt = '20041001'
ALTER PARTITION FUNCTION PF_History()
SPLIT RANGE(@dt)
GO


--=========================================
-- 清除历史存档记录中的过期数据
--=========================================
-- 1. 创建用于保存过期的历史存档数据的表
CREATE TABLE Production.TransactionHistoryArchive_2001_temp(
    TransactionID int NOT NULL,
    ProductID int NOT NULL,
    ReferenceOrderID int NOT NULL,
    ReferenceOrderLineID int NOT NULL
        DEFAULT ((0)),
    TransactionDate datetime NOT NULL
        DEFAULT (GETDATE()),
    TransactionType nchar(1) NOT NULL,
    Quantity int NOT NULL,
    ActualCost money NOT NULL,
    ModifiedDate datetime NOT NULL
        DEFAULT (GETDATE()),
    CONSTRAINT PK_TransactionHistoryArchive_2001_temp_TransactionID
        PRIMARY KEY CLUSTERED(
            TransactionID,
            TransactionDate)
)

-- 2. 将数据从历史存档记录分区表移动到第1步创建的表中
ALTER TABLE Production.TransactionHistoryArchive
    SWITCH PARTITION 1
        TO Production.TransactionHistoryArchive_2001_temp

-- 3. 删除不再包含数据的分区
DECLARE @dt datetime
SET @dt = '20020101'
ALTER PARTITION FUNCTION PF_HistoryArchive()
MERGE RANGE(@dt)

-- 4. 修改分区架构, 增加用以接受新分区的文件组
ALTER PARTITION SCHEME PS_HistoryArchive
NEXT USED [PRIMARY]

-- 5. 修改分区函数, 增加分区用以接受新数据
SET @dt = '20040101'
ALTER PARTITION FUNCTION PF_HistoryArchive()
SPLIT RANGE(@dt)


查询分区信息：

;WITH
TBINFO AS(
    SELECT
        SchemaName = S.name,
        TableName = TB.name,
        PartitionScheme = PS.name,
        PartitionFunction = PF.name,
        PartitionFunctionRangeType = CASE
                WHEN boundary_value_on_right = 0 THEN 'LEFT'
                ELSE 'RIGHT' END,
        PartitionFunctionFanout = PF.fanout,
        SchemaID = S.schema_id,
        ObjectID = TB.object_id,
        PartitionSchemeID = PS.data_space_id,
        PartitionFunctionID = PS.function_id
    FROM sys.schemas S
        INNER JOIN sys.tables TB
            ON S.schema_id = TB.schema_id
        INNER JOIN sys.indexes IDX
            on TB.object_id = IDX.object_id
                AND IDX.index_id < 2
        INNER JOIN sys.partition_schemes PS
            ON PS.data_space_id = IDX.data_space_id
        INNER JOIN sys.partition_functions PF
            ON PS.function_id = PF.function_id
),
PF1 AS(
    SELECT
        PFP.function_id, PFR.boundary_id, PFR.value,
        Type = CONVERT(sysname,
            CASE T.name
                WHEN 'numeric' THEN 'decimal'
                WHEN 'real' THEN 'float'
                ELSE T.name END
            + CASE
                WHEN T.name IN('decimal', 'numeric')
                    THEN QUOTENAME(RTRIM(PFP.precision)
                        + CASE WHEN PFP.scale > 0 THEN ',' + RTRIM(PFP.scale) ELSE '' END, '()')
                WHEN T.name IN('float', 'real')
                    THEN QUOTENAME(PFP.precision, '()')
                WHEN T.name LIKE 'n%char'
                    THEN QUOTENAME(PFP.max_length / 2, '()')
                WHEN T.name LIKE '%char' OR T.name LIKE '%binary'
                    THEN QUOTENAME(PFP.max_length, '()')
                ELSE '' END)
    FROM sys.partition_parameters PFP
        LEFT JOIN sys.partition_range_values PFR
            ON PFR.function_id = PFP.function_id
                AND PFR.parameter_id = PFP.parameter_id
        INNER JOIN sys.types T
            ON PFP.system_type_id = T.system_type_id
),
PF2 AS(
    SELECT * FROM PF1
    UNION ALL
    SELECT
        function_id, boundary_id = boundary_id - 1, value, type
    FROM PF1
    WHERE boundary_id = 1
),
PF AS(
    SELECT
        B.function_id, boundary_id = ISNULL(B.boundary_id + 1, 1),
        value = STUFF(
            CASE
                WHEN A.boundary_id IS NULL THEN ''
                ELSE ' AND [partition_column_name] ' + PF.LessThan + ' ' + CONVERT(varchar(max), A.value) END
            + CASE
                WHEN A.boundary_id = 1 THEN ''
                ELSE ' AND [partition_column_name] ' + PF.MoreThan + ' ' + CONVERT(varchar(max), B.value) END,
            1, 5, ''),
        B.Type
    FROM PF1 A        
        RIGHT JOIN PF2 B
            ON A.function_id = B.function_id
                AND (A.boundary_id - 1 = B.boundary_id
                    OR(A.boundary_id IS NULL AND B.boundary_id IS NULL))
        INNER JOIN(
            SELECT
                function_id,
                LessThan = CASE
                        WHEN boundary_value_on_right = 0 THEN '<='
                        ELSE '<' END,
                MoreThan = CASE
                        WHEN boundary_value_on_right = 0 THEN '>'
                        ELSE '>=' END
            FROM sys.partition_functions
        )PF
            ON B.function_id = PF.function_id
),
PS AS(
    SELECT
        DDS.partition_scheme_id, DDS.destination_id,
        FileGroupName = FG.name, IsReadOnly = FG.is_read_only
    FROM sys.destination_data_spaces DDS
        INNER JOIN sys.filegroups FG
            ON DDS.data_space_id = FG.data_space_id
),
PINFO AS(
    SELECT
        RowID = ROW_NUMBER() OVER(ORDER BY SchemaID, ObjectID, PS.destination_id),
        TB.SchemaName, TB.TableName,
        TB.PartitionScheme, PS.destination_id, PS.FileGroupName, PS.IsReadOnly,
        TB.PartitionFunction, TB.PartitionFunctionRangeType, TB.PartitionFunctionFanout,
        PF.boundary_id, PF.Type, PF.value
    FROM TBINFO TB
        INNER JOIN PS
            ON TB.PartitionSchemeID = PS.partition_scheme_id
        LEFT JOIN PF
            ON TB.PartitionFunctionID = PF.function_id
                AND PS.destination_id = PF.boundary_id
)
SELECT
    RowID,
    SchemaName = CASE destination_id
            WHEN 1 THEN SchemaName
            ELSE N'' END,
    TableName = CASE destination_id
            WHEN 1 THEN TableName
            ELSE N'' END,
    PartitionScheme = CASE destination_id
            WHEN 1 THEN PartitionScheme
            ELSE N'' END,
    destination_id, FileGroupName, IsReadOnly,
    PartitionFunction = CASE destination_id
            WHEN 1 THEN PartitionFunction
            ELSE N'' END,
    PartitionFunctionRangeType = CASE destination_id
            WHEN 1 THEN PartitionFunctionRangeType
            ELSE N'' END,
    PartitionFunctionFanout = CASE destination_id
            WHEN 1 THEN CONVERT(varchar(20), PartitionFunctionFanout)
            ELSE N'' END,
    boundary_id = ISNULL(CONVERT(varchar(20), boundary_id), ''),
    Type = ISNULL(Type, N''),
    value = CASE PartitionFunctionFanout
            WHEN 1 THEN '<ALL Data>'
            ELSE ISNULL(value, N'<NEXT USED>') END
FROM PINFO
ORDER BY RowID




--＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝
--drop database dbPartitionTest
　　--测试数据库
　　create database dbPartitionTest
　　go
　　use
　　dbPartitionTest
　　go
　　--增加分组
　　alter database dbPartitionTest ADD FILEGROUP P200801
　　alter database dbPartitionTest ADD FILEGROUP P200802
　　alter database dbPartitionTest ADD FILEGROUP P200803
　　go
　　--分区函数
　　CREATE PARTITION FUNCTION part_Year(datetime)
　　AS RANGE LEFT FOR VALUES
　　(
　　’20080131 23:59:59.997’,’20080229 23:59:59.997’,’20080331 23:59:59.997’
　　)
　　go
　　--增加文件组
　　ALTER DATABASE dbPartitionTest ADD FILE (NAME = N’P200801’,FILENAME = N’c:tb_P200801.ndf’,SIZE = 1MB,MAXSIZE = 500MB,FILEGROWTH = 1MB)TO FILEGROUP P200801
　　ALTER DATABASE dbPartitionTest ADD FILE (NAME = N’P200802’,FILENAME = N’c:tb_P200802.ndf’,SIZE = 1MB,MAXSIZE = 500MB,FILEGROWTH = 1MB)TO FILEGROUP P200802
　　ALTER DATABASE dbPartitionTest ADD FILE (NAME = N’P200803’,FILENAME = N’c:tb_P200803.ndf’,SIZE = 1MB,MAXSIZE = 500MB,FILEGROWTH = 1MB)TO FILEGROUP P200803
　　go
　　--分区架构
　　CREATE PARTITION SCHEME part_YearScheme　AS PARTITION part_Year　 TO (P200801,P200802,P200803,[PRIMARY])
　　go
　　CREATE TABLE [dbo].t_part
　　(name varchar(100) default newid(),date datetime NOT NULL)
　　ON part_YearScheme (date)
　　go
　　--添加测试数据,每天1条
　　declare @date datetime
　　set @date=’2007-12-31’
　　while @date<=’2008-04-0’
　　1 begin
　　insert into t_part(date)values(@date)
　　set @date=@date+1
　　end
　　go
　　--查询数据分布在哪些分区
　　select $partition.part_Year(date) as 分区编号,* from t_part　order by date
　　--查询数据库文件
　　go
　　sp_helpfile

===================================================