The FILESTREAM feature of SQL Server 2008, which allows
storage of and efficient access to BLOB data using a combination of SQL Server
2008 and the NTFS file system. It covers choices for BLOB storage, configuring
Windows and SQL Server for using FILESTREAM data, considerations for combining
FILESTREAM with other features, and implementation details such as partitioning
and performance.
1. Enable FILESTREAM support
To enable FILESTREAM support, run the following SQL:
On SQL Server 2008 Express:
USE master;
EXEC sys.sp_configure N'filestream access level',
N'2'
GO
RECONFIGURE WITH OVERRIDE
GO
On SQL Server 2008 Enterprise :
exec [sp_filestream_configure] @enable_level = 3;
This command enables or disables the support according to
the parameter. The value 3 means that FILESTREAM will be enabled for
Transact-SQL, local file system access, and remote file system access.
2. Create a database with a File Group that contains FILESTREAM
To create a database instance with a file group that
contains FILESTREAM, run the following sql command. This creates the database
with 3 file groups, but only one of them contains FILESTREAM as you can see in
the command.
CREATE DATABASE CSAFMS
ON
PRIMARY (
NAME = CSAFMS_Primary,
FILENAME = 'D:\Databases\CSAFMS\CSAFMS.mdf'),
FILEGROUP FileStreamGroup CONTAINS FILESTREAM (
NAME =
FileManagement_FileGroup,
FILENAME = 'D:\Databases\CSAFMS\CSAFMS')
LOG ON ( NAME = FileManagement_Log,
FILENAME = 'D:\Databases\CSAFMS\CSAFMSLog.ldf')
GO
This command will create the following directories
structure:
In the Data Directory, I could find the files specified
above:
and in the CSAFMS directory, I could find the following
content:
3. Create a Table with FILESTREAM
Note: FILESTREAM is not a
type of a column, but it is a property you put on a varbinary(max) column.
So, in order to create a table with a varbinary(max) column
that will be used for FILESTREAM, run the following command. Note that I don't
have any additional columns to the file itself.
CREATE TABLE [dbo].[Files]
(
FileID uniqueidentifier NOT NULL ROWGUIDCOL PRIMARY KEY,
FileName Varchar(300) NOT NULL,
Length bigint
NOT NULL,
FileContents varbinary(max) FILESTREAM DEFAULT NULL
)
4. Add Test Data
In order to get the feeling of the experience of working
with files, let add an empty file:
insert into Files
(FileID,[FileName],Length,FileContents)
values (newid(),'Dummy File 2',0, CAST ('my test file' as varbinary(max)));
If we now query the database to see the files in it:
select * from Files
We will get the following result (the GUID will probably be
different...)
FileID
FileContents
------------------------------------
------------------------------------------------
8247CE78-74DF-4BDE-A08D-9760AE0B1555
my test file
and If we look at the FileManagement directory, we can see
that a new directory has been created
