Question

Question is old and i guess has no 100% right answer. But would like to hear more experienced advice.

Using SQL Server 2008 R2.

I have table where will be stored millions of rows. Most of the columns is description (date, status, title, ..) of varbinary(max) column data. There are also 2 columns of XML data type. these XML’s are small and will be queried often. So:

   MyTable
   (
        SomeID varchar(20)-- queried most often
        Date DateTime -- queried most often
        Status VarChar(10) -- queried most often
        Title VarChar(50) -- queried most often
        -- some more columns here
        SomeSmallXML xml -- queried quite often 
        SomeOtherSmallXML xml -- queried quite often 
        MyData varbinary(max) -- queried rarely
        MyOtherData varbinary(max) -- queried rarely
    )

IF i move all large value types to other table:

• can do online reindexing of clustered index. But then i have to move also xml types to
  other table. as they are queried quite often, its does not seem
  reasonable. (i expect fragmentation, because SomeID column is coming
  from client app. Its not reasonable to make other surrogate key as
  clustered index, so SomeID will be key of clustered index.)
• can move large data to slower storage. But guess can achieve
  the same by table partitioning (old data in slow filegroup) + indexes
  on fast storage.

In this case do not see very good reasons to move large value data types to other table.
i do see reason to set “sp_tableoption N’MyTable’, ‘large value types out of row’, ‘ON'”.

What is your advice? What else i have to take in account?

Answer 1

I made decision based on discussion with other colleagues: separated SOME of LOB data (also SomeID and Date columns) in separated data in other table.

Most important: i missed to consider update rate of columns and how long data is queried often and when they get old enough not to be interesting in vast majority (but not all) cases.

And that is what makes difference in this case.

So came up with:

  MyTable
   (
        SomeID varchar(20)-- queried most often / Updated never
        Date DateTime -- queried most often / Updated never
        Status VarChar(10) -- queried most often / Updated few times after insert
        Title VarChar(50) -- queried most often / Updated never
        -- some more columns here
        SomeSmallXML xml -- queried quite often /  Updated few times after insert
        SomeOtherSmallXML xml -- queried quite often / Updated never
        MyData varbinary(max) -- queried rarely / Updated never
        MyOtherData varbinary(max) -- queried rarely / Updated 1 shortly after insert
    )

So, as can see, some LOB data- MyData and MyOtherData varbinary(max) gets static after short time. They are large enough, so i would like to store them on cheep disk and put on read only partition at some point of time. As more recent is Date, as more often i need “MyData” or “MyOtherData”.

So final design looks approximately like this:

 MyTable
   (
        SomeID varchar(20)
        Date DateTime 
        Status VarChar(10) 
        Title VarChar(50) 
        -- some more columns here
        SomeSmallXML xml 
        SomeOtherSmallXML xml 
    )
  MyTableLOB
   (
        SomeID varchar(20) 
        Date DateTime -- used for partitioning
        MyData varbinary(max)
        MyOtherData varbinary(max)
    )