You can specify the collation at database level or table level, see BOL ---> COLLATE clause

See if this helps

http://www.sqlteam.com/article/introduction-to-locking-in-sql-server

Which is your isolation level?

See if this helps

http://www.sqlservercentral.com/articles/Advanced/designingcrossdatabasequeries/1753/

(see the "security" section).

Enable the "db" role to log in to DB2 and grant the necessary rights to the base table(s).

To list permissions for a principal on...

The amount of fragmentation depends on the table and indexes (clustered or not) structure.

Kimberly Tripp's blog on sqlskills.com offers plenty of detail.

Do you index your XML data?

See if this helps

http://msdn.microsoft.com/en-us/library/ms345118%28SQL.90%29.aspx

Do you store that column as varchar(MAX), or as XML data type?

I'm not sure that I understand your environment (for example I see that there are a MySQL database and a SQL Server database involved (?)) and your requirements. Can you...

Using an "application role" may be helpful in this case.

Alternatively, make your application access the database through a user which has no other rights than the "EXEC" right on the...

Try to delete chunks of records, see

http://www.sqlservercentral.com/articles/T-SQL/67898/

See if this helps:

http://research.microsoft.com/apps/pubs/default.aspx?id=64525

You're welcome...

Check the execution plan of those queries, see why they are slow. Indexing may be your problem.

Is your database set to auto update statistics? Are you aware of any unusual...

SQL 2008 (you may consider upgrading) offers the "transparent data encryption", which protects the data "at rest":

http://blogs.msdn.com/lcris/archive/2007/10/03/sql-server-2008-transparent-data-encryption-feature-a-quick-overview.aspx

http://www.mssqltips.com/tip.asp?tip=1514

The users in the db should have no other rights than the "EXEC" rights...

See if this helps. Look for the UNPIVOT usage.

http://technet.microsoft.com/en-us/library/ms177410.aspx

I had recently the same problem. The main "culprit" was a query which executed repeatedly (each time a product detail page was visited) and required thousands of physical reads. Reviewing...