|
|
Posted by Erland Sommarskog on 03/30/06 00:59
T.H.N. (newsgroup.replies@spam.la) writes:
> I'm trying to work out a database design to make it quicker for my client
> program to read and display updates to the data set. Currently it reads in
> the entire data set again after each change, which was acceptable when the
> data set was small but now it's large enough to start causing noticable
> delays. I've come up with a possible solution but am looking for others'
> input on its suitability to the problem.
I assume that these updates are not performed by your client program,
but by some other process?
Rather than using a datetime column, you could use a timestamp column.
A timestamp column is automatically updated with a monotonically
increasing value each time a row is inserted or updated. Each value is
unique in the database. Note that the name of the type is misleading.
The value is a binary(8) and has to relation to time.
You would still need that table for deleted rows. (Unless you add a
"deleted" bit to the table; the client program would then actually
perform the deletion once it has consumed the update.
--
Erland Sommarskog, SQL Server MVP, esquel@sommarskog.se
Books Online for SQL Server 2005 at
http://www.microsoft.com/technet/prodtechnol/sql/2005/downloads/books.mspx
Books Online for SQL Server 2000 at
http://www.microsoft.com/sql/prodinfo/previousversions/books.mspx
Navigation:
[Reply to this message]
|