Question

I do an UPDATE request from my client app to the database (SQL Server 2008 R2), and it takes 6 seconds to complete.

Currently I have around 5 millions rows in that table.

Here is the query:

UPDATE MyTable 
SET subject2 = 'information abx', dateProcessed = '2012-02-27 23:02:44' 
WHERE id = 3712028;

Where columns are:

[id] [int] IDENTITY(1,1) NOT NULL,
[dateProcessed] [datetime] NULL,
[subject2] [nvarchar](150) NULL,

Any ways to defragment database of something like that?? 🙂

I don’t think that 6 seconds is normal time to update 1 row searching by indexed field…

Any help is highly appreciated! Thanks,

UPDATE 1: I’m sorry guys! I don’t have indexes in the table. I thought IDENTITY is someth like INDEX also. So probably I should add INDEX to id field or make it PRIMARY KEY

Answer 1

Declaring a column identity does not automatically makes it indexed; declaring it a primary key does:

[id] [int] IDENTITY(1,1) NOT NULL PRIMARY KEY

You can add the constraint to an existing database like this:

ALTER TABLE MyTable ADD PRIMARY KEY (id)