Split table

[Carlo Martinez]

Hi all,
I have to store the filesize of each file on a users directory, for example:

C:\Users\marc\file1.txt 2500
C:\Users\marc\file2.txt 3270

and the most executed query are:
SELECT filesize FROM files WHERE filename = 'C:\Users\marc\file1.txt'
or
SELECT filesize FROM files WHERE filename LIKE 'C:\Users\marc\file%'

Index on filename field of course.
filename UNIQUE

Now I'm thinking to split the user part of the filename on a separate table:

Table Users:
Id, Username

Table Files:
UserId, Filename, Filesize

So the query become:
SELECT filesize FROM files WHERE filename = 'file1.txt' AND UserId = 'xx'
Index on filename and UserId
filename no UNIQUE

Any suggestion about perfromance gain/loss of this change ?

Table have about 1,5M rows (800 users)

Thanks
Carip