This integer column should be unique, så prior to altering it to unique i
need to traverse all rows and update the rows with the same value. This table
has another colun which is the primary key and when 2 rows has the same value
the row with the highest value of the primary key will be updated.
Any easy T-SQL way to do it?

Regards,
Olav

update your_table
set your_column = whatever you wish to set it to
where exists(select 1 from your_table yt
where your_table.your_column = yt.your_column
and your_table.PK_column > yt.PK_column)

Just curious: what kind of information do you have in that column so
that you can easily update it just to enforse uniqueness?

convert flat file to native file format for bcp
Disappearing data (happens rarely)
Query Help: Counts & Sums
Year To Date?
trying to do some weird math in an update statement