Regarding performance

Question

Hi,

 

Back in the old days they said that you should never use select distinct from a database, because the optimizer cannot use index. I want to know if it is some difference in performance between:

select distinct *

with cte as(

   select col1,

               row_number() over (partition by col1 order by col1) as DuplicateCount

   from   table1

) 

select col1 from cte where DuplicateCount = 1;

 

select col1

from   table1

group by col1

 

Maybe it is another way to write an sql statement with best performance regarding select distinct values from on one or more columns.