Group by improvements?

Asked Dec 14 '16 at 22:38

Active Dec 14 '16 at 22:45

Viewed 50 times

I have a requirement to improve below distinct slow query to use group by(or any other suggestion...)

insert into #temp1
    select distinct col1, '', 'known',col2, col3
    from mytable

Help please.

Index is only on col1 which is int type. Rest of columns are varchar(10)

It's a Sybase query but any SQL compliant code would do(i.e. no CTE). regards.

edited Dec 14 '16 at 22:45

Gordon Linoff

asked Dec 14 '16 at 22:38

Sandiyan

Hi, Welcome to Stack Overflow. I suggest checking out http://stackoverflow.com/help/mcve to help ask a better question. – Eric S Dec 14 '16 at 22:40
1

Adding a compound index on `col1, col2, col3` would likely improve the query performance. – D Stanley Dec 14 '16 at 22:41
4

For a query as simple as that, distinct and group by should use the exact same execution plan... – ZLK Dec 14 '16 at 22:45
using group by will have no effect on that query – Paul Maxwell Dec 14 '16 at 22:46

0 Answers0