Re: speed up select distinct

creating index on table cause slower insert, update, delete (dml statements)...so i wouldn't create them, instead of them i'm looking for different syntax of query resulting the same goal...only dinstinct values of column1, column2... etc...

"Tomislav Kokoska" <tkokoska_at_varteks.com> wrote in message news:d5sb22$huo$1_at_ss405.t-com.hr...
> table TABLE has about 5 000 0000 rows, (4438252 rows) and non unique index
> on column1, column2, column3 with cardinality 254241..
>
> column5 has cardinality about 20 so it is expect that bitmap index will
> surely increase perfomance (where clause). What about distinct? how we
> could achive
> better performance. creating another index on selected distinct column...
> ??
>
>
> oracle 9.2.0.5.0
>
>
>
>
>
>
>
>
>
> ----- Original Message -----
> From: "Dave" <solomons_dad.w.marks_and_whom_at_oracle.com>
> Newsgroups: comp.databases.oracle.misc
> Sent: Tuesday, May 10, 2005 12:05 PM
> Subject: Re: speed up select distinct
>
>
>>
>>
>> Tomislav Kokoska wrote:
>>> What is the best way to speed up query looks like:
>>>
>>> SELECT DISTINCT COLUMN1, COLUMN2, COLUMN3, COLUMN4 FROM TABLE WHERE
>>> COLUMN5 = 5
>>>
>>>
>>>
>>
>> Define "best."
>
> "Tomislav Kokoska" <tkokoska_at_varteks.com> wrote in message
> news:d5pniu$97$1_at_ss405.t-com.hr...
>> What is the best way to speed up query looks like:
>>
>> SELECT DISTINCT COLUMN1, COLUMN2, COLUMN3, COLUMN4 FROM TABLE WHERE
>> COLUMN5 = 5
>>
>>
>>
>
>
Received on Thu May 12 2005 - 08:30:07 CDT