Michal Politowski writes:
> Cannot there be a (system/hardware) setup where there is a perceptible
> performance difference between forward and backward index scans?
I think it's been reported already that backward index scans indeed can
be much slower than forward index scan, but that how to mod
On Tue, 2 Nov 2010 10:10:19 -0400, Tom Lane wrote:
> Szymon Guz writes:
> > On 2 November 2010 10:36, AI Rumman wrote:
> >> Is it possible to create an index in descending order?
>
> > create index i on t(i desc);
>
> Note that there is actually no point at all in such a declaration.
> The pla
Szymon Guz writes:
> On 2 November 2010 10:36, AI Rumman wrote:
>> Is it possible to create an index in descending order?
> create index i on t(i desc);
Note that there is actually no point at all in such a declaration.
The planner is perfectly capable of using backwards indexscans at
need, so
On 2 November 2010 12:57, AI Rumman wrote:
> But I am using Postgresql 8.1. Is it possible here?
I am afraid not. You could try to do the index using kind of 1/field
trick but I am not sure if it performs better than backward index scan
in general.
>
> On Tue, Nov 2, 2010 at 3:42 PM, Szymon Guz
But I am using Postgresql 8.1. Is it possible here?
On Tue, Nov 2, 2010 at 3:42 PM, Szymon Guz wrote:
>
>
> On 2 November 2010 10:36, AI Rumman wrote:
>
>> Is it possible to create an index in descending order?
>>
>
> yes...
>
> create index i on t(i desc);
>
>
> regards
> Szymon
>
On 2 November 2010 10:36, AI Rumman wrote:
> Is it possible to create an index in descending order?
>
yes...
create index i on t(i desc);
regards
Szymon
On 2 November 2010 12:36, AI Rumman wrote:
> Is it possible to create an index in descending order?
>
Yes it is -
http://www.postgresql.org/docs/current/interactive/indexes-ordering.html
--
Sergey Konoplev
Blog: http://gray-hemp.blogspot.com /
Linkedin: http://ru.linkedin.com/in/grayhemp /
JI
Is it possible to create an index in descending order?
