Re: GiST index performance - Mailing list pgsql-performance

From Craig Ringer
Subject Re: GiST index performance
Date
Msg-id 49E7D9C0.90901@postnewspapers.com.au
Whole thread Raw
In response to Re: GiST index performance  (dforum <dforums@vieonet.com>)
List pgsql-performance
dforum wrote:
> hello,
>
> there is other performance problem on this request.
>
> If you analyse query plan, you see that most of the time are lost during
> sequencial scan, and you have 2 seq scan.
>
> You have to create other indexes to match the request.
>
> Postgresq is totally dependant on index to reach is performance.

That depends a lot on your queries. Sometimes a sequential scan is a
faster and better choice. It may also be faster for small tables.

I've usually found that when I (for performance testing purposes) force
the planner to an index scan instead of its preferred sequential scan,
the query runs slower than it did with a sequential scan.

Sure, there are queries that are horrifyingly slow without appropriate
indexes, but I wouldn't go so far as to say that Pg is totally dependent
on indexes to perform well. It depends a lot on the query.

--
Craig Ringer

pgsql-performance by date:

Previous
From: Kris Jurka
Date:
Subject: Re: No hash join across partitioned tables?
Next
From: Vlad Arkhipov
Date:
Subject: Optimizer's issue