From: | "Francisco Reyes" <lists(at)stringsutils(dot)com> |
---|---|
To: | <pgsql-general(at)postgresql(dot)org>, <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Subject: | Re: Any way to favor index scans, but not bitmap index scans? |
Date: | 2008-07-23 20:44:42 |
Message-ID: | c76fec09f01dcea1bea994d67b67c067@stringsutils.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On 4:12 pm 07/23/08 "Francisco Reyes" <lists(at)stringsutils(dot)com> wrote:
> Also, that plan is only 3 seconds.
Minor update.
A co-worker is using another DB.. and re-running my query after he did his
work.. now the query using the index scans takes 2 minutes instead of 3
seconds. 3 seconds was likely data cached.
To re-list the times..
Sequential scan 12 minutes
Bitmap scans 30 minutes
index scan with not bitmap 2 minutes
It is worth pointing out that the bitmap test was run AFTER the sequential
scan test.. right after it.. so it should have benefited from OS caching.
The join_ids table fits completely in memory.
select pg_size_pretty(pg_total_relation_size('join_ids'));
pg_size_pretty
----------------
291 MB
(1 row)
par4mo=# select pg_size_pretty(pg_relation_size('join_ids'));
pg_size_pretty
----------------
94 MB
(1 row)
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2008-07-23 20:47:34 | Re: mac install question |
Previous Message | Oleg Bartunov | 2008-07-23 20:28:46 | Re: [GENERAL] Fragments in tsearch2 headline |