| From: | Shaun Thomas <sthomas(at)optionshouse(dot)com> |
|---|---|
| To: | Thomas Kellerer <spam_eater(at)gmx(dot)net> |
| Cc: | <pgsql-performance(at)postgresql(dot)org> |
| Subject: | Re: Thinking About Correlated Columns (again) |
| Date: | 2013-05-16 12:47:08 |
| Message-ID: | 5194D54C.6030901@optionshouse.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
On 05/15/2013 03:31 PM, Thomas Kellerer wrote:
> What happens if you create one index for each column? (instead of one
> combined index)
I just created the combined index to simplify the output. With two
indexes, it does the usual bitmap index scan. Even then the row estimate
is off by a factor of 1000, which is the actual problem.
The other reason I tried the combined index was that there was some
noise a while back about having the planner check for possible column
correlations on composite indexes. Clearly that ended up not being the
case, but it was worth examination.
--
Shaun Thomas
OptionsHouse | 141 W. Jackson Blvd. | Suite 500 | Chicago IL, 60604
312-676-8870
sthomas(at)optionshouse(dot)com
______________________________________________
See http://www.peak6.com/email_disclaimer/ for terms and conditions related to this email
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Cuong Hoang | 2013-05-16 14:46:10 | Reliability with RAID 10 SSD and Streaming Replication |
| Previous Message | Andrea Suisani | 2013-05-16 08:47:07 | Re: [OT] linux 3.10 kernel will improve ipc,sysv semaphore scalability |