| From: | Alban Hertroys <alban(at)magproductions(dot)nl> | 
|---|---|
| To: | Postgres General <pgsql-general(at)postgresql(dot)org> | 
| Subject: | Plan for outer joins | 
| Date: | 2006-09-11 09:58:56 | 
| Message-ID: | 45053360.6020906@magproductions.nl | 
| Views: | Whole Thread | Raw Message | Download mbox | Resend email | 
| Thread: | |
| Lists: | pgsql-general | 
Hi,
I was tuning a join on a few tables for a SELECT COUNT(*) query, when I 
realized that the content of the second table didn't actually matter to 
the count. So, I figured a LEFT OUTER JOIN would be faster, but... 
apparently it's not.
Shouldn't the planner notice that the right part of SELECT COUNT(*) FROM 
x LEFT OUTER JOIN y USING (id) isn't relevant for the result and skip 
the join?
Attached are the explain plans for both versions and a few variations 
(TB wraps it if I paste).
Regards,
-- 
Alban Hertroys
alban(at)magproductions(dot)nl
magproductions b.v.
T: ++31(0)534346874
F: ++31(0)534346876
M:
I: www.magproductions.nl
A: Postbus 416
    7500 AK Enschede
// Integrate Your World //
| Attachment | Content-Type | Size | 
|---|---|---|
| explain_joins.txt | text/plain | 3.7 KB | 
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Martijn van Oosterhout | 2006-09-11 10:00:18 | Re: Child program using parent program's transaction? | 
| Previous Message | Purusothaman A | 2006-09-11 09:57:09 | Re: [GENERAL] Problem with lo_export() and lo_import() from remote machine. |