| From: | Michael Fuhr <mike(at)fuhr(dot)org> |
|---|---|
| To: | Matthew Sackman <matthew(at)lshift(dot)net> |
| Cc: | pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Massive performance issues |
| Date: | 2005-09-01 22:34:36 |
| Message-ID: | 20050901223436.GA44175@winnie.fuhr.org |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
On Thu, Sep 01, 2005 at 06:42:31PM +0100, Matthew Sackman wrote:
>
> "address_pc_top_index" btree (postcode_top)
> "address_pc_top_middle_bottom_index" btree (postcode_top,
> postcode_middle, postcode_bottom)
> "address_pc_top_middle_index" btree (postcode_top, postcode_middle)
This doesn't address the query performance problem, but isn't only
one of these indexes necessary? The second one, on all three
columns, because searches involving only postcode_top or only
postcode_top and postcode_middle could use it, making the indexes
on only those columns superfluous. Or am I missing something?
--
Michael Fuhr
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Dan Harris | 2005-09-01 23:02:49 | Re: Poor performance on HP Package Cluster |
| Previous Message | Matthew Sackman | 2005-09-01 22:22:39 | Re: Massive performance issues |