| From: | Simon Riggs <simon(at)2ndquadrant(dot)com> |
|---|---|
| To: | Dann Corbit <DCorbit(at)connx(dot)com> |
| Cc: | "Jonah H(dot) Harris" <jonah(dot)harris(at)gmail(dot)com>, Xiao Meng <mx(dot)cogito(at)gmail(dot)com>, pgsql-hackers(at)postgresql(dot)org, Kenneth Marshall <ktm(at)rice(dot)edu> |
| Subject: | Re: [PATCH]-hash index improving |
| Date: | 2008-07-18 07:52:55 |
| Message-ID: | 1216367575.19656.700.camel@ebony.2ndQuadrant |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Thu, 2008-07-17 at 16:37 -0700, Dann Corbit wrote:
> Large table unique index equality search should be very fast with hashed
> index (and the only place where any advantage will be seen). Hashed
> indexes are useless for any search besides equality and gain more and
> more when the levels of the b-tree index increase.
I think a comparison with a btree using a functional index should be
shown.
> The only way to get better performance from hash based indexes is to
> read fewer index pages than if a tree-based index were used. So I think
> that the scheme used to create the index pages is the focus to make them
> worthwhile.
Agreed. Some math on that, plus a clear focus on making this faster than
a btree is critical to this project.
--
Simon Riggs www.2ndQuadrant.com
PostgreSQL Training, Services and Support
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Marko Kreen | 2008-07-18 08:22:30 | Re: TABLE-function patch vs plpgsql |
| Previous Message | Andrew Sullivan | 2008-07-18 06:23:43 | Re: Load spikes on 8.1.11 |