Re: Asking advice on speeding up a big table - Mailing list pgsql-general

From felix-accts-pgsql@crowfix.com
Subject Re: Asking advice on speeding up a big table
Date
Msg-id 20060415154911.GA12672@crowfix.com
Whole thread Raw
In response to Re: Asking advice on speeding up a big table  (Francisco Reyes <lists@stringsutils.com>)
List pgsql-general
On Sat, Apr 15, 2006 at 10:31:26AM -0400, Francisco Reyes wrote:
> felix@crowfix.com writes:
>
> >Usage is to match data from the key and val tables to fetch the data
> >value from the sid table.
>
> What is the relation between key and val tables?
> Will key.id and val.id be equal?

This benchmark explores an idea for a simple berkeley-db-like lookup,
but faster and allowing ranges, and specialized for looking up info in
some other tables I have.  The key table data is table.column, and
1000 is a rough guess on how many unique column names there might be.
The val table is the contents of those columns, and 100K is nother
rough guess.  The end result, the sid table, is a generic ID I have,
coudl be anything, like lat/lon, room-bookshelf-shelf-book, etc.

key.id and val.id have no bearing on each other.

I have made some minor changes and speeded things up to around 15-20
lookups/sec, good enough, but not exciting :-) and in the process,
come across some odd misbehavior.  I have a writeup, almost ready to
post, but I want to make sure I cxross my Is and dot my Ts properly.

--
            ... _._. ._ ._. . _._. ._. ___ .__ ._. . .__. ._ .. ._.
     Felix Finch: scarecrow repairman & rocket surgeon / felix@crowfix.com
  GPG = E987 4493 C860 246C 3B1E  6477 7838 76E9 182E 8151 ITAR license #4933
I've found a solution to Fermat's Last Theorem but I see I've run out of room o

pgsql-general by date:

Previous
From: Andreas Kretschmer
Date:
Subject: Re: inet value validation
Next
From: felix-accts-pgsql@crowfix.com
Date:
Subject: Query runs fast or slow