>> Are you suggesting that I:
>> a. use a different datatype in plpgsql or
>> b. cast the values or
>> c. change the data type in the table
>Yes ;-). Presumably (b) would be the least painful route, but any of
>these would do the trick.
It turns out that the text and char(5) were not the problem - it was
an integer field that was passed into the function as a text (the
proximity field)! It now utilizes the index and has gone from taking
over 7 seconds to complete to .011 seconds!
I love postgres!
In response to
pgsql-hackers by date
|Next:||From: Darren Johnson||Date: 2002-12-22 01:58:58|
|Subject: Re: PostgreSQL-R|
|Previous:||From: Tom Lane||Date: 2002-12-21 16:40:03|
|Subject: Re: Resource management in 7.4 |