Re: [PERFORM] Big IN() clauses etc : feature proposal - Mailing list pgsql-hackers

From Martijn van Oosterhout
Subject Re: [PERFORM] Big IN() clauses etc : feature proposal
Date
Msg-id 20060509103632.GD29652@svana.org
Whole thread Raw
In response to Re: [PERFORM] Big IN() clauses etc : feature proposal  (PFC <lists@peufeu.com>)
Responses Re: [PERFORM] Big IN() clauses etc : feature proposal
List pgsql-hackers
On Tue, May 09, 2006 at 12:10:37PM +0200, PFC wrote:
>     Yes, but in this case temp tables add too much overhead. I wish
>     there  were RAM based temp tables like in mysql. However I guess the
> current temp  table slowness comes from the need to mark their existence in
> the system  catalogs or something. That's why I proposed using cursors...

It would be interesting to know what the bottleneck is for temp tables
for you. They do not go via the buffer-cache, they are stored in
private memory in the backend, they are not xlogged. Nor flushed to
disk on backend exit. They're about as close to in-memory tables as
you're going to get...

Have a nice day,
--
Martijn van Oosterhout   <kleptog@svana.org>   http://svana.org/kleptog/
> From each according to his ability. To each according to his ability to litigate.

pgsql-hackers by date:

Previous
From: PFC
Date:
Subject: Re: [PERFORM] Big IN() clauses etc : feature proposal
Next
From: PFC
Date:
Subject: Re: [PERFORM] Big IN() clauses etc : feature proposal