Re: SET variables - Mailing list pgsql-hackers

From Bruce Momjian
Subject Re: SET variables
Date
Msg-id 200105140016.f4E0G7e26177@candle.pha.pa.us
Whole thread Raw
In response to Re: SET variables  (mlw <markw@mohawksoft.com>)
List pgsql-hackers
> > Wow, that really does stink.  I know Informix can't have a temp table
> > with the same name as a real table, and I thought that was bad, but
> > Oracle is much worse.
> >
> 
> Although....
> 
> One can see the advantage in a globally shared temporary table. For instance,
> something like user web session management. One can insert and update against
> the temp table and never have to worry about disk I/O or vacuuming. (Assuming a
> temp table is implemented as a memory buffer)

Yes, but having a temp table never hit disk is a different issue from
its visibility.  We could eventually implement the memory-only feature
if we wanted to.  Right now, we have it dumping to disk as a backing
store for the table, assuming it wouldn't fit all in memory.

--  Bruce Momjian                        |  http://candle.pha.pa.us pgman@candle.pha.pa.us               |  (610)
853-3000+  If your life is a hard drive,     |  830 Blythe Avenue +  Christ can be your backup.        |  Drexel Hill,
Pennsylvania19026
 


pgsql-hackers by date:

Previous
From: mlw
Date:
Subject: Re: SET variables
Next
From: "Joe Conway"
Date:
Subject: Isn't pg_statistic a security hole - Solution Proposal