From: | Neil Conway <neilc(at)samurai(dot)com> |
---|---|
To: | "Marc G(dot) Fournier" <scrappy(at)postgresql(dot)org> |
Cc: | Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us>, Josh Berkus <josh(at)agliodbs(dot)com>, "Arcadius A(dot)" <ahouans(at)sh(dot)cvut(dot)cz>, PostgreSQL Advocacy <pgsql-advocacy(at)postgresql(dot)org> |
Subject: | Re: MySQL interview, no mention of PostgreSQL |
Date: | 2003-10-17 19:35:38 |
Message-ID: | 1066419337.580.44.camel@tokyo |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-advocacy |
On Fri, 2003-10-17 at 15:21, Marc G. Fournier wrote:
> On Fri, 17 Oct 2003, Neil Conway wrote:
> > We do the fork() before we even look at the client's connection packet,
> > so I'm sceptical that going through all that trouble really makes us
> > significantly more resistant to a DOS attack.
>
> The point is that its alot faster for the backend to open a text file and
> parse that
Actually, we don't do that: the file is pre-parsed and already in
memory, so we just binary search through it.
And yes, I realize that this is somewhat faster than doing a system
catalog lookup. I'm just sceptical that the in-memory search is THAT
much faster that it's worth worrying about, considering (a) we need to
do the fork() regardless, which is expensive (b) the content of the
system catalog will almost certainly be in the buffer pool. (The fact
that the code in question is (IMHO) rather ugly doesn't add to my
sympathy for it...)
-Neil
From | Date | Subject | |
---|---|---|---|
Next Message | Bruno LEVEQUE | 2003-10-17 20:07:35 | Re: first version of the PostgreSQL flyer |
Previous Message | Marc G. Fournier | 2003-10-17 19:21:52 | Re: MySQL interview, no mention of PostgreSQL |