From: "Shridhar Daithankar" <shridhar_daithankar@persistent.co.in>
To: pgsql-performance@postgresql.org
Date: Thu, 03 Oct 2002 22:05:24 +0530
MIME-Version: 1.0
Subject: Re: Large databases, performance
Reply-To: shridhar_daithankar@persistent.co.in
Cc: "pankaj M. Tolani" <pankaj@pspl.co.in>
Message-ID: <3D9CBF24.4516.AA1560C@localhost>
References: <3D9CBAD7.23509.A908C61@localhost>
In-reply-to: <1033662394.21324.59.camel@camel>
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
Content-description: Mail message body

On 3 Oct 2002 at 12:26, Robert Treat wrote:

> On Thu, 2002-10-03 at 12:17, Shridhar Daithankar wrote:
> > On 3 Oct 2002 at 11:57, Robert Treat wrote:
> > May be it's time to rewrite famous myth that postgresql is slow. 
> 
> That myth has been dis-proven long ago, it just takes awhile for
> everyone to catch on ;-)

:-)

> Hmm... been awhile since I dug into mysql internals, but IIRC once the
> table was locked, you had to wait for the insert to complete so the
> table would be unlocked and the select could go through. (maybe this is
> a myth that I need to get clued in on)

If that turns out to be true, I guess mysql will nose dive out of window.. May 
be time to run a test that's nearer to real world expectation, especially in 
terms on concurrency..

I don't think tat will be an issue with mysql with transaction support. The 
vanilla one might suffer.. Not the other one.. At least theoretically..

> My thinking was that if your just doing inserts, you need to update the
> statistics but don't need to check on unused tuples. 

Any other way of doing that other than vacuum analyze? I thought that was the 
only way..

Bye
 Shridhar

--
"Even more amazing was the realization that God has Internet access.  Iwonder 
if He has a full newsfeed?"(By Matt Welsh)