Re: Any better plan for this query?..

On 5/12/09, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> On Tue, May 12, 2009 at 1:00 PM, Dimitri <dimitrik(dot)fr(at)gmail(dot)com> wrote:
>> On MySQL there is no changes if I set the number of sessions in the
>> config file to 400 or to 2000 - for 2000 it'll just allocate more
>> memory.
>
> I don't care whether the setting affects the speed of MySQL. I want
> to know if it affects the speed of PostgreSQL.

the problem is they both have "max_connections" parameter, so as you
asked for MySQL I answered for MySQL, did not test yet for PostgreSQL,
will be in the next series..

>
>> After latest fix with default_statistics_target=5, version 8.3.7 is
>> running as fast as 8.4, even 8.4 is little little bit slower.
>>
>> I understand your position with a pooler, but I also want you think
>> about idea that 128 cores system will become a commodity server very
>> soon, and to use these cores on their full power you'll need a
>> database engine capable to run 256 users without pooler, because a
>> pooler will not help you here anymore..
>
> So what? People with 128-core systems will not be running trivial
> joins that return in 1-2ms and have one second think times between
> them. And if they are, and if they have nothing better to do than
> worry about whether MySQL can process those queries in 1/2000th of the
> think time rather than 1/1000th of the think time, then they can use
> MySQL. If we're going to worry about performance on 128-core system,
> we would be much better advised to put our efforts into parallel query
> execution than how many microseconds it takes to execute very simple
> queries.

Do you really think nowdays for example a web forum application having
PG as a backend will have queries running slower than 1-2ms to print a
thread message within your browser??? or banking transactions??

>
> Still, I have no problem with making PostgreSQL faster in the case
> you're describing. I'm just not interested in doing it on my own time
> for free. I am sure there are a number of people who read this list
> regularly who would be willing to do it for money, though. Maybe even
> me. :-)
>
> ...Robert
>

You don't need to believe me, but I'm doing it for free - I still have
my work to finish in parallel :-)) And on the same time I don't see
any other way to learn and improve my knowledge, but nobody is perfect
:-))

Rgds,
-Dimitri