Re: Postgres performance

--- PFC <lists(at)boutiquenumerique(dot)com> wrote:
> You mean, you have no foreign keys in your database
?
> In SELECT they are definitely useful (think select
> for update, isolation
> level serializable...)
No, I haven't foreign keys in the older version, in
that new I've it... however I manage relations from
app code (PHP)... doesn't MYSQL allow to use 'foreign
keys' in sure and fast way then?

> > - why postgres use a new process for every query ?
> > (mySQL, if I'm not wrong, use threads... I think
> its
> > faster)
>
> Not for every query, for every CONNECTION.
> You are using persistant connections are you. Are
you ?
I'm using PHP and every user (can be from 1 user to
100 users) must connect to the database... do you know
how I can use persistant connection? I think it's
impossible... I'm wrong?
>
> > - why connection time is slower? (compared to
> mySQL)?
>
> This is of no importance as everyone uses
> persistent connections anyway.
See last answer...

> > - why postgres require analyze? (mySQL, if I'm not
> > wrong, don't require it)
> > Yours answers will be very apreciated! Thx
>
> So it has a planner which knows what it's doing ;)
> instead of just
> guessing in the dark.
>
> And MySQL requires analyze too (read the docs),
> optimize table which
> looks like vacuum to me, and sometimes repair
> table...
Ok... they are conceptually implemented in the same
mode...
Thx a lot! Now my ideas about pg & mySQL are cleared
(and corrects ;) )
Mauro

___________________________________
Nuovo Yahoo! Messenger: E' molto più divertente: Audibles, Avatar, Webcam, Giochi, Rubrica… Scaricalo ora!
http://it.messenger.yahoo.it