2010/6/30 Thom Brown <thombrown(at)gmail(dot)com>
> On 30 June 2010 19:43, Joshua D. Drake <jd(at)commandprompt(dot)com> wrote:
> > On Wed, 2010-06-30 at 20:42 +0200, Szymon Guz wrote:
> >> Hi,
> >> in a PHP application working on Postgres normally the new connection
> >> to the database is made per request.
> >> This can potentially cause too big overhead, so I've got some
> >> questions:
> >> - is the overhead really noticeable?
> > It can be.
> >> - could this be solved using persistent connections, or the persistent
> >> connections in php and postgres don't work properly?
> > Don't use them.
> >> - could this be solved using something like pgpool?
> > Yes, using a connection pooler will solve the problem. I prefer
> > pgbouncer.
> I can't really add to that.
Thanks for the answers that ensured me in what I was thinking about.
In response to
pgsql-general by date
|Next:||From: Zoid||Date: 2010-06-30 19:08:38|
|Subject: LINE 1: IDENTIFY_SYSTEM error infinitum|
|Previous:||From: MD||Date: 2010-06-30 18:53:10|
|Subject: Re: Looking for multithreaded build of libpq for Windows|